When payoffs from different actions are unknown, agents use their own past experience as well as the experience of their neighbors to guide their decision making. In this paper, the authors develop a general framework to study the relationship between the structure of these neighborhoods and the process of social learning. They show that, in a connected society, local learning ensures that all agents obtain the same payoffs in the long run. Thus, if actions have different payoffs, then all agents choose the same action, and social conformism obtains. The authors develop conditions on the distribution of prior beliefs, the structure of neighborhoods and the informativeness of actions under which this action is optimal. In particular, they identify a property of neighborhood structures--local independence--which greatly facilitates social learning. Simulations of the model generate spatial and temporal patterns of adoption that are consistent with empirical work.

We present an approach to network formation based on the notion that social networks are formed by individual decisions that trade off the costs of forming and maintaining links against the potential rewards from doing so. We suppose that a link with another agent allows access, in part and in due course, to the benefits available to the latter via his own links. Thus individual links generate externalities whose value depends on the level of decay/delay associated with indirect links. A distinctive aspect of our approach is that the costs of link formation are incurred only by the person who initiates the link. This allows us to formulate the network formation process as a noncooperative game. We first provide a characterization of the architecture of equilibrium networks. We then study the dynamics of network formation. We find that individual efforts to access benefits offered by others lead, rapidly, to the emergence of an equilibrium social network, under a variety of circumstances. The limiting networks have simple architectures, e.g., the wheel, the star, or generalizations of these networks. In many cases, such networks are also socially efficient.