Social networking and individual outcomes beyond the mean field case

We study individually optimized continuous outcomes in a dynamic environment in the presence of social interactions, and where the interaction topology may be either exogenous and time varying, or endogenous. The model accommodates more general social effects than those of the mean-field type. We address endogenous networking by assuming that each individual chooses the weights she attaches to the characteristics and the decisions of other agents. This amounts to choosing the elements of a row of a weighted adjacency matrix. Endogenous social networking thus involves simultaneity between decisions and patterns of directed social interactions.