Optimal Packaging of Insurance and Credit for Smallholder Farmers in Africa

Lead Research Organisation: University of Greenwich
Department Name: Natural Resources Institute, FES

Abstract

Farm households in Africa must cope with bad conditions as to soil quality, weather and infrastructure. The variability of rainfall causes yields to vary strongly from one year to the next. With yields already low (due to poor soil condition) these variations can be life threatening. Meanwhile, inadequate infrastructure makes it difficult to help the households with access to financial services, insurance and inputs that could stabilize their access to resources, and enhance yields.

Solving a single aspect, say bringing inputs to the farm, will not be sufficient as credit is also needed. But credit can only be provided if sufficient likelihood exists that loans will be repaid. Here, insurance can help. If insurance of the loan makes it attractive enough for the lender, a package can be composed of inputs, with credit and insurance, that solves all these problems with one bundle. Yet, the households will remain exposed to some risks as insuring against all is prohibitively expensive. What is the appropriate degree of insurance in such bundles? That is the core question addressed in this research. It aims at supplying inputs to farmers on credit, with insurance, in such a way that a good balance is found between the benefits and risks to the farmers and the profits and risks to the credit provider.

We investigate the possibilities for such a balanced approach in Kenya and Zambia in collaboration with a large insurance provider and a farmers organisation. Together with them we collect information on the costs, benefits and risks involved in using the inputs, the alternatives open to them, and the costs and benefits involved in providing credit to finance the purchase of inputs, with and without an insurance against crop failure.

With all this information, we go and talk to the stakeholders concerned to find out how they would respond if more or less insurance would be provided. Will credit suppliers lower their prices, if repayment of loan is more likely because the crop is insured? Will households decide to take higher yielding (but more risky) crops if part of the downside risk is insured? We establish this for the parties concerned in Kenya and Zambia, but also in other African countries.

Having established how these stakeholders respond to changes in insurance, we can proceed to derive what the best degree of insurance might be. And this is then finally tested in a field experiment.

With this knowledge we can help other suppliers of insurance and credit, and farm organisations to establish similar packages that are adapted to the local conditions for input supply, and financial services.

Planned Impact

Our impact plan is structured around eight key objectives, which will be achieved through cost-effective engagement activities that are targeted to specific beneficiaries in order to maximize potential impacts of our research on practice, policy and academia. The primary beneficiary is the group of smallholder farmers in low-income countries. They are expected to benefit from improved insurance products that yield high agricultural productivity. In the long run, the families and communities of direct users will also gain. Among practitioners and policy-makers, the project will encourage key stakeholders in the agricultural value chain to collaborate together towards increasing the provision of improved bundled financial products, and to introduce policy changes conducive to innovation.

Thus, the first steps in our impact plan are to build awareness and secure commitment of stakeholders. To this end, we will hold an initial workshop with stakeholders at the start of the project in Kenya and in Zambia with the aim of achieving greater ownership of project objectives and co-production of knowledge, leading to higher likelihood of implementation success. For cost-effectiveness purposes, we will undertake these activities during the same visit planned for project launch in the two countries. The target beneficiaries are: farmers' associations; insurance companies; MFIs; financial institutions; NGOs; national and local governments; agriculture input providers; agri-businesses; weather station officers; environmental and climate change organisations; journalists; policy-makers; universities and research institutes; Kilimo Salama; ZNFU. Press news will be released in local media, and a dedicated project website will be publicly accessible and regularly updated. This website will be inbuilt in existing NRI/WUR website facilities.

Great emphasis is placed on building capacity within our local partner institutions through co-production of research design from the proposal developmental stage and then through training in survey design, research methodology and analysis. Most importantly, capacity will be strengthened through intensive personal interactions taking place during fieldwork whereby our Northern researchers will be working alongside our Southern researchers. Co-authorship of journal papers and conference presentations are also planned. In addition, we will encourage strong collaboration and mutual capacity building between partners in Kenya and Zambia in order to create the foundations for a lasting exchange of knowledge in the region.

