Index-based insurance is an innovative financial product that has been introduced in recent years in countries as diverse as India, Malawi, Mongolia, and Thailand. It allows individual smallholder farmers to hedge against agricultural production risks, such as droughts or floods. The product pays out in events that are triggered by a publicly observable index, such as rainfall recorded on a local rain gauge. As a result, advocates argue that payouts can be calculated and disbursed quickly and automatically without the need for households to formally file a claim. This arrangement in turn reduces transaction costs, which would otherwise tend to drive up the price of the insurance. Fast payouts are also likely to be valued by policyholders in an environment where households are poor and often liquidity-constrained. Finally, the insurance product is free of adverse selection and moral hazard problems that often plague insurance markets because payouts are based only on publicly observed data rather than on private information reported by the person filing claims.
Index-based insurance appears, therefore, to hold significant promise for rural households. Weather shocks to agricultural income generate fluctuations in household consumption that are not perfectly insured; at the extreme, they may lead to famine or death. Indeed, plenty of evidence suggests that households in developing countries are only partially insured and may thus avoid more profitable but riskier investments.
Despite the benefits of index-based insurance, there are also some concerns. First, the product pays depending on the realization of an index, not on the actual crop yield or income of the farmer. Crop yields will relate to the index in complex ways that depend on soil moisture, evaporation, soil type, water runoff, and a variety of other factors. A good product will be one that maximizes the correlation between the index and what the client cares about. A variety of factors may discourage participation, such as household credit constraints, limited understanding of the product among potential consumers, limited trust in the insurance provider, or high transaction costs that raise the price of insurance.
In a 2007 study on southern India by Giné, Townsend, and Vickery, the ratio of expected payouts on rainfall index insurance relative to premiums was estimated at only around 30 percent on average, compared with expected payouts equal to 65–76 percent of premiums for automobile and homeowners’ insurance in the United States. This relatively low payout rate may reflect a number of factors, including a lack of economies of scale given the small initial market for the product and the fact that the market is still in its infancy. More important in India, however, is the fact that high weather insurance payouts are correlated with poor macroeconomic conditions because of the dependence of the Indian economy on agriculture and the monsoon. These properties of insurance contracts are problematic to an insurer from a risk-management perspective. If rainfall insurance were written at a large scale, underwriters could limit their risk exposure by selling part of their rainfall risk to a reinsurer or by holding a significant capital buffer against potential losses. But both of these options are likely to be costly because of transaction costs, informational frictions, and tax concerns.
The pros and cons of these types of products raise an important set of interrelated questions: What types of households buy index insurance? What factors prevent the remaining households from participating? And does the purchase of index insurance result in more efficient risk taking?