C.1.2 (An Anarchist FAQ)

C.1.2 Is economics a science? is the third chapter of Section C of An Anarchist FAQ.

Transcript

In a word, no. If by “scientific” it is meant in the usual sense of being based on empirical observation and on developing an analysis that was consistent with and made sense of the data, then most forms of economics are not a science. Rather than base itself on a study of reality and the generalisation of theory based on the data gathered, economics has almost always been based on generating theories rooted on whatever assumptions were required to make the theory work. Empirical confirmation, if it happens at all, is usually done decades later and if the facts contradict the economics, so much the worse for the facts.

A classic example of this is the neo-classical theory of production. As noted previously, neoclassical economics is focused on individual evaluations of existing products and, unsurprisingly, economics is indelibly marked by “the dominance of a theoretical vision that treats the inner workings of the production process as a ‘black box.’” This means that the “neoclassical theory of the ‘capitalist’ economy makes no qualitative distinction between the corporate enterprise that employs tens of thousands of people and the small family undertaking that does no employ any wage labour at all. As far as theory is concerned, it is technology and market forces, not structures of social power, that govern the activities of corporate capitalists and petty proprietors alike.” [William Lazonick, Competitive Advantage on the Shop Floor, p. 34 and pp. 33–4]

Production in this schema just happens — inputs go in, outputs go out — and what happens inside is considered irrelevant, a technical issue independent of the social relationships those who do the actual production form between themselves — and the conflicts that ensure. The theory does have a few key assumptions associated with it, however. First, there are diminishing returns. This plays a central role. In mainstream diminishing returns are required to produce a downward sloping demand curve for a given factor. Second, there is a rising supply curve based on rising marginal costs produced by diminishing returns. The average variable cost curve for a firm is assumed to be U-shaped, the result of first increasing and then diminishing returns. These are logically necessary for the neo-classical theory to work.

Non-economists would, of course, think that these assumptions are generalisations based on empirical evidence. However, they are not. Take the U-shaped average cost curve. This was simply invented by A. C. Pigou, “a loyal disciple of [leading neo-classical Alfred] Marshall and quite innocent of any knowledge of industry. He therefore constructed a U-shaped average cost curve for a firm, showing economies of scale up to a certain size and rising costs beyond it.” [[[Joan Robinson]], Collected Economic Papers, vol. 5, p. 11] The invention was driven by need of the theory, not the facts. With increasing returns to scale, then large firms would have cost advantages against small ones and would drive them out of business in competition.

This would destroy the concept of perfect competition. However, the invention of the average cost curve allowed the theory to work as “proved” that a competitive market could not become dominated by a few large firms, as feared. The model, in other words, was adjusted to ensure that it produced the desired result rather than reflect reality. The theory was required to prove that markets remained competitive and the existence of diminishing marginal returns to scale of production did tend by itself to limit the size of individual firms. That markets did become dominated by a few large firms was neither here nor there. It did not happen in theory and, consequently, that was the important thing and so “when the great concentrations of power in the multinational corporations are bringing the age of national employment policy to an end, the text books are still illustrated by U-shaped curves showing the limitation on the size of firms in a perfectly competitive market.” [Joan Robinson, Contributions to Modern Economics, p. 5]

To be good, a theory must have two attributes: They accurately describe the phenomena in question and they make accurate predictions. Neither holds for Pigou’s invention: reality keeps getting in the way. Not only did the rise of a few large firms dominating markets indirectly show that the theory was nonsense, when empirical testing was finally done decades after the theory was proposed it showed that in most cases the opposite is the case: that there were constant or even falling costs in production. Just as the theories of marginality and diminishing marginal returns taking over economics, the real world was showing how wrong it was with the rise of corporations across the world. So the reason why the market become dominated by a few firms should be obvious enough: actual corporate price is utterly different from the economic theory.

This was discovered when researchers did what the original theorists did not think was relevant: they actually asked firms what they did and the researchers consistently found that, for the vast majority of manufacturing firms their average costs of production declined as output rose, their marginal costs were always well below their average costs, and substantially smaller than ‘marginal revenue’, and the concept of a ‘demand curve’ (and therefore its derivative ‘marginal revenue’) was simply irrelevant. Unsurprisingly, real firms set their prices prior to sales, based on a mark-up on costs at a target rate of output.