Positive impacts on practitioners will lead to adoption of improved bundled financial products under better conditions. Through a series of workshops and seminars, financial institutions; MFIs; NGOs; Women's groups; insurance companies; farmers associations; input providers; agri-businesses; and extension service officers will become aware of the benefits of bundled financial products on input investment and farm productivity, and will increase their supply and demand for these improved financial products.

Policy dialogue and wider practitioners and policy-making impacts, meanwhile, are strongly aided by the support gained from the African Development Bank, which will contribute to our project significantly through facilitating engagement with African banks, policy-makers and practitioners in the field - ensuring that impacts are manifested across Africa and beyond the project lifetime. Our links to FaRMAf, an EC-funded project being led by NRI and WUR, which applies a collaborative approach involving a wide range of national and regional stakeholders including the Pan-African Farmers' Organisation and the East African Farmers' Federation, will also help. It is therefore expected that practice and policy changes conducive to adoption of innovation will occur at the local, national and regional level in the low-income countries of Africa.

Publications

10 25 50

 
Description Our research investigates whether bundles of insurance, credit and inputs encourage smallholder farmers' adoption of modern technologies that leads to increased productivity, as well as how the supply of credit responds - in Ethiopia (as substitute for Zambia; approved by ESRC) and Kenya. Main findings are:

1. Improving index-based insurance increases the demand for insurance but not the investment in modern technologies. Only when insurance is bundled with credit and input, investment and productivity grow:
We conducted randomised control trials (RCTs) in Ethiopia and compared four products that we offered to farmers: (1) standard index-based insurance; (2) newly improved index-based insurance (IBI); (3) the new IBI plus credit; (4) the new IBI plus credit and inputs. We found that improving delivery channels and trust in the new IBI -via trusting farmer groups- increased the uptake/demand for insurance but this did not translate into higher investment in agricultural inputs. While, bundles of new IBI with credit and inputs, resulted in higher investment in better seeds and fertiliser, which led to greater productivity. This is the clearest outcome that shows the impact of bundling products as opposed to adopting them in isolation.

2. Bundled insurance with improved seeds leads to increase in investment in additional modern inputs. Also, farmers' willingness to pay is high:
Our RCTs in Kenya offered farmers a free insurance product conditional to buying improved seeds that can lead to greater productivity. Our results show that (1) bundling seeds with free insurance increases demand for seeds but by less than the implicit value of the subsidy; (2) the bundle also enhanced the demand for other modern inputs, such as fertiliser, and raised the area people cleared for farming, which suggests the existence of either production complementarities or risk reduction; and (3) farmers' willingness-to-pay for insurance, if offered next year, is high. We interpret this as farmers being less distrustful of the insurance product and more able to internalise the resulting reduced risk.

3. Supply of and demand for insurance-linked credit product show diverse and somewhat conflicting outcomes in relation to product attributes:
We used discreet choice experiments to examine supply- and demand-side preferences for attributes of insurance-linked credit product in Kenya. Willingness-to-offer (WTO) by suppliers and willingness-to-pay (WTP) by farmers were also explored. WTO-WTP figures are most conflicting around credit term, collateral requirement and loan flexibility. However, both WTO and WTP increase with partial (as opposed to full) loan collateral, which suggests that contract design would need to adopt less stringent collateral so as to encourage demand for bundles that can lead to greater productivity.

4. Rainfall patterns are erratic and, therefore, the design of index-based insurance needs to change in order to reduce basis risk:
Our research findings show that rainfall patterns are so erratic that increase the IBI basis risk, i.e. risk of actual crop failure not being captured by the designed weather index. But, by applying dynamic triggers, which would account for erratic patterns, we found that basis risk reduces and demand for insurance rises.
Exploitation Route The findings have been immediately useful for our partners who have been able to put them in practice in their daily business. This is particularly the case for the insurance companies we partnered with, i.e. Oromia Insurance Company in Ethiopia and APA Insurance Ltd in Kenya. We designed new insurance products (often bundled with credit and inputs) for them to offer smallholder farmers. Given our research findings, insurance companies have expressed an interest in continuing offering these and similar products to their clientele.

