Abstract

This paper provides a synthesis of recent contributions to the literature of equilibrium diffusion of technological change, points out a difficulty common to all, and offers a proposal to resolve the difficulty.

Diffusion can be regarded as the result of processes involving changing circumstances in an environment of heterogeneous agents. If diffusion arises from simultaneous operation of multiple processes, the ceteris paribus qualification necessary for theoretical work may be stringent; for empirical work, parameters of individual processes cannot be identified without specifying a multiple process model.

But non-trivial multiple process models are generally analytically intractable. The paper's synthesis of the literature is based on a heuristic multi-process equilibrium model of the adoption decision and proposes numerical simulation as an avenue to escape the related problems of tractability and identification. A numerical experiment with a prototype multi-process simulation model suggests the possible importance of interaction among processes. Recent econometric advances offer new prospects for estimation of the parameters of such models.