We provide empirical evidence that an early form of “mobile money” is used to share risk. Our analysis uses a unique dataset containing the entire universe of mobile phone communications over a four-year period in Rwanda, and exploits spatio-temporal variation in communication caused by earthquakes and floods. We show that individuals are significantly more likely to send money to people affected by economic shocks, and use these shock-induced transfers to test between two models of phone-based giving. We find that the patterns of interaction are more consistent with the predictions a model of giving based on reciprocity, rather than a model of giving based on pure charity.From a new paper by Blumenstock, Eagle, and Fafchamps (August 2012).
HT: IPA