We also conducted meetings with policy makers in Africa and shared our research findings. Clear interest arose from development banks at the national and regional levels about using our findings to inform policy regarding facilitation of financial markets for smallholder farmers. Most particularly, the African Development Bank and the World Bank requested a copy of our findings to discuss them in internal meetings. Our journal papers, on the other hand, have been submitted for publication and have the potential to incite the discovery of new areas of academic research, based on our outcomes, especially around the question of the extent of impact of packages - comprising insurance, credit and inputs - on agricultural productivity.
Sectors Agriculture, Food and Drink,Financial Services, and Management Consultancy

URL http://agricreditplus.nri.org
 
Description During our randomised control trials in Kenya, we decided to assess the contributing value of subsidising insurance (by providing free insurance to smallholder farmers) and to test the effect of this on farmers' behaviour/decision on whether to invest in good-quality certified seeds, fertilisers, labour, etc. For this, we designed a brand-new insurance product and persuaded a local insurance company to offer this to smallholder farmers. While in Ethiopia, we designed a brand-new product that combined insurance, credit and agricultural inputs to see their impact on farmers' behaviour. Policy/practice: Changes in policymakers' and practitioners' knowledge, understanding and attitudes about the issue: The changes in policy/practice that we observed are that local insurance companies learnt how to design insurance products that are more suitable to smallholder farmers. Policy-makers learn that subsidising insurance and bundling it with credit and inputs are effective because they lead to investment in better agricultural inputs that lead to greater productivity and lower poverty levels. Also, smallholder farmers learnt that insurance is useful and protects them from crop failures, which was also evident from their willingness to pay for insurance in future. Key insights & impact include: • Insurance companies significantly increased their clientele portfolio and business profits during 2015-2018. • The portion of land devoted to farming the four types of crops under study (maize, sorghum, soya and sunflower) increased. This indicates that farmers respond to the presence of the insurance by increasing their farming efforts. (Note: these farmers never before used insurance; and common belief is that, if insurance is adopted, farmers will reduce farming efforts, i.e. moral hazard). • There was a significant increment in the use of improved/certified crop seeds. • Large impact on the purchase of fertilizer and agricultural chemicals. • Greater demand for credit, which was associated with higher use of inputs. • Income expectations were shown to be higher. • Farmers' uptake of insurance products is higher when insurance is bundled with credit. • Farmers demonstrated high willingness to pay for insurance products, which shows the usefulness placed on insurance protection. 1. Impact on Practitioners: As part of the research, we conducted personal interviews with a large number of insurance companies to ascertain their willingness to collaborate with us by allowing us to conduct an impact study about insurance for poor smallholder farmers. We quickly realised that insurance companies were providing insurance to large-scale farmers but not to low-income smallholder farmers due to their belief that poor farmers were too risky. After an intensive period of negotiations, we reached a collaborative agreement with APA Insurance Limited in Kenya and Oromia Insurance Company SC in Ethiopia. During the research, we raised awareness within the insurance companies of the benefits of offering insurance to small-scale farmers and we designed brand-new insurance products to be offered to these farmers. The insurance companies adopted our designed insurance products and extended them to smallholder farmers. The impact of our research on APA Insurance Limited and Oromia Insurance Company SC is threefold: • Insurance companies have learnt to adopt and implement brand new insurance products, whereby the insurance product was designed using the principles described in Marr et al (2018, 2019), Bulte et al (2018, 2019, 2020), Belissa et al (2018, 2019, 2020), Marr (2022). • They have acquired a new set of clientele, i.e. smallholder farmers; leading to increased portfolio and business. For example, APA Insurance Limited stated that their clientele portfolio of smallholder farmers increased by 100% over 2015-2017. APA evidence letter on request. • Due to the workings of the new insurance products, insurance companies made payouts to smallholder farmers when hit by drought. Accordingly, APA Insurance Limited made payouts of about Kenyan Shillings 2,152,804 in total during the 2017 drought, while smallholder farmers in Ethiopia stated that they received Birr 3,000 on average per farmer from Oromia Insurance Company SC as payout. APA and Oromia evidence letters on request. 