Can financial inclusion beat the odds by combining micro-credit and micro-insurance?

Blackjack is one of the few casino games where sufficient knowledge about the order of the cards allows you to systematically beat the odds. The problem is that its hard both to keep track of the cards and to calculate whether to stick, twist or fold. In 1982 some adventurous geeks figured out that several pairs of eyes and some basic wearable computing might allow them to collaborate and win big; after a couple of practice runs they headed to Las Vegas, and, amazingly, filmed their heist under the noses of the casino bosses. They called the subsequent documentary ’The Wedding Party’ because that’s what they posed as to avoid attracting suspicion about one group all being at the same table at once. At the end they get surrounded by a mob of burly security goons desperate to figure out how these guys’ luck was lasting so long.

Wouldn’t it be good if we could similarly guarantee the outcome from tech bets on financial inclusion and poverty alleviation? Well, perhaps we can, using a similar analytical approach.

In 2014-15 I was asked to design the digitalisation of a paper-based agricultural micro-credit scheme in Ethiopia. This Gates-funded initiative has ambitious goals: provide farmers with credit to increase the take-up of fertiliser and other costly agricultural inputs (like irrigation pumps) in the hope of higher crop yields. The theory runs that so long as the bigger harvest can pay for the cost of credit, smallholder farmer incomes will rise. In 2015 this scheme aimed to reach 1.8 million farmers, a volume that’s hard to manage without some form of automation. Using the impressive Addis Ababa-based software supplier Apposit, in the pilot areas for digitalisation we deployed smartphones at agricultural co-operatives, issued NFC stickers to farmers, and used the community and local MFIs to assess creditworthiness and issue microcredit. Despite the scheme launching in the lead up to one of the worst droughts in living memory it still transacted nearly US$10 million [ETB 160M] in the last couple of months alone. Because we designed massive scalability into the solution, we expect very significant growth in the years ahead.

There are obvious pitfalls, not least being what happens if the rains fail again - last year’s drought may have been related to el Nino and we hope it’s a decade until the next big one. When it next happens, wont farmers who have taken out microcredit find themselves applying expensive chemicals to seeds that will never grow? Who will be liable for these bad debts, and what will be the impact on the digital ecosystem? How to mitigate this?

At this point we need to look at the development of micro-insurance. This is the incredibly fast-growing service that helps excluded communities to protect themselves from catastrophic downside risks; there have been some amazing recent successes, such as Telenor India's Suraksha micro-insurance product (by MicroEnsure) which acquired 22 million opted-in customers in just 148 days from launch. Today, micro-insurance focuses on life (and credit-life) products (see table below), though cover for healthcare (e.g. emergency hospitalisation) is growing fast. Micro-insurance is very much a work-in-progress, and the reasons for the current product mix is owing to the complexity of property or agriculture insurance products; life and credit-life use cases are just easier to set up, opt-into, under-write and manage. Agricultural micro-insurance, for reasons that would make a great case study, makes up just 0.5%-1.75% of the market in LatAm and Africa, with just US$1.1m in gross written premiums reported in the whole of Africa in 2014. Successful agricultural micro-insurance use cases are very hard to find.

But what if we layer catastrophic agricultural micro-insurance on to the micro-credit service discussed above? We know the micro-credit service is profitable so long as the revenue from the increased crop yield exceeds the cost of servicing the micro-credit; using round numbers, fertiliser use typically doubles or triples the harvest, so with loan interest between, say 15-50%, it’s clearly very advantageous for farmers to borrow this money and get a higher yield - unless there’s a catastrophic failure of the harvest. At this point, agricultural catastrophic micro-insurance can pick up the economic cost. What’s inventive about this step is the following: we know the expected outcome from the micro-credit, because applying the recommended fertiliser gives a quantifiable expected yield increase for the fertilised crops; we also have the predicted frequency at which there’s a catastrophic failure in the harvest from historical data. Combining these two gives us both the likelihood of the micro-credit upside and the probability of a catastrophe triggering a micro-insurance claim. Since the insured farmers all have the micro-credit, we know there's a good chance their crop yields are going to be significantly up, which then lowers the risk on the corresponding micro-insurance product, if applied to the same community over time. The combined return for this hybrid product can therefore be estimated and so developed as a single product with known costs and benefits. It would be interesting to see if this might catalyse the deployment of agricultural micro-insurance as a more attractive service to providers, and help the micro-insurance industry continue its impressive growth trajectory.

Malcolm Vernon 9th Sept 2016

Malcolm heads Social Mobile Ventures Ltd, a consultancy firm working to use mobile and internet technologies to create positive social outcomes in frontier markets.