Session 2 looked at two impact evaluations of new microinsurance schemes, how such schemes should be marketed, and what role government should play in encouraging the sector.
First off James Vickery spoke about an evaluation of a microinsurance project in India working with ICICI bank to offer monsoon insurance. Rain is a great risk to insure because the risk is exogenous. The payoff can’t be influenced by the farmer so there are no moral hazard issues.
What might prevent take-up of insurance products? The study investigates price (too expensive) and other barriers, and finds a mix of the two. Insurance demand is price sensitive, but price isn’t everything. Even when the expected value of payouts is larger than premiums, there is still limited take-up. The reasons are likely to be liquidity constraints to some extent, and also trust and understanding of the product
The bottom-line: Preliminary evidence suggest that insurance does affect household risk-taking incentives. Adoption rates are modest but growing, and design improvements are important. But does underinsurance cause underinvestment? Probably still an open question.
Chris Udry then spoke about a similar project in Northern Ghana, offering rainfall insurance to farmers. So do farmers underinvest? And if so, why? We have 2 hypotheses, that farmers are risk-averse, and that they are capital-constrained. The study addresses these 2 constraints by offering insurance (for risk), and cash (for capital).
The study looks at 4 groups.
- Group 1: Offered the simple rainfall insurance product.
- Group 2: Offered the cash grant.
- Group 3: Offered both the rainfall insurance and the cash grant.
- Group 4: Offered nothing (the control group).
The program was designed to be very simple and understandable – if it rained for either too many or too few consecutive days, there is a payout. An immediate response was seen after 1 year, with farmers using more inputs and expanding production, leaving them better off - eating more, and sending their children to school more.
In the second year the insurance was sold, at different (randomized) prices to different farmers. Again a demand curve was seen, with almost 100% of farmers taking up the insurance for free, about half of farmers demanding the insurance at actuarially fair prices, and fewer but still a substantial number at a higher commercial price.
Michael McCord of the MicroInsurance Centre spoke about the importance of product design and marketing. MicroInsurance is a harder sell than microcredit, but the potential market could be 3 billion people.
Finally Barbara Magnoni spoke about what the role of the government and subsidies for microinsurance should be. There are certainly risks. Are subsidies sustainable? Will companies work less hard if they get subsidies?
Session 3: Targeting the Ultra-Poor
Microfinance institutions have had great success at reaching poor and low-income households, but have been much less successful at reaching the very poor. BRAC recently piloted a new project in Bangladesh with the aim of bringing financial access to the ultra-poor (those earning less than $1 a day). This model is now being scaled-up with pilots being carried out around the world coordinated by CGAP, and with many being evaluated through randomized evaluations.
Aude de Montesquiou (CGAP) explained the idea behind the intervention, combining an intensive period of training, along with access to savings, along with asset transfers and consumption support, with the aim of "graduating" people from extreme poverty.
William M. Abrams (Trickle Up) discussed the experiences of Trickle-up in operationalizing the model. Trickle-up have been doing similar projects since 1979, offering households training, grants and savings. Changes that have been made to their traditional program include a focus on health training and “hand-holding” - a personalized mentoring at the household rather than in a classroom. Watch the trickle-up program video here.
Jeremy Shapiro (Yale) discussed some early (18 month) results from one of the impact evaluations ongoing in India. People participating in the program skipped fewer meals, had improved health knowledge, increased life satisfaction, and increases in formal savings. What seemed to be driving much of these results was non-agricultural enterprises - raising the issue of those heterogeneous effects again.
Jonathan Bauchet (NYU Wagner) discussed results from the SKS study in India. The study found no impact on income, schooling, or household savings. It did however find some potentially positive results whereby households shift their sources of income from wage labor to livestock - which is more predictable and under the control of the household. There was also a decrease in household borrowing from money-lenders, which would seem a positive development, but then raises questions about what the ultimate goal of the project should be - graduating into or out of debt?
More to come.