The network structure of informal arrangements : evidence from rural Tanzania

In developing countries, whenever formal economic and financial institutions lack strength, households are forced to rely on risk sharing and other informal arrangements based on pre-existing interpersonal relationships. This paper takes a network perspective to investigate how rural households form the links through which they provide and/or get economic support, and whether the connection structure of the community affects the formation of these links. I test the hypothesis that indirect contacts matter, that is, agents take into account not only potential partners’ characteristics, but also their position with respect to all other agents. A network formation framework with fully heterogeneous agents is first presented, following Jackson and Wolinsky (1996), an estimation procedure is then proposed and applied to data on a village in rural Tanzania. Results show that when agents evaluate the net advantage of forming a link they also consider the relative position and the wealth of indirect partners. My paper contributes to both network theory and the literature on risk sharing arrangements in that it proposes an innovative procedure to estimate endogenous network formation models, and provides evidence that network structure has an explanatory value disregarded by all previous studies, which are focused on direct relations only.