Insuring the Poor: Experimental evidence from contract farming in Kenya


Around the world, the poor do not typically purchase insurance, despite having the most to gain from risk reduction. In the case of crop insurance, studies consistently find low adoption among small farmers, despite the large risks they face from drought, flood, and pests. One explanation is that insurance products typically require the premium to be paid upfront. Farmers must pay premiums at the time of planting, when they are typically low on cash, and they are paid at harvest, when most farmers receive at least some income.

Using a randomized controlled trial, we evaluate demand for a pay-at-harvest insurance product, where the premium is deducted from the farmer’s harvest revenues. We find:

· Seventy-two percent of farmers in the study took up pay-at-harvest insurance, compared to just 5% of farmers that were offered the standard pay-up-front contract, and the difference was largest among the poorest farmers.

· In contrast, when offering a 30% price discount on the premium of the pay-up-front contract, only 6% of farmers opted for insurance, hardly different from the full price premium.

· Giving farmers a cash gift before offering them insurance did little to close the gap between demand for pay-upfront and pay-at-harvest insurance products, showing the difference cannot be explained by farmers simply not having the cash to pay upfront.

· Delaying payment of the premium by just one month increased the proportion of farmers who purchased insurance by 21 percentage points.

The results show farmers have high demand for insurance, but are unwilling to pay for it upfront, with the poorest farmers having higher demand for pay-at-harvest insurance.

Main article

Standard insurance products require premiums to be paid upfront. Could this explain why the poor – who have the most to gain from risk reduction – typically shun insurance products? The authors test the impact of a new crop insurance product in Kenya, where the premium payment is delayed until harvest. They find that offering this new product to farmers leads to a significant increase in their demand for insurance, especially among the poorest farmers.

The puzzle of low insurance demand among the poor

Throughout the world, the poor purchase less insurance (Rampini and Viswanathan 2016), and insurance markets are especially thin in the developing world. This could reflect supply-side problems: insurance is a complicated and highly regulated product, reliant on effective financial and legal institutions. Yet surprisingly, the binding constraint is often on the demand side. Across many settings, the poor have shown little interest in insurance, even though they are likely to have the most to gain from risk reduction.

This puzzle of low demand for insurance is especially stark in the case of crop insurance. Subsistence farmers rely on their crops for their livelihoods and face many risks. For them, the consequences of a failed harvest can be severe and insuring these risks can give farmers the confidence to buy fertilizer and make other high return investments (Karlan et al. 2014). Yet, despite more than 20 years of concerted efforts to improve crop insurance markets in the developing world, poor farmers still rarely buy insurance.

Researchers have considered many possible explanations as to why this is the case, and insurers have modified their products accordingly:

  • To increase trust in the insurer – insurers have trusted organizations endorse the product (Cole et al. 2013)
  • To improve poor quality insurance – insurers use new data to improve the design of insurance products (Clarke 2016, Elabed et al. 2013)
  • To counteract high costs – insurers subsidize the premiums (Cole et al. 2013, Karlan et al. 2014)
  • To reduce unfamiliarity with the benefits of insurance – insurers teach farmers about insurance and advertise (Cai et al. 2015, Karlan et al. 2014)

These efforts have increased demand, but from a low base rate, and take-up has remained low. When farmers do buy insurance, if it doesn’t pay out in the first year, they often don’t buy it again (Cole et al. 2014). 

The impact of premium payment timing

In Casaburi and Willis (2018), we ask whether the timing of insurance premium payments could explain the low demand for insurance among the poor. The welfare gains from insurance come from transferring income across circumstances, from good conditions to bad (risk reduction). That is how basic economic models of insurance work. Yet in practice, standard insurance products also transfer income across time: the premium must be paid upfront, and any payouts are made in the future, if an adverse event occurs.

Figure 1: Time vs. State in Insurance

The goal of insurance is risk reduction: the transfer of income across circumstances of the world, from good situations ( ) to bad ( ). However, standard insurance products also transfer income across time: the premium is paid up front with certainty (at time ), and any payouts are made in the future, if a bad outcome occurs (at time ).


