Adoption and diffusion of an innovation of uncertain profitability

Abstract

The argument has frequently been made that the pattern of diffusion associated with most new innovations will typically have certain characteristics. A diffusion curve for an innovation is usually defined as the proportion of its potential users who have already adopted as a function of time (measured from the first adoption). Many empirical studies by both economists and other social scientists have shown that diffusion curves nearly always have two distinct characteristics: (i) they are “S-shaped” (the proportion adopted is an increasing function of time which is initially convex but eventually becomes concave); and (ii) they tend to be right-hand skewed (the inflection point occurs at a time corresponding to a proportion adopted which is less than one half, so that the function is concave over the greatest amount of time; in fact, this skewing is occasionally severe enough to make the diffusion curve appear concave everywhere). ’ The focus of most empirical studies of innovation diffusion has centered on identification of either the major determinants of the speed of diffusion or the characteristics of firms which determine how long they delay adoption.* The seminal contributions are Griliches’ [6] study of hybrid corn and Mansfield’s [ 121 study of 12 industrial innovations. These studies, and most of the similar ones following, conclude that diffusion is faster for innovations with higher profitability and that large firms tend to be the earliest adopters. Some studies (e.g., Mansfield [ 121, Herregat [ 71) have attempted to ascertain