PLEASE DON’T STOP THE RAIN Understanding behavioural barriers to take-up of rainfall insurance in India

Wagstaff, Benedict [Browse]
Senior thesis
125 pages


Shafir, Eldar [Browse]
Woodrow Wilson School of Public and International Affairs [Browse]
Class year
Summary note
Weather-related shocks present the greatest risk to the incomes of nearly 900,000 people in India who rely on agriculture for income and consumption. Due to climate change weather patterns are expected to continue to become increasingly unpredictable and variable, and finding effective ways to manage this intrinsic weather risk is of pivotal importance to ensure the progress of the poverty reduction and development agendas in India. Existing risk diversification and consumption-smoothing measures are incomplete solutions to the problem, and formal rainfall insurance has become the central topic of discussion as a potential solution. However, take-up rates of formal insurance have been worryingly low, and experts are still struggling to identify the precise reasons for low take-up. This research hopes to add to the debate by exploring the role subsidies play in shaping the demand for rainfall insurance in India, as well as the role of framing. The study tests how demand varies in response to exogenous variations in price, and how it responds to different marketing approaches. We hypothesized that take-up would be higher if the discount is larger, if insurance is marketed as a financial product (offering a chance to bet against nature rather than a chance to buy insurance), if historical rainfall data accompany the sales pitch, and if the respondent is told the seller will return at the end of the year (this ought to increase perceived trustworthiness). A large scale Randomized Control Trial that spanned three states of India was used to test our hypotheses and found that: i) Demand was highly price-elastic (price definitely matters) ii) Offering the product as a gamble (potential to gain) led to a 5-percentage point higher take-up rate than presenting it as an insurance product (mitigating losses) iii) Accompanying the insurance product pitch with historical data and committing to returning next year led to the same 5-percentage point higher level of take-up. This suggests that both economic and behavioural factors are important in the decision to purchase rainfall insurance products, but further research is needed to identify the ideal combination to promote take-up. The research also inquired of respondents who did not buy insurance why they did not do so. The modal reason given was that the respondent did not have the cash at the time, suggesting policymakers could usefully explore alternative payment options, such as an instalment payment plans and electronic debits from a prepaid cell phone.

Supplementary Information