This paper investigates costly social investments, in the context of risk-sharing. Extending the bargaining micro-foundations of Stole and Zwiebel (1996), we postulate that the benefits of risk-sharing are distributed according to the Myerson value. In particular, more centrally connected individuals receive  a higher share of the surplus. Our main focus is comparing individual versus social incentives to establish relationships, and examining the resulting inefficiencies of equilibrium networks. If agents in the community are homogenous, there is never underinvestment relative to the socially efficient benchmark. In contrast, there can be severe overinvestment. We also find a novel trade-off between efficiency and equality, and show that the most stable efficient network is also the most unequal one. When there are multiple groups in a society, and incomes are more correlated within  groups, there can be overinvestment within groups whereas underinvestment across groups is possible, and more central agents have better incentives to form across groups links. This reinforces the efficiency-quality trade-off. Using data from 75 Indian village networks, we provide empirical evidence consistent with our model’s predictions. We show that the comparative statics of network structure as economic environmental parameters are changed are consistent with our theory, using a difference-in-difference approach looking at the risk-sharing network versus the friendship network. Joint with Attila Ambrus and Arun G. Chandrasekhar.

Paper: Social Investments, Informal Risk Sharing, and Inequality