The Growth of Demand for New Commodities

Abstract

In this paper we consider the use of the logistic curve in analyses of the demand for new commodities and in studies of innovation. We demonstrate that the theoretical basis for the curve is weak and that the results obtained in empirical analyses have sometimes been unsatisfactory. The choice of a symmetric curve seems unjustified on theoretical grounds, and empirical evidence indicates that a positively skew curve would be preferable. We then present a cross-section analysis of the growth of British television ownership, using a model based upon the cumulative lognormal curve. The results obtained are intuitively plausible and there is no evidence to suggest that a positively skew curve provides an inadequate explanation of the data.