Network Externalities Based on Social Composition

The market-level adoption of many goods and services is often a consequence of interdependent individual-level decisions. The interdependence arises in cases when the value to a single user depends on the quantity of others also utilizing the good or service. At the macro-level, the theoretical consequences of such network externalities provide a clear example of how individual rationality can lead to suboptimal collective outcomes: an early advantage in the number of users can accumulate into an insurmountable advantage, opening the door for the lock-in of an inferior product. Through the development of a computational, agent-based model of product selection, I illustrate that incorporating the sociological idea that user composition matters in determining the value of a product introduces an endogenous check on the self-reinforcing effects of positive network externalities. I also show that when large initial differences among products exist, introducing a composition-based aspect to a network externalities leads to a disconnect between micro- and macro-level behavior; in the sense that even large differences in the determinants of the attractiveness of products at the micro-level, yield similar macro-level equilibria.