Agriculture is the lifeblood of Ghana’s economy. The agrarian sector employs 60 percent of the Ghanaian workforce and accounts for 44 percent of the country’s GDP, according to the Ghanaian government.
Despite the importance of agriculture to the African nation, there are few insurance options to protect the investments of farmers and the banks that support them. However Katie School of Insurance and Financial Services faculty and students are working to create microinsurance products—low-cost insurance designed for individuals with low incomes—that would provide support and peace of mind for Ghanaian farmers, who depend on strong harvests of maize, rice, sorghum, and other crops.
The research project began in 2008 when the Katie School received a grant from the International Labor Organization to study the potential for developing microinsurance in Ghana. This country was chosen because a significant portion of its population could benefit from the research. In addition the country has a good relationship with the United States, a less-corrupt government, financial and critical infrastructure to facilitate the work, and is English speaking.
While Ghana already has some insurance products available, most of which are health-related, the Katie School team decided to focus on insurance for the poor. Lack of insurance can be problematic in a year with poor yields. In some cases farmers who are unable to pay back loans are forced to sell vital capital or move to avoid the aggressive pursuit of banks. Likewise when banks do not receive repayment, they are less likely to lend to other farmers who require high-quality seed and fertilizer to ensure a strong harvest.
Microinsurance could break that cycle, ensuring that loans are repaid and that farmers and banks are not left to the mercy of unpredictable climates. Yet with premiums for microinsurance amounting to the equivalent of only $15 per year, cutting administrative overhead to avoid quickly depleting premiums becomes critical.
“The idea is unlike traditional insurance, which you have in the States, where you actually go out and have people look at the losses,” said Jim Jones, executive director of the Katie School. “The infrastructure, remoteness, and size of the farms make that unworkable in developing countries. You have to come up with something else that simulates what the loss would be in aggregate form.”
For the last four years, Jones and Management and Quantitative Methods Professor Askar Choudhury have been exploring methods to assess yields and losses without creating overhead. This goal has seen the team consider several approaches, one of which is using rainfall data in an attempt to create a correlation with crop yield. This would allow assessors to examine data, not the fields.
“At some point the ultimate objective is to put together an insurance instrument with indemnity and everything and how it is going to be paid,” Choudhury said. “Right now we are trying to identify factors that correlate with yield.”
When there is a disaster, usually the government comes to the aid of farmers. This index insurance would be a way to take the place of the government coming to the aid of farmers anytime there is a disaster.
Creating a model that relies on rainfall data has proved challenging. While temperature, rainfall, and other weather data have the potential to estimate yields, weather stations are sparse. Given that Ghana has several microclimates, distances from weather stations may obscure correlations.
Since rainfall alone cannot determine yields, Jones and Choudhury have also included Normalized Difference Vegetation Index (NDVI) satellite data in their research as a way to estimate yields. NDVI images can place a numeric value on the level of greenness in fields, which can be a major factor in predicting a farm’s crop yield.
“I think the most promising thing we’ve had here are NDVI satellite images,” Jones said. “The level of granularity is a challenge—it is 200 meters by 200 meters, so that’s a pretty big area. But it is very close proximity to what you’re trying to measure.”
The team has examined NDVI data for McLean County to determine how well it correlates with actual yield and has been pleased with the initial results. Using McLean in building a model is advantageous as Choudhury and Jones have access to information on what was planted, when it was planted, and what the final yield was. To fully utilize NDVI data in Ghana, similar information would need to be collected.
Several solutions to overcome the lack of data are being explored, including using drones with infrared cameras to more closely examine fields and working with phone companies to establish a system in which farmers can call in or text information. This information, along with the images from NDVI would paint a more accurate picture of what would need to be covered in case of a disaster.
Recent Illinois State alums Adolph Okine (left) and Frank Danquah, both of Ghana, helped research potential insurance options for farmers in their native country.
The project has also given Frank Danquah, M.S. ’15, and Adolph Okine, M.S. ’15, Illinois State actuarial master’s students from Ghana, the opportunity to contribute to the research.
“When there is a disaster, usually the government comes to the aid of farmers,” Okine said. “This index insurance would be a way to take the place of the government coming to the aid of farmers anytime there is a disaster. These farmers have people lending them money. It puts these lenders at ease since they know they will get their money back if a disaster occurs.”
As Jones, Choudhury, Danquah, and Okine continue to create an index model through which a microinsurance product could be based, they are also considering further ways to eliminate overhead. One such idea is to redefine who the policyholder would be for this insurance. Jones noted in one example that a village could actually be the policyholder. With their knowledge of what was planted and other factors, should an indemnity be paid, officials could then distribute money to the farmers. Likewise a co-op, or even the banks themselves, could become the policyholders, making loans more accessible while carrying less risk to lenders.
Though work continues on defining an index that can be used to construct a working model, the project has already allowed for initial findings to be shared in the Journal of Economics and Economic Education Research. Additionally, the broad scope of the project has created opportunities for collaboration with other departments on campus, such as Agriculture, Geography, and Marketing.
In the end, it is all about creating a model that can help Ghanaian farmers.
“I worked in insurance (in Ghana) for a year before grad school,” Danquah said. “I didn’t see any insurance for farmers. So I see this as a very great step. We are able to build models and make recommendations. Some companies will be able to pick these models up and implement them.”