2. Impact on Direct Beneficiaries: The ultimate aim of the research is to generate positive impact on smallholder farmers, the direct beneficiaries of insurance/credit. During the course of the research, smallholder farmers have benefited in the following ways: • Increased awareness about the workings of insurance/credit: We conducted several workshops and focus group discussion meetings with about 2,700 smallholder farmers, whereby they acquired thorough knowledge of how insurance/credit products work and what the benefits are in protecting farm land from drought. • Introduction to an insurance company: We acted as the liaison between the insurance companies and the smallholder farmers, which led to farmers' improved understanding of the availability of financial products and the consequent demand for them. • Proven benefit of insurance: As detailed in Bulte, Cecchi, Lensink, Marr, Van Asseldonk (2018, 2019, 2020), the impact of insurance on smallholder farmers in Kenya has been the following: 1. Greater use of improved/certified seeds: Having conducted Randomised Control Trials with about 1,000 smallholder farmers in Kenya, our research demonstrates that adoption of insurance leads to higher purchase of improved crop seeds and hence to increased productivity. It also shows that farmers change their risk behaviour and begin to use certified seeds instead of the poor-quality ones traditionally used. Thus, our research findings demonstrate that farmers purchased 15% more certified seeds of the four crops under study (maize, sorghum, soya and sunflower) as a result of insurance cover protection, given than the risk of crop failure, due to drought, was reduced. 2. Most interestingly, we found important crowding-in effects of insurance protection, i.e. farmers bought higher amount of what we named 'unconditional inputs' i.e. fertiliser, labour and agricultural machinery. Research findings show that the overall impact of insurance on unconditional input costs is almost 1700 Kenyan Shillings (circa US$18), more than twice the average insurance price waived. Similarly, we found positive and significant impact on total land farmed (26% increment) and the likelihood of farmers to obtain credit from banks, i.e. 8% increment. • Proven impact of insurance/credit on demand/uptake of products by smallholder farmers in Ethiopia: As detailed in Marr et al (2018, 2019, 2020), Marr (2022) and Belissa et al (2018, 2019, 2020), the impacts are as follows: 1. Theoretical background: It has been shown that insurance has the potential to overcome moral hazard and adverse selection problems that often plague the development of rural financial markets. However, prior evidence shows that adoption of insurance has met low uptake/demand by smallholder farmers. 2. Our research shows that uptake/demand increases if we combine insurance with credit and agricultural inputs. In other words, when we offer smallholder farmers a bundle of useful products (insurance+credit+inputs), several constraints are overcome and smallholders' demand for this bundled product increases, which in turn is beneficial for farmers' crop protection and productivity. 3. We conducted Randomised Control Trials with 1,661 smallholder farmers in Ethiopia. There were three distinct groups that we analysed: (1) farmers that received only insurance; (2) those that received insurance plus credit; and (3) those that received insurance plus credit plus agricultural inputs. 4. Impacts are: the uptake of the standalone insurance is low amounting to only 8.8% of the total demand; while bundling insurance and credit increases the uptake to 24.5% and when we combine insurance, credit and agricultural inputs, the uptake increases by 32.2%. 5. Similarly, added impacts show that insurance significantly reduces supply-side credit rationing, whereby the coefficient of significance is -0.911. Also, risk-adverse farmers exhibit higher uptake of insurance than other risk-type smallholder farmers. Changes in the number and ability of other researchers to use the data or analytical methods in future (eg via teaching or training): So far, we have trained about 100 African young early-career researchers in Kenya and Ethiopia. The topics that we covered were brand-new themes that these researchers have never before knew about. These included: 1. The techniques of designing and undertaking randomised control trials. 2. How to design and administer questionnaires. 3. How to select and evaluate potential participants in the survey. 4. How to collect and verify data. 5. How to enter, clean and validate data collected via surveys. 6. How to interview stakeholders. 7. How to explain about project details to participants and institutions. 8. How to manage budgets and costs. Main factors contributing to impacts: 1. Our project's contribution: We conducted several one-week training programmes during two years in each of the countries under study, Kenya and Ethiopia. Four senior researchers, members of our research project, were the trainers in these programmes. The topics were delivered in English and in the local languages. 2. Others' contributions: Our partner universities contributed with translators to local languages, venues, insights on how to select local researchers, and the like. We have developed a strong and trusting relationship between the University of Greenwich (United Kingdom), Wageningen University (The Netherlands), and local universities and institutions in Kenya and Ethiopia, which is working well. Also, with policy-making institutions such as the Ethiopian Development Bank and local NGOs in both Kenya and Ethiopia. These institutions have requested our research outcomes to discuss them in internal meetings for possible implementation in policy and practice changes. Policy/practice: Changes in the existence and strength of networks of policymakers and practitioners who can understand and make use of the research results: We are starting to engage again with the African Development Bank in order to make our research impacts more widely in Africa.
First Year Of Impact 2017
Sector Agriculture, Food and Drink,Communities and Social Services/Policy,Financial Services, and Management Consultancy
Impact Types Cultural,Societal,Economic,Policy & public services

