#636 2009 · Kilimo Salama / ACRE Africa (Syngenta Foundation for Sustainable Agriculture) · Agricultural insurance / microinsurance
Kilimo Salama insured farmers nobody could afford to visit by never sending anyone to visit them
the problem
Insuring a smallholder farmer's one-acre crop against drought is uneconomical because verifying each individual farmer's loss costs more than the entire premium they could ever afford to pay
background
Crop insurance for a farmer with one or two acres had been considered structurally impossible for decades, not because insurers doubted the need — drought traps African smallholders in poverty precisely because it can wipe out an entire season's investment in seed and fertilizer at once — but because traditional insurance requires a claims assessor to travel out and verify each individual farmer's loss before paying, and that single farm visit costs more than the whole premium a subsistence farmer could pay for a policy that size. Insurers had simply written off the market at any price the target customer could afford.
The standard response to a market an industry can't reach profitably is subsidy: have a government or NGO absorb the assessor's cost so the insurer can still serve the customer. That fixes the accounting but not the underlying problem, and it depends on continuous outside funding rather than a business the insurer could run on its own indefinitely.
what everyone would do
Subsidize the cost of sending assessors to verify individual farmers' losses, so insurers can still be paid enough to cover the visit. This treats the assessor's visit as a fixed cost of doing insurance and only argues about who pays it, which keeps the program dependent on continuous outside funding rather than fixing the fact that verifying a one-acre farmer's loss will always cost more than they can pay for coverage.
what they saw
The expensive part of insuring a smallholder farmer was never the payout — it was proving, farm by farm, that a loss had genuinely occurred. Once you accept that an area-wide weather event (a drought, an excess-rain season) affects every farmer registered to the same weather station roughly the same way, you no longer need to verify each individual's loss at all; you only need to verify the weather, once, automatically, for everyone in that zone at the same time.
the move
Kilimo Salama removed the claims assessor from the transaction entirely rather than trying to subsidize their cost. Automated, solar-powered weather stations installed across farming regions track rainfall against a predetermined index; at the end of the growing season, the weather-station data feeds an agronomic model, and a payout is calculated and sent automatically whenever the rainfall index crosses a drought or excess-rain threshold — with no individual farm visit, no loss claim, and no paperwork from the farmer required. Farmers buy the policy through the same local agro-dealers who already sell them seed and fertilizer, and premiums and payouts move through the M-PESA mobile-money system farmers already use for other transactions.
why it works
Replacing individual claim verification with an automated index shared across every farmer near one weather station converts a cost that scaled with the number of policyholders (one assessor visit per claim) into a cost that doesn't scale with policyholder count at all (one weather station serving thousands of farmers). That is what allows a premium small enough for a one-acre farmer to still be profitable for the insurer — the fixed cost of the weather station and its data feed is amortized across an entire region's worth of policies instead of being duplicated per farm. Routing premiums and payouts through M-PESA, a payment rail farmers already trusted and used, removes the remaining friction (cash handling, physical branch visits) that would otherwise still require a costly human presence in the loop.
the payoff
Beginning as a 2009 pilot with 200 farmers, the program had grown to 51,000 insured farmers in Kenya and 14,000 in Rwanda by 2012, with premium revenue collected by partner UAP Insurance nearly doubling within six months (from KSh 19 million in 2011 to KSh 33 million in the first half of 2012). In 2011 alone, automated weather-station triggers released KSh 8.7 million (roughly $103,000) in payouts to 2,168 farmers across five Kenyan regions with no claims process at all. By 2015, the program (rebranded ACRE Africa) covered over 394,000 farmers across three countries, insuring $11.7 million, with Syngenta's own data indicating insured farmers earned 16% more than uninsured peers.
where it breaks
The model only works where a farmer's actual loss correlates tightly with the regional weather index — a farmer whose crop fails from a localized pest outbreak, a bad seed batch, or a flooded low-lying plot gets no payout even in a year the regional index shows normal rainfall, and conversely a farmer on unusually favorable microterrain may collect a payout despite suffering no real loss. This basis risk is the structural price of removing individual verification, and it grows the more localized and variable the actual damage mechanism is relative to the weather station's coverage radius. It also depends on weather-station density and reliable data transmission (solar power, mobile network coverage) being present in the target region — the mechanism cannot function anywhere the underlying sensor infrastructure hasn't been built out.
what came after
UAP Insurance's executive director credited the design specifically with avoiding 'claims procedures that have created mistrust' in traditional agricultural insurance, arguing the automated-index model is what made the product 'economically viable for insurance companies in developing countries that had previously written off the agricultural sector' — the removed verification cost, not a smarter subsidy, is what the program's own backers point to as the structural unlock. The index-insurance-via-mobile-money model has since been cited as a template distinct from statistical basis-risk innovations in the same space (multi-scale cotton insurance in Mali solves a different problem — verifying genuine loss under an index — while Kilimo Salama's contribution is eliminating the verification step altogether).
references
- [1]First micro-insurance plan uses mobile phones and weather stations to shield Kenya's farmersScienceDaily (Burness Communications release), 2010sciencedaily.com
- [2]Kilimo Salama (Safe Farming) Weather Index Insurance in Kenya: Early Market SuccessIndex Insurance Forum (World Bank / GIIF, hosted via FARM-D), 2012farm-d.org