Risk-sharing and endogenous network formation

In economic literature insurance networks are often treated as exogenous institutions. Frequently, the assumption is made that some clearly identifiable group (e.g. ‘the whole village’ or ‘the extended family’) constitutes an insurance network. Still, theory suggests that the formation of insurance links depends on a myriad of factors related to smooth information flows, norms, trust, the ability to punish, discount rates, group size and the potential gains of cooperation (e.g. how correlated income streams are and the amount of other income smoothing strategies available). Recent research has shown that risk is more likely to be shared in small, tightly knit networks, which do not necessarily coincide with a collection of households delineated on the basis of just one factor like kinship or place of residence. However, so far only few empirical attempts have been made at investigating the determinants of network formation. This paper suggests an approach by which one might present stylized facts on endogenous network formation. Applying it to a data set collected in a Haya village in rural Tanzania, we find that kinship, geographical proximity, the number of common friends, clan membership, religious affiliation and wealth strongly determine the formation of risk-sharing networks. Our data suggest that poor households have less dense networks than the rich, making them more vulnerable in the face of idiosyncratic risk.