If I were a donor looking to support a breakthrough to protect small holder farmers in sub-Saharan Africa, I would take a close look at Lorenzo Casaburi's and Jack Willis’ new working paper, “Time vs. State in Insurance: Experimental Evidence from Contract Farming in Kenya”. These researchers — who presented their work last month at the Annual Bank Conference on Africa, hosted by the Center for Effective Global Action at the University of California, Berkeley — seem to have cracked the code, at least in part, for a policy goal that has eluded the agricultural development community for years: finding insurance products that works for smallholders in developing countries.

Farming in Africa is risky business. Droughts, pests and myriad other factors frequently destroy harvests and edge the region’s farmers toward financial peril. Insurance could mitigate the severity of these calamities, but financial constraints, mistrust of insurance providers and numerous other factors have left small African farmers among the most under-insured in the world. Research into how to address constraints to the adoption of insurance products in a large-scale and commercially-viable fashion have to date been fairly inconclusive (though tons of important work is ongoing). Without immediate answers, several large international donors have deprioritized formal insurance within their funding strategies for agricultural development.

However, Casaburi and Willis demonstrate that fairly simple insurance adjustments can have a big impact. These researchers specifically collaborated to offer Kenyan smallholders working under a contract-farming arrangement the opportunity to purchase insurance with premium payments deducted from their revenues at harvest time. Farming is a cyclical business, with producers typically cash constrained during the planting season and relatively flush at harvest. As a result, an option to "pay later" seems like a pretty straightforward adjustment; however, the results in Kenya were rather astounding. The delayed payment option had a take-up rate of roughly 70 percent, while the standard model, where farmers pay upfront, was adopted by just five percent of the participants. Willis and Casaburi drilled down further in their study to assess why delayed payments were so effective in increasing demand. They find that lack of cash and a preference for an immediate return (“present bias”) were primary constraints to the standard model.

A lot of important questions remain unanswered in this study. Researchers did not have significant scope to understand how holding insurance might have affected the welfare of those farmers that purchased it. Moreover, results like these warrant additional experimentation in different geographic contexts and sectors. The Center for Effective Global Action's Agricultural Technology Adoption Initiative, generously funded by UK Aid and the Bill and Melinda Gates Foundation and co-implemented with The Abdul Latif Jameel Poverty Action Lab, have supported Willis and Casaburi as they explore where their work might go. More resources and partners, however, are really required to make a tangible difference.

Many other interesting research findings were presented at the Annual Bank Conference on Africa. The great folks from the World Bank’s Office of the Chief Economist for the Africa Region just posted a full run-down and video is available here.