 
Title Design of Randomised Control Trials - Impact Assessment 
Description We randomise the assignment of a free crop-insurance to 832 farmers belonging to 40 farmer groups, conditional on uptake of certain quality seeds. The seeds include improved varieties of maize, sorghum, soya and sunflower. After a lottery assigning participants to a treatment (40%) or control group, treatment subjects are awarded a free insurance on the land they farm using certified improved seeds. If they do not buy any improved seeds they do not get the insurance-even if they have won the lottery. Of 832 farmers correctly reached by the intervention, 366 won the insurance lottery and 466 did not. Table 1 shows that the randomization worked as expected; there are no significant differences across the two groups. The main objective of the project is to see to what extent the presence of free insurance increases the appeal of improved seed varieties (conditional crowding-in) as well as other inputs (unconditional crowding-in), and to what extent this leads to different farming decisions and outcomes. It is possible to conduct an intention to treat (ITT) analysis, taking all lottery winners as if they had indeed benefitted from the insurance and vice versa (as the insurance is conditional on purchasing quality seeds, this is not necessarily the case). This will yield conservative estimates of the impact. It is also possible to conduct LATE and TOT estimates. By design, it is expected that the presence of insurance may induce some farmers that otherwise would not have purchased improved seeds do to so. Indeed, 434 farmers purchased at least one packet of quality seeds: 46% of the control group and 59% of treatment. The difference in uptake may mean that "worse" farmers are taking up improved seeds that they would not otherwise have purchased. This may downwardly bias the estimates with respect to average productivity and income when comparing these two groups. To measure this effect, in 28 out of 40 farmer groups we also provided a random subsample of control farmers a surprise lottery (40%) in case they had previously decided to purchase quality seeds independently of the project. The insurance is based on the amount of seeds purchased within three days after participation to the surprise lottery, allowing everybody to increase the amount of packets purchased regardless of its outcome. This allows for differences in total input purchases between surprise lottery winners and losers. In total 228 free crop insurances conditional on quality seeds were awarded. Inputted by Professor Ana Marr 
Type Of Material Model of mechanisms or symptoms - human 
Provided To Others? No  
Impact Table 1. Summary statistics by lottery outcome Variables Lost N Lost Mean Won N Won Mean ? Age 455 46.215 358 45.617 0.598 Female 466 0.908 366 0.904 0.003 Education 466 6.328 366 6.470 -0.142 HH size 466 5.652 366 5.751 -0.099 Income generating members 466 2.167 366 2.164 0.003 Mpesa account 466 0.811 366 0.836 -0.025 Bank account 466 0.253 366 0.290 -0.036 Plan to borrow 466 1.3e+04 366 1.3e+04 -570.729 Land under 4 crops in study 466 3.798 366 3.809 -0.010 Total land (acres) 466 9.485 366 9.232 0.253 Produced maize last year 466 0.989 366 0.973 0.017 Produced sorghum last year 466 0.067 366 0.087 -0.021 Produced sunflower last year 466 0.021 366 0.014 0.008 Produced soya last year 466 0.006 366 0.011 -0.004 Likely drought 466 