Informal Insurance, Basis Risk and the Demand for Microinsurance

The literature on microinsurance is growing. A series of studies have been published in recent years that look at determinants of (generally low) take-up of microinsurance products, particularly index insurance products. Some work is also being done looking at the impact of offering insurance to farmers.

At the Third European conference on microfinance, held June 10-12 in Norway, Mark Rosenzweig shared some interesting results on the take-up of rainfall microinsurance in India, and its impact on risk-taking. He particularly addresses two important determinants of take-up and impact that have so far received limited attention: basis risk (that is, the risk for a client that the insurance may not pay out when he or she experiences an actual loss due to the possible difference in rainfall at the weather station and on his or her plot), and the complementarity of formal microinsurance with informal insurance mechanisms.

Rozenzweig and Ahmed Mushfiq Mobarak took advantage of detailed data on caste in India to understand informal insurance networks. To look at the causal impact of basis risk on take-up and impact, they randomly determined whether new weather stations would be built within or outside villages.

The results are nuanced, and highlight the complexity of insurance decisions of households (and of financial decisions in general). First, informal insurance networks can be both complement and substitute to formal rainfall insurance, depending on the nature of the shock and the level of basis risk. When informal insurance networks correctly protect against aggregate shocks—in other words, shocks that affect most members—demand for rainfall insurance is lower. Formal rainfall insurance, however, is in stronger demand when the caste network does not protect against idiosyncratic shocks (ones that only affect a few members of the network).

Second, a higher level of basis risk does reduce the demand for microinsurance, but the effect is smaller when informal networks are good at covering farmers’ individual losses.

Finally, in terms of impact, the paper is consistent with previous literature in finding that increasing the amount of insurance, formal or informal, that households have leads to more agricultural risk-taking, which, insofar as it allows higher returns, can be a good thing. In this setting, farmers are more likely to plant a riskier (higher yield but less drought tolerant) variety of rice when their combined informal and formal insurance protection is higher.

All in all, the results highlight the need to carefully understand informal insurance arrangements in order to design and offer formal products that really increase households’ total protection against risk.