Product of Industry
1. The problems of the distribution of the product of industry in long periods are usually approached on the assumption of long-run equilibrium. It is tacitly assumed that it is the distribution corresponding to long-run equilibrium which dominates the picture if cyclical fluctuations are eliminated. And the long-period changes in the distribution of the product of industry are accordingly attributed to those in the basic parameters of long-run equilibrium such as the rate of interest and the state of technical knowledge.
This approach might be considered correct if our economy were quasi-stationary--i.e. if all its 'dynamics' consisted of cyclical fluctuations round a certain stationary level. Once, however, we realize that what we call more or less vaguely a 'trend' must also be considered, the matter becomes very complicated. Indeed, the existence of a 'secular trend' actually presupposes that long-run equilibrium in a strict sense (i.e. in the sense of the system's being at rest) is never reached. To that it may be replied that, if the trend is uniform, the modifications required in the equilibrium theory can be easily introduced. This is quite true, but why should the trend be uniform? Secular trend is at least as complex a dynamic process as the business cycle, and no simplifications representing it as a 'natural growth', etc., will do. If we admit this fact, the usefulness of the long-run equilibrium theory of distribution becomes rather doubtful.
In section III below we give an example showing how remote the real long-period dynamic process may be from the pattern of long-run equilibrium: in the period 1899-1923 the rate of profit in US manufacturing fell drastically, while the long-term rate of interest showed a steady rise.
2. If, however, we reject the long-run equilibrium theory of distribution as unrealistic, what else may be put in its place? It is clear that the problem becomes indeterminate unless some additional parameter