Asymptotic Properties for Models of Strategic Network Formation

New York University, New York NY

Investigators

Abstract

Network models can be used to describe the structure and the effects of various systems of contracts, transactions, and other formal or informal relationships between economic agents. In these contexts, the consequences of economic policies often depend crucially on the structure of a social or economic network, as well as the forces and incentives shaping that structure. This project develops a tractable empirical model of network formation that has credible micro-foundations. Given the broad range of applications for network models, the techniques and insights envisioned in the proposed project may also be useful for researchers in other social sciences, education, public health and related fields. We consider static models of network formation, where links are undirected and a link can be severed unilaterally by either agent. We formulate a random utility model to describe agents' preferences over potential links, and assume pairwise stability as a solution concept. The main objective of the proposed research is to derive the asymptotic distribution of pairwise stable links, assuming that the number of agents in the network grows large. Our formal results will then be used for identification, estimation, and welfare analysis, where we plan to develop specific applications in the context of the proposed activity. In addition to providing a tractable empirical model that can be used to uncover the incentives behind link formation decisions, the limiting approximation can serve as a building block for modeling and estimation of peer effects and other forms of strategic interaction on networks with endogenous link formation.