0.442 366 0.415 0.027 Likely excessive rain 466 0.247 366 0.311 -0.065* Likely pest 466 0.685 366 0.678 0.007 Risk game investment 466 59.227 366 64.809 -5.581 Openness 466 0.003 366 -0.004 0.006 Conscientiousness 466 -0.027 366 0.034 -0.061 Extraversion 466 0.018 366 -0.023 0.040 Agreeableness 466 0.001 366 -0.001 0.003 Neuroticism 466 0.021 366 -0.027 0.048 * p < .05, ** p < .01, *** p < .001. The main objective of the project is to see to what extent the presence of free insurance increases the appeal of improved seed varieties (conditional crowding-in) as well as other inputs (unconditional crowding-in), and to what extent this leads to different farming decisions and outcomes. It is possible to conduct an intention to treat (ITT) analysis, taking all lottery winners as if they had indeed benefitted from the insurance and vice versa (as the insurance is conditional on purchasing quality seeds, this is not necessarily the case). This will yield conservative estimates of the impact. It is also possible to conduct LATE and TOT estimates. By design, it is expected that the presence of insurance may induce some farmers that otherwise would not have purchased improved seeds do to so. Indeed, 434 farmers purchased at least one packet of quality seeds: 46% of the control group and 59% of treatment. The difference in uptake may mean that "worse" farmers are taking up improved seeds that they would not otherwise have purchased. This may downwardly bias the estimates with respect to average productivity and income when comparing these two groups. To measure this effect, in 28 out of 40 farmer groups we also provided a random subsample of control farmers a surprise lottery (40%) in case they had previously decided to purchase quality seeds independently of the project. The insurance is based on the amount of seeds purchased within three days after participation to the surprise lottery, allowing everybody to increase the amount of packets purchased regardless of its outcome. This allows for differences in total input purchases between surprise lottery winners and losers. In total 228 free crop insurances conditional on quality seeds were awarded. Notable impact: We trained approximately 20 local university students. Impacts included: (1) raising research capabilities among local university students; (2) acquisition of new research skills such as new research methodologies including randomised control trials; (3) capacity building of other local university researcher on state-of-the art research methodologies. 
URL http://agricreditplus.nri.org/images/documents/Publications/Designed_RCT-ESL0122351project-MARR.pdf
 
Title Baseline Database - Kenya 
Description Baseline Database - Kenya country case 
Type Of Material Database/Collection of data 
Provided To Others? No  
Impact For the collection of this baseline database, we, the research team, trained about 20 Kenyan students on the techniques of implementation of surveys, pilot studies, data collection, interview techniques, liaison with smallholder farmers. The impact resulting from the development of this database included: (1) raising research capabilities among local (female and male) students in Kenya; (2) acquisition of new research skills of these students; (3) capacity building of other research staff within the local university. 
URL http://agricreditplus.nri.org/images/documents/Publications/Designed_RCT-ESL0122351project-MARR.pdf