Unveiling diffusion dynamics: an autocatalytic percolation model of environmental innovation diffusion and the optimal dynamic path of adoption subsidies

This article applies the autocatalytic percolation model developed by Cantono and Solomon (2010) to the diffusion of environmental innovation. It contributes to the recent applied microeconomic diffusion literature by unveiling diffusion dynamics, by determining under what conditions is diffusion self-sustaining and by defining the optimal dynamic schedule of adoption subsidies which insures autonomous propagation. To this end a model which combines in a unique framework a learning curve model of dynamic cost reductions, a discrete choice model of heterogeneous technology adoption and a contagion model of technology diffusion is developed. It is shown that the system dynamics are discontinuous, path-dependent and irreversible. Propagation dynamics are uncovered: diffusion occurs along subsequent conquers of islands of potential adopters. Under certain circumstances diffusion is self-sustaining. In other occasions diffusion is confined to anegligible sub-set of the entire population of potential adopters. In the latter case a policy intervention can drive the system to overall propagation. This can be achieved by adoption subsidies which, in order to be effective and to avoid a waste of resources, must follow an optimal dynamic schedule. It is shown that the phasing-out stage is as important as the early stage of the intervention.