Agriculture Insurance: Its Hard...

Pula was recently mentioned in the Economist and their thoughtful and balanced piece led me to reflect on working in agriculture insurance for the last 10 years. It is tough making stuff work in Africa. It is tough making stuff work in Agriculture. Now, try adding Insurance to this recipe and we’ve made one of the most challenging concoctions possible. I started working on Agriculture Insurance in Rwanda in 2008, and I remember thinking at the end of one supremely unsuccessful early pilot: This stuff is HARD. 

I was shepherding a group of tomato farmers who wanted credit, bankers who hated risk, and insurers who didn’t understand weather, all in a country where only one weather station remained after a genocide. We insured farmers for drought, but they lost their crop to a disease. It was just never-endingly HARD.

But I also remember thinking: this is what I want to do with my life. I want to build insurance products for farmers.

I still think that today, as we build PULA, an insurance technology company that develops and deploys insurance products for farmers across 8 countries across Africa and Asia. But things have changed in the last 10 years. 

When I started working on insurance, long before we started PULA, we needed weather stations. We could only find one weather station in the whole of Rwanda with sufficiently long time series acceptable to the only reinsurer that was willing to take on this kind of risk at the time, Swiss Re. I was at the Syngenta Foundation and we liked to tackle problems head on, so we started building weather stations. A couple of years in, we were running east Africa’s largest weather data network and we realised that the solar panels on the weather stations were too valuable without a permanent security guard. 

We didn’t give-up; we found a solution in satellite data. Using satellites was cheaper, the data was often free, and they had zero security costs.

But we also lost something when we moved to outer space: Farmers couldn't touch or see the satellites. They had called the weather stations ‘musema kweli’s or ‘truth tellers’. Now they’d ask to meet “Mr. Satellite”, so he could inspect the farm. And they certainly were NOT joking around when they disagreed with the outcome of the insurance compensation, as determined by Mr Satellite. 

Our satellite algorithms would sometimes simply not catch farmers experience and  I remember calculating in my mind how my life savings could cover a group of farmers who had lost their livelihood, but our algorithm didn’t reflect this. The insurance I was so "proudly building" was mercilessly failing. Satellites and the mathematical formulas we built to simulate droughts were often not better than the proverbial monkey in simulating farmers experiences.

I remember the day we came up with our yield index insurance, where we insured farmers against a whole range of risks through less fancy but much more down to earth ‘Harvest measurements’. My favourite and relentless negotiator client (in equal parts) at One Acre Fund, Andrew Youn nudged me in the right direction. He said: “These weather insurances get it wrong most of the time, can't we do better? Do you think any insurer will accept working with field collected yield data?” 

So, when I started Pula with my cofounders, we all but mothballed our weather insurances and started running teams of young enumerators, equipping them with things like tablets, weighing scales, harvest bags and rope. We advanced them airtime and transport money and off they went with their randomly sampled farmer list to harvest with farmers. The surveyors set standard sized boxes on farmers’ fields and while they could not visit every farmer, farmers would often know of a neighbour who had been surveyed. And their harvest measurements reflected losses in yields caused by floods, birds, fall army worm as well as droughts, dramatically expanding our insurance product coverage.

The first time we got reports of floods and bird attacks, I held my breath as our enumerators visited the farms. I could not have been prouder when the yield results came out. Our statistics matched farmers experiences. I still remember that day. It's the day I started sleeping better. 

These last couple of months, PULA jumped a very big hurdle in increasing farmer demand for insurance in Malawi and Zambia. It’s a story for another time, but safe to say that it feels awesome to get over 130,000 farmers insured.

So when I read this recent article about insurance in Africa, I thought: yes this stuff is HARD, but I know that we will get there.