Pure Profit: Innovation, Uncertainty and Market Power
The neoclassical theories covered in the preceding three chapters are all theories of interest. The profits analysed by these theories are profits in competitive equilibria and assume that uncertainty is of no importance. In such a context no agent will pay or receive a premium above the ruling rate of interest, nor will any agent pay or receive interest at a lower rate. There are zero pure profits. The status of such theories is a controversial issue among neoclassical economists. Disputes arise over a number of matters, including the relevance of the competitive assumption, how likely it is that economies will operate in, or close to, equilibria, and the significance which various types of uncertainty have. However, even if one takes the view that competition is limited, that uncertainty results in very different economic patterns, and that economies are rarely close to equilibria of supplies and demands, the theories of competitive equilibria in a context of certainty are still useful. They provide a benchmark with which to assess the importance of any deviation from the assumptions upon which they rest. Indeed, the principal theorists of pure profit, Schumpeter (1912; 1939) and Knight (1921), both developed their analyses in this way.
Unable to display preview. Download preview PDF.
Notes to Chapter 13
- 33.However, Knight (1921) does recognise continually that his bold dichotomies, in fact, cannot be easily made. He accepts that all knowledge is partial, that the estimated frequencies of future outcomes can never be really ‘objective’ and subject to precise measurement (pp. 199, 223, 227, 231). Measurement is a matter of degree (pp. 246–7) and more or less informed estimates of consequences resulting from any action can always be made (p. 227). Insurance services are available even when estimates of future outcomes have to build upon a very slender past experience (p. 250). The entrepreneurial activity of uncertainty-bearing is split up and combined with other economic functions (pp. 300, 304, 307, 350, 355). No income is purely contractual or ‘certain’ and pure profit forms a component of virtually all incomes (pp. 272, 277–8, 290, 366). Yet despite all this fudging, Knight still maintains that the central categories, on which his theory of pure profit is based, retain their validity. Subsequent work in the theory of uncertainty has tended to jettison them. This is particularly the case with respect to Knight’s pivotal distinction between risk and uncertainty. For example, Hicks (1931, p. 175) writes: Professor Knight’s doctrine of ‘measurable risks’ is one of the parts of his teaching that I am quite unable to accept, at any rate in the uncompromising form in which he first states it (Risk, Uncertainty and Profit, pp. 43ff). It is quite true that there are certain kinds of risk that are practically eliminated in a business of reasonable size — Mangoldt’s ‘Champagnerfabrikant’ with his broken bottles is the classical example of this. Experience has shown that the chance of failure is expressible by a definite fraction. But even here the possibility of elimination depends on the size of the business. It will not necessarily be desirable to extend a business beyond what would from other points of view be the optimum size in order to eliminate completely a small risk. Nor will it necessarily be worth while to eliminate such a risk by insurance. Insurance involves costs of administration and it is once more a question of balancing advantages whether these costs should be insured or not. Further, the grouping of measurable risks is simply a limiting case, and not a very important one, of the general principle of reduction. Reduction is applicable even when experience does not give us sufficient ground for a knowledge of the exact chances. Even Professor Knight himself admits this (ibid, p. 239) and the whole case has been admirably stated by Professor Hardy: ‘All applications of the law of averages rest on a grouping of things, unlike in many respects, into classes, on the basis of certain similarities; if cases nearly alike are infrequent, we must do our grouping on the basis of less homogeneous classes. If the classification is crude, or if the cases are not numerous, the statistical method loses its accuracy. But these cases certainly shade off into Professor Knight’s “true uncertainties” by imperceptible degrees, the margin of error getting larger as the evidence gets more scanty.’ [G. O. Hardy, Readings in Risk and Risk Bearing, University of Chicago Press, 1924, p. 55] See also Arrow (1971b, pp. 1–43), Weston (1954), and Friedman (1976, pp. 279–82). The criticisms made of Knight in this text do not, however, depend upon rejecting his categories.Google Scholar