In other words, they did not passively react to the market. These prices are an essential feature of capitalism as prices are set to maintain the long-term viability of the firm. This, and the underlying reality that per-unit costs fell as output levels rose, resulted in far more stable prices than were predicted by traditional economic theory. One researcher concluded that administered prices “differ so sharply from the behaviour to be expected from” the theory “as to challenge the basic conclusions” of it. He warned that until such time as “economic theory can explain and take into account the implications” of this empirical data, “it provides a poor basis for public policy.” Needless to say, this did not disturb neo-classical economists or stop them providing public policy recommendations. [Gardiner C. Means, “The Administered-Price Thesis Reconfirmed”,The American Economic Review, pp. 292–306, Vol. 62, No. 3, p. 304]

One study in 1952 showed firms a range of hypothetical cost curves, and asked firms which ones most closely approximated their own costs. Over 90% of firms chose a graph with a declining average cost rather than one showing the conventional economic theory of rising marginal costs. These firms faced declining average cost, and their marginal revenues were much greater than marginal cost at all levels of output. Unsurprisingly, the study’s authors concluded if this sample was typical then it was “obvious that short-run marginal price theory should be revised in the light of reality.” We are still waiting. [Eiteman and Guthrie, “The Shape of the Average Cost Curve”, The American Economic Review, pp. 832–8, Vol. 42, No. 5, p. 838]

A more recent study of the empirical data came to the same conclusions, arguing that it is “overwhelming bad news ... for economic theory.” While economists treat rising marginal cost as the rule, 89% of firms in the study reported marginal costs which were either constant or declined with output. As for price elasticity, it is not a vital operational concept for corporations. In other words, the “firms that sell 40 percent of GDP believe their demand is totally insensitive to price” while “only about one-sixth of GDP is sold under conditions of elastic demand.” [A.S. Blinder, E. Cabetti, D. Lebow and J. Rudd, Asking About Prices, p. 102 and p. 101]

Thus empirical research has concluded that actual price setting has nothing to do with clearing the market by equating market supply to market demand (i.e. what economic theory sees as the role of prices). Rather, prices are set to enable the firm to continue as a going concern and equating supply and demand in any arbitrary period of time is irrelevant to a firm which hopes to exist for the indefinite future. As Lee put it, basing himself on extensive use of empirical research, “market prices are not market-clearing or profit-maximising prices, but rather are enterprise-, and hence transaction-reproducing prices.” Rather than a non-existent equilibrium or profit maximisation at a given moment determining prices, the market price is “set and the market managed for the purpose of ensuring continual transactions for those enterprises in the market, that is for the benefit of the business leaders and their enterprises.” A significant proportion of goods have prices based on mark-up, normal cost and target rate of return pricing procedures and are relatively stable over time. Thus “the existence of stable, administered market prices implies that the markets in which they exist are not organised like auction markets or like the early retail markets and oriental bazaars” as imagined in mainstream economic ideology. [Frederic S. Lee, Post Keynesian Price Theory, p. 228 and p. 212]

