Heterogeneity vs. externalities in technological competition: A tale of possible technological landscapes

The article presents a stochastic interaction model based on Gibbs random fields to analyze technological competition in a population of heterogeneous adopters with local or global externalities. The relationships between both heterogeneity and externalities and imperfect and asymmetric information are first emphasized. When local externalities and heterogeneity coexist, the technological landscapes of the industry are then shown to depend on the relative influence of these two parameters, with a phase transition: technologies coexist either in approximately equal market shares when heterogeneity is high enough or with one of the technologies only surviving in technological niches when local externalities dominate. Niches do also spontaneously appear: technological options survive in economic space due to the existence of some amount of heterogeneity among agents. On the contrary, when global externalities are added, pure standardization almost always occurs. We finally argue that different public policies should be designed so as to fit with different technological landscapes.