Rainfall Insurance

By Shawn Cole & Petia Topalova | Shawn Cole is a professor at Harvard Business School interested in corporate finance and banking in developing countries. Petia Topalova is an economist in the research department at the IMF. Her research interests include trade and development, financial services, and development in South Asia.

Over 500 million people in India rely on agriculture as a main source of income. While some have access to irrigation, the quality of the monsoon is still tremendously important. When the rain fails, the households suffer: not only do their own lands yield less, but they may have difficulty finding work near their village, since other farms are also suffering. Traditional methods of coping with risk, such as borrowing from nearby friends or relatives, may be less helpful when everyone is in the same predicament.


In cooperation with ICICI/Lombard, the Self-Employed Women’s Association (SEWA), in Gujarat, has begun to help farmers by offering them insurance against poor monsoons. We are currently in the middle of a study involving 100 villages to assess the potential benefits of rainfall insurance. Without insurance, a failed monsoon may force a household to sell productive assets, forgo medical care, or reduce food consumption. Moreover, anticipating this vulnerability, households may shy away from planting high-yield crops which produce more but are more vulnerable to rain shortfalls.

For decades the government has attempted to protect farmers against weather risk by offering crop insurance. Indeed, many government banks require the purchase of crop insurance as a condition of providing an agricultural loan. Yet, few are eligible to purchase the insurance, and those who do, often receive payments only years after their loss. Private insurers have been reluctant to insure individual farmers’ crops. It is very expensive to pay someone to measure the extent of crop loss of an individual who has planted only an acre of land. Insuring private crop loss is difficult: companies attract households with the lowest land quality, and the fact that the crop itself is insured may reduce farmers’ willingness to spend money or resources to ensure the highest possible yield. Recent innovations in the insurance sector have led to the development of rainfall insurance which may dramatically improve the livelihoods of rural poor (both farmers and landless laborers) by substantially reducing their vulnerability to adverse weather conditions. For several years, BASIX, a microfinance institution based in Hyderabad, has offered weather insurance to its clients, and these policies currently reach thousands of households. Rainfall insurance policies, including those offered through SEWA, are based on measurements from an official rainfall station located close to a farmer’s land.

The insurance policy pays the farmer a specified amount of money if there is a rain shortfall, and the payment increases with the severity of the shortfall. This solves the problems of high transaction costs—the insurance policy pays if the official weather station reports insufficient rain—and also eliminates adverse selection and moral hazard. Yet merely offering a product to farmers and agricultural laborers does not guarantee that they will take it. Financial contracts are complicated, and it can be difficult for a farmer to understand why she should spend money now to get an uncertain payout in the future. Indeed, our survey in three districts of Gujarat found that SEWA members had very low levels of financial literacy: when given a short test that included questions about risk, inflation, and interest rates, members scored on average no higher than the would have simply by guessing. This was despite relatively high math skills—respondents answered, on average, two thirds of the math questions correctly. Our ongoing study began with a baseline survey of 1,500 households in 100 villages. SEWA offered rainfall insurance to individuals in 33 randomly selected villages, with 64 serving as a comparison group. Insurance will be phased into these villages at a later date. Of the 500 individuals surveyed in the villages where insurance was offered, approximately one-third purchased insurance. Preliminary results indicate that rich households and households in which the head of household was literate or educated were much more likely to purchase insurance. Members’ attitudes towards risk also mattered. In the baseline survey, respondents played a simple game, which let them choose between “safe” option (Rs. 50 for sure), and increasingly risky options that offered a higher return.

Two months after the survey, households who had picked safer choices were much more likely to purchase insurance than those who had selected riskier option. To improve our understanding of how households think about risk, SEWA targeted different households with specific marketing messages. These messages were randomly assigned. Households that received a flyer emphasizing the benefits of rainfall insurance (“protection against drought”) were more likely to purchase insurance than those receiving flyers emphasizing the potential negative impact of not purchasing insurance (“without insurance, you may be vulnerable to a drought”). The most important question remains unanswered: does rainfall insurance improve the welfare of the households that buy it? In the coming years, SEWA will expand its marketing efforts, target a greater share of the eligible population, and introduce insurance in 33 of the 67 remaining study villages. By carefully tracking households in these 100 villages, we hope to gain a clear understanding of the benefits, and limitations of rainfall insurance.