Unsurprisingly, most of these researchers were highly critical the conventional economic theory of markets and price setting. One viewed the economists’ concepts of perfect competition and monopoly as virtual nonsense and “the product of the itching imaginations of uninformed and inexperienced armchair theorisers.” [Tucker, quoted by Lee, Op. Cit., p. 73f] Which was exactly how it was produced. No other science would think it appropriate to develop theory utterly independently of phenomenon under analysis. No other science would wait decades before testing a theory against reality. No other science would then simply ignore the facts which utterly contradicted the theory and continue to teach that theory as if it were a valid generalisation of the facts. But, then, economics is not a science. This strange perspective makes sense once it is realised how key the notion of diminishing costs is to economics. In fact, if the assumption of increasing marginal costs is abandoned then so is perfect competition and “the basis of which economic laws can be constructed ... is shorn away,” causing the “wreckage of the greater part of general equilibrium theory.” This will have “a very destructive consequence for economic theory,” in the words of one leading neo-classical economist. [John Hicks, Value and Capital, pp. 83–4] As Steve Keen notes, this is extremely significant: <quote> “Strange as it may seem ... this is a very big deal. If marginal returns are constant rather than falling, then the neo-classical explanation of everything collapses. Not only can economic theory no longer explain how much a firm produces, it can explain nothing else. “Take, for example, the economic theory of employment and wage determination ... The theory asserts that the real wage is equivalent to the marginal product of labour ... An employer will employ an additional worker if the amount the worker adds to output — the worker’s marginal product — exceeds the real wage ... [This] explains the economic predilection for blaming everything on wages being too high — neo-classical economics can be summed up, as [John Kenneth] Galbraith once remarked, in the twin propositions that the poor don’t work hard enough because they’re paid too much, and the rich don’t work hard enough because they’re not paid enough ... “If in fact the output to employment relationship is relatively constant, then the neo-classical explanation for employment and output determination collapses. With a flat production function, the marginal product of labour will be constant, and it will never intersect the real wage. The output of the form then can’t be explained by the cost of employing labour... [This means that] neo-classical economics simply cannot explain anything: neither the level of employment, nor output, nor, ultimately, what determines the real wage ...the entire edifice of economics collapses.” [Debunking Economics, pp. 76–7] </quote> It should be noted that the empirical research simply confirmed an earlier critique of neo-classical economics presented by Piero Sraffa in 1926. He argued that while the neo-classical model of production works in theory only if we accept its assumptions. If those assumptions do not apply in practice, then it is irrelevant. He therefore “focussed upon the economic assumptions that there were ‘factors of production’ which were fixed in the short run, and that supply and demand were independent of each other. He argued that these two assumptions could be fulfilled simultaneously. In circumstances where it was valid to say some factor of production was fixed in the short term, supply and demand could not independent, so that every point on the supply curve would be associated with a different demand curve. On the other hand, in circumstances where supply and demand could justifiably be treated as independent, then it would be impossible for any factor of production to be fixed. Hence the marginal costs of production would be constant.” He stressed firms would have to be irrational to act otherwise, foregoing the chance to make profits simply to allow economists to build their models of how they should act. [Keen, Op. Cit., pp. 66–72] Another key problem in economics is that of time. This has been known, and admitted, by economists for some time. Marshall, for example, stated that “the element of time” was “the source of many of the greatest difficulties of economics.” [Principles of Economics, p. 109] The founder of general equilibrium theory, Walras, recognised that the passage of time wrecked his whole model and stated that we “shall resolve the ... difficulty purely and simply by ignoring the time element at this point.” This was due, in part, because production “requires a certain lapse of time.” [Elements of Pure Economics, p. 242] This was generalised by Gerard Debreu (in his Nobel Prize for economics winning Theory of Value ) who postulated that everyone makes their sales and purchases for all time in one instant. Thus the cutting edge of neo-classical economics, general equilibrium ignores both time and production. It is based on making time stop, looking at finished goods, getting individuals to bid for them and, once all goods are at equilibrium, allowing the transactions to take place. For Walras, this was for a certain moment of time and was repeated, for his followers it happened once for all eternity. This is obviously not the way markets work in the real world and, consequently, the dominant branch of economics is hardly scientific. Sadly, the notion of individuals having full knowledge of both now and the future crops up with alarming regularly in the “science” of economics. Even if we ignore such minor issues as empirical evidence and time, economics has problems even with its favoured tool, mathematics. As Steve Keen has indicated, economists have “obscured reality using mathematics because they have practised mathematics badly, and because they have not realised the limits of mathematics.” indeed, there are “numerous theorems in economics that reply upon mathematically fallacious propositions.” [Op. Cit., p. 258 and p. 259] For a theory born from the desire to apply calculus to economics, this is deeply ironic. As an example, Keen points to the theory of perfect competition which assumes that while the demand curve for the market as a whole is downward sloping, an individual firm in perfect competition is so small that it cannot affect the market price and, consequently, faces a horizontal demand curve. Which is utterly impossible. In other words, economics breaks the laws of mathematics. These are just two examples, there are many, many more. However, these two are pretty fundamental to the whole edifice of modern economic theory. Much, if not most, of mainstream economics is based upon theories which have little or no relation to reality. Kropotkin’s dismissal of “the metaphysical definitions of the academical economists” is as applicable today. [Evolution and Environment, p. 92] Little wonder dissident economist Nicholas Kaldor argued that: <quote> “The Walrasian [i.e. general] equilibrium theory is a highly developed intellectual system, much refined and elaborated by mathematical economists since World War II — an intellectual experiment ... But it does not constitute a scientific hypothesis, like Einstein’s theory of relativity or Newton’s law of gravitation, in that its basic assumptions are axiomatic and not empirical, and no specific methods have been put forward by which the validity or relevance of its results could be tested. The assumptions make assertions about reality in their implications, but these are not founded on direct observation, and, in the opinion of practitioners of the theory at any rate, they cannot be contradicted by observation or experiment.” [The Essential Kaldor, p. 416] </quote>