Comparative Statics and Centrality Measures in Oligopolies with Interdependent Preferences

Considering the Cournot oligopoly with interdependent preferences proposed in [5], we analyze the effects of a change in the network of social interactions. Reconsidering some of the main centrality measures proposed in the literature, we show how intercentrality, Bonacich and Friedkin-Jensen centrality measures can be related in a network described by a general matrix of interaction. This allows showing under what conditions a firm can benefit, in terms of equilibrium performance, from a change in the weight of interaction of a firm with respect to one of its competitors can benefit. Extending the approach to the study of a uniform change in the behavior of all the firms, we show that it is collectively beneficial only if the structure of social interaction is characterized by a sufficient degree of homogeneity in terms of weight distributions