Evidence on Informal Insurance in Rural Zimbabwe

The test on the presence of full insurance that is commonly employed does not take into account that households also rely on buffer stocks to shield their consumption from income shocks. In this paper a test is developed that deals with this omission. It is shown that in the presence of partial insurance, the common test on the degree of informal insurance underestimates the degree of protection offered by informal arrangements. This finding has policy implications because if the degree of informal insurance is partial, but high nonetheless, targeting of transfers (in case of drought for instance) is of less importance while transitory movement out of poverty is limited. In the empirical part the newly developed test is employed for data on rural Zimbabwe. The presence of full insurance is rejected. 1 For correspondence: Hans Hoogeveen, Vrije Universiteit, Room 4A29, De Boelelaan 1105, 1081 HV Amsterdam. [email protected]. 2 Comments and suggestions provided by Rob Alessie, Paul Frijters, Jan-Willem Gunning, and Peter Lanjouw were very helpful and are much appreciated. I am indebted to Bill Kinsey, who granted me access to his data set.