Segmentation and Increasing Returns in the Evolutionary Dynamics of Competing Techniques

The diffusion process of two competing technologies is investigated by means of a multilayer niche model modified to encompass several hierarchically related levels, such as the case of fragmentation of the population of adopters between two competing technologies due, for instance, to product differentiation. The multilevel model describes how in each of two markets two groups of firms are producing different products, giving four products in total that can be produced by means of two techniques. The result is thus a model of competition within competition and diffusion within diffusion that generates a non‐linear dynamic system. The characteristics of the system are investigated by means of simulation. The long‐run behaviour of the model is sensitive to initial conditions, though an unfavourable start‐up can be overcome by means of increasing returns to scale. Indeed, returns to scale play a crucial role in determining the final outcome, although very high returns to scale imply instability. Furthermore, the winning technology need not be the most efficient and stability of a trajectory does not imply stability of the relative technique.