Requiring upfront payment of the premium means the demand for insurance depends not only on preferences over risk, but also on additional factors including liquidity constraints, intertemporal preferences, and trust in the insurer. The ability to self-insure depends on these same factors, thus charging the premium upfront may reduce demand for insurance precisely when the potential gains are largest.

The timing issue is especially stark for crop insurance. The product reduces risk, smoothing income across harvest outcomes. On average, farmers will receive money during infertile seasons and pay in fruitful seasons. But crop insurance gets the timing wrong, making income less smooth over time. Premiums are due at planting, when farmers are investing in their crops, while any insurance payouts are made at harvest, when farmers usually receive their income.

We develop an intertemporal model of insurance demand, featuring liquidity constraints, to understand the theoretical implications of requiring upfront premium payment. The framework shows that the gains from a pure risk-reduction contract are larger for the poor, yet the poor also face a higher cost of paying the premium upfront. This high relative cost of the premium may deter poor farmers from opting for standard upfront insurance products. It arises, in part, because the transfer across time introduces a trade-off between insurance and the ability to self-insure against other risks. 

Experimental evidence from contract farming in Kenya

We conducted three randomized controlled trials in Kenya to test a pay-at-harvest crop insurance product – removing the upfront premium – in partnership with a sugarcane contract farming company. For the insurer, the challenge with delaying the premium is getting farmers to pay when it is due. We use a credit mechanism to enforce premium payment: the company offers the insurance product and deducts the premium (plus interest) at harvest (which was approximately 12 months later in our setting).

Experiment 1: Demand for pay-at-harvest insurance is much higher, especially among the poor

In our main experiment, we offered insurance to 605 farmers and randomized the timing of the premium payment. Seventy-two percent of farmers opted for the pay-at-harvest insurance, among the highest rates seen for full-priced agricultural insurance (the insurance was priced actuarially fairly, meaning that on average, payouts equal premiums). In contrast, only 5% of farmers opted for the standard pay-upfront insurance, which is low but in line with results in other settings. To benchmark the difference, we offered a 30% price discount on the upfront premium to a third group. Even with this discount, only 6% opted for pay-upfront insurance, hardly different from the full price premium.

Figure 2: Main Experiment: Insurance Take-Up by Treatment Group

The figure shows insurance take-up rates across the three treatment groups in the main experiment. The Pay-upfront group was offered the standard insurance product requiring premium upon sign-up. In the Pay-upfront with 30% discount group, farmers were offered a 30% discount on the premium of the same standard insurance product. In the Pay-at-harvest group, farmers could sign up to the insurance under the agreement that the premium would be deducted from their revenues at harvest time.


The results show farmers demand insurance but are unwilling to pay for it upfront. The poorest and most liquidity-constrained farmers had higher demand for pay-at-harvest insurance, with their demand decreasing more when they had to pay upfront.

Experiment 2: Giving farmers cash does not close the gap

One possible explanation is that farmers simply do not have the cash to pay the upfront premium. To test this, we gave some farmers cash before offering them insurance (similar to Cole et al. 2013). The cash gift was slightly larger than the premium, ensuring farmers had money to purchase the insurance if they desired. The cash gift did little to close the gap between pay-upfront and pay-at-harvest insurance.

Experiment 3: Even a one-month delay in the premium increases take-up, but not as much 

Another possible explanation is that farmers are “present biased” – preferring consumption now, rather than later (Laibson 1997, Duflo et al. 2011). Pay-upfront demand is low because the premium must be paid immediately at sign-up. Even a small delay in premium payment could increase demand substantially, similar to “Save More Tomorrow” – a program designed to nudge individuals into a higher savings rate for retirement by asking them to commit now to saving more in the future (Thaler and Benartzi 2004). We tested this idea by comparing take-up when the premium was due immediately upon signing up, to take-up when payment was delayed by just one month. This small delay increased take-up by 21 percentage points; smaller than the effect of delaying payment until harvest time, but still large and suggestive of present bias. A pay-in-one-month insurance product has fewer of the enforcement concerns inherent in a pay-at-harvest product.

