Kenyan farmers champion crop insurance for climate resilience
In Kenya, smallholder farmers face the threat of extreme weather conditions due to climate change. These vulnerable farmers also lack access to traditional insurance schemes – where payments are linked to individual farmer yields – that could safeguard their crops and assets.
To overcome this constraint, index-based insurance products have been introduced to measure an independently observable outcome, such as local rainfall, which impacts yields. This approach reduces the high costs of the insurer to monitor thousands of individual households and, in turn, reduces premium costs for farmers. However, in practice, index-based insurance comes with its own challenges, including inaccurate pay-outs when farmers experience crop damage. This unreliability has resulted in a lack of trust among farmers. As a result, insurance uptake has remained low, which has impacted millions of people who could potentially benefit.
To address these issues, ACRE Africa, the Kenya Agricultural and Livestock Research Organization, the International Food Policy Research Institute and the Netherlands’ Wageningen University worked together to develop, test and evaluate an innovative, picture-based insurance (PBI) product. The technology of the PBI project, which is part of the Cultivate Africa’s Future Fund (a partnership of IDRC and the Australian Centre for International Agricultural Research), capitalizes on advancements in remote sensing, image processing, machine learning and smartphone ownership to deliver fast and accurate insurance coverage for crop damage. Initially focusing on maize, the project scaled up to also target green grams, beans and sorghum, and uses satellite and smartphone imagery of farmers’ fields uploaded to an app to rapidly assess crop losses.
More than 8,480 farmers (65% women) are covered for climate insurance.
More than 4,000 farmers have received payment for their claims.
181 champion farmers (58% female) collected more than 60,000 images from farmers’ fields.
Three machine-learning models have been developed to speed up image processing and claim settlement.
Known as SeeItGrow, the PBI smartphone app is used by “champion” farmers – trained by the project team to use the app and work with local farmers – to capture and submit photos from farmers’ fields at different crop growth stages. Working with a community of local farmers on the ground, as opposed to sending an agent across the country to visit thousands of individual farms, reduces monitoring costs and helps improve smallholder trust in the insurance process. In conjunction with satellite imagery from the local area, the field images are used by agronomists and insurance companies to assess actual crop yield losses and determine if conditions warrant a payout.
To join the cost-effective scheme, farmers dial a telephone code – using Unstructured Supplementary Service Data (USSD) technology – and follow prompts that guide them on purchasing crop insurance. For an acre of land, insurance cover worth 200 Kenyan shillings (KSh), or CAD2.00, provides a payout of up to KSh2,000 (CAD21) in the event of weather-related crop loss, which helps farmers recover the initial investment paid to buy seed. In addition, the premium offers farmers information on regional weather conditions and the provision of tailor-made advisories to farmers every two weeks.
To enhance efficiency in the insurance process, the project team has developed three machine-learning models to automate the classification of growth stages, types of damage and extent of damage based on maize images. This is important to reduce the time needed to assess high numbers of photos and settle claims.
In seven counties in Kenya (Bungoma, Busia, Embu, Tharaka-nithi, Meru, Makueni and Machakos), champion farmers or village extension service providers, have been engaged and trained to share information on the climate-smart insurance product and upload images from farmers’ fields. As they register farmers, the champions log farmer details, including their names, location, the size of their land and the crops they grow. This information is used to determine the reach of the scheme and to understand what kind of crops the farmers are growing.
As well as promoting crop insurance, the champion farmers are also passing on the training they have received in good agronomic and sustainable land-management practices, including mulching (a departure from burning farm waste after harvesting, which undermines soil fertility), intercropping, crop rotation and agroforestry. This information helps them, as well as the farmers they train, to diversify income streams and further enhance climate resilience. The champions are also being linked to the formal input sector to increase local access to quality agricultural inputs, such as seed, which they sell for a commission.
Working with champion farmers since 2019, the project has registered 8,500 climate-smart insurance policies, resulting in more than 4,000 insurance payouts (with a relatively larger number of payouts for female clients), and has collected close to 65,000 crop images for seasonal crop monitoring.
Enrolling female farmers
As elsewhere, women farmers in Kenya generally have less access to land, agricultural financing, training and education than their male counterparts. The research found that women are also more disempowered than men due to heavy workloads, lack of autonomy in income and a strong acceptance of domestic violence. Therefore, concerns for gender equality and social inclusion were core to the project’s objectives and considered at all stages of design, implementation and reporting.
To enhance women’s inclusion, more female champion farmers were recruited (58%) than males. Consequently, not only have women reported feeling more at ease when being enrolled, but female champions have also been found to recruit more women (67%) than men (52%). Among the women who were offered a policy, 68% took up the product in 2021.
Elizabeth Musembi, a 43-year-old mother of three from Machakos County, is one farmer who enrolled with PBI after engaging with a champion farmer. She signed up prior to the rainy season of 2021, paying a premium of CAD3.00. Farmers were hit by drought during this season, losing most of their crops. Fortunately for Elizabeth, she was able to claim CAD20 in compensation through the scheme – enough to buy three bags of seed to plant for the next season, which she would not have been able to afford without the insurance.
By covering farmers for visible damage to their fields, the PBI products are reducing the disparity between insurance pay-outs and actual crop losses, reducing risks for smallholders. The project has been able to carry out automation of image analytics for more than 65,000 crop images, increasing the speed and accuracy of claim settlement processing, which represents a crucial step towards achieving scalability and sustainability. The next step will be to assess farmer data to see which other crops should be insured.
Working through champion farmers has increased the reach of insurance products among women and, in fact, increased demand for insurance among more women than men. The champion farmer model has therefore been essential for providing climate information and technologies, and in promoting gender inclusivity.