Incorporating Technological Change in Diffusion Models

Abstract

The theoretical and empirical implications of a static and dynamic logistic diffusion model are compared. The dynamic model relaxes some assumptions of the static model by allowing for a flexible adoption ceiling, for changes in the technology, and for disadoption. Both models were used to estimate the diffusion of semi‐dwarf wheat varieties in the United States. The dynamic model provides a better fit to the data and provides additional insights into the economic determinants of adoption. In particular, the importance of technological change in other areas (here, in fertilizer) on the diffusion path of an innovation was shown.