Counterparty risk and trust in insurance

Twelve months after our experiment started, due to financial problems, the contract farming company had to delay harvesting (in our setting, the company manages harvesting), which caused nearly half of the farmers to sell to other buyers. Multiple tests show that pay-at-harvest insurance did not induce farmers to side-sell, but the episode does highlight another possible channel: counterparty risk. Insurance requires considerable trust from the farmer – trust that the insurer will actually pay when supposed to. Delaying the premium payment eases this: if the insurer defaults before harvest, at least the farmer saves the premium. Although we find little evidence for this channel, it may account for part of our main result.

Implications for policy and further questions

Our findings suggest further work is needed on the timing of insurance:

  • Are there lasting effects? From a policy view, we want to see if the results can be replicated in other insurance settings. Liu et al. (2016) found that delaying the premium payment tripled demand for livestock in China, and Belissa et al. (2019) found sizeable effects for crop insurance in Ethiopia.
  • Are there other ways to enforce the payment of premiums which are not charged upfront? Comparisons with credit are encouraging, and perhaps the enforcement innovations used in microcredit could be applied to microinsurance. Belissa et al. (2019) offer a pay-later crop insurance product through Iddirs – informal risk-sharing institutions in Ethiopia – although they find non-trivial default rates.
  • What are the implications of upfront premium payment for other types of insurance, especially those with premiums long before payouts, such as life insurance, annuities, and rare-disaster insurance?
  • Do other timings for the premium payment work for crop insurance? At the previous harvest, for example, when farmers still have liquidity.

Our results show how small changes to sophisticated products, which are mostly employed by the world’s rich, may have an outsized positive impact on the poorest people in the developing world.

Further reading

Belissa, T, Bulte, E, Cecchi, F, Gangopadhyay, S and Lensink, R, (2019), “Liquidity constraints, informal institutions, and the adoption of weather insurance: A randomized controlled Trial in Ethiopia”, Journal of Development Economics, 140, pp.269-278.

Casaburi, L, & Willis, J (2018), “Time versus state in insurance: Experimental evidence from contract farming in Kenya”, American Economic Review, 108(12), 3778-3813.

Clarke, D J (2016), “A Theory of Rational Demand for Index Insurance", American Economic Journal: Microeconomics, 8(1): 283-306. 

Cole, S, Gine, X, Tobacman, J, Topalova, P, Townsend, R and Vickery, J (2013), “Barriers to Household Risk Management: Evidence from India", American Economic Journal: Applied Economics, 5(1): 104-35.

Cole, S, & Xiong, W (2017), “Agricultural insurance and economic development”, Annual Review of Economics, 9, 235-262.

Cole, S, Stein, D and Tobacman, J (2014), “Dynamics of Demand for Index Insurance: Evidence from a Long-Run Field Experiment", American Economic Review, 104(5): 284-90.

Duflo, E, Kremer, M and Robinson, J (2011), “Nudging Farmers to Use Fertilizer: Theory and Experimental Evidence from Kenya", American Economic Review, 101(6): 2350-90.

Ghada, E, Bellemare, M, Carter, M R and Guirkinger, C (2013), "Managing Basis Risk with Multi-scale Index Insurance", Agricultural Economics, 44: 419–431.

Jing, C, Janry, A and Sadoulet, E (2015), “Social Networks and the Decision to Insure”, American Economic Journal: Applied Economics, 7(2): 81-108.

Karlan, D, Osei, R, Osei-Akoto, I and Udry, C (2014), “Agricultural Decisions after Relaxing Credit and Risk Constraints", The Quarterly Journal of Economics, 129(2): 597-652.

Laibson, D (1997), “Golden eggs and hyperbolic discounting", The Quarterly Journal of Economics, 443-477.

Rampini, A A and Viswanathan, S (2016), Household Risk Management, Working Paper 22293, National Bureau of Economic Research.

Thaler, R H and Benartzi, S (2004), “Save More Tomorrow: Using Behavioral Economics to Increase Employee Saving", Journal of Political Economy, 112(S1): S164-S187.

Yanyan, L, Chen, K, Hill, R and Xiao, C (2016), Delayed Premium Payment, Insurance Adoption, and Household Investment in Rural China, IFPRI Discussion Paper (01306).