Share

Discussion Paper 01020, International Food Policy Research Institute, Washington, DC. Karlan, Dean, Ed Kutsoati, Margaret McMillan, and Chris Udry

An experiment in Ghana examines how risk mitigation impacts credit and investment decisions, finding that although indemnified loans did not change borrowing behaviour, farmers are more willing to bear the risk of bringing goods to market rather than accept lower prices at farm gate.

A farmer’s decision to borrow money from a bank will invariably involve an assessment of their ability to pay back the loan. Preferably, a bank loan would finance inputs to agricultural production to generate profits which would then serve to repay the loan. Yet, since agriculture profits involve unforeseeable and uncontrollable factors, as in weather and crop prices, even the most creditworthy farmers can be reluctant to take a loan if they are sufficiently risk averse. If that is the case, then there is reason to believe that mitigating those risks would influence farmers to request more credit.

A social experiment in rural Ghana examines how smallholder farmers respond to management of crop price risks. Specifically, loans include an indemnification clause whereby 50% of the loan would be forgiven if prices fall below a certain threshold. Data is obtained for 169 eggplant and maize farmers from baseline and follow-up surveys, and from the partnering rural bank who administered the loans.

The empirical strategy applies a “treatment” and “control” design, where inference can be drawn by comparing the group who receives the loan with the clause to the group who receives a regular loan. Ultimately, loan indemnification is meant to encourage investment, so the experiment seeks to measure outcomes in loan take-up rate and investment behaviour. 

The experiment results are ambiguous as to the effectiveness of an indemnification clause, though implications can be drawn from three main findings. First, the loan take-up rate is not found to change when price-risk, or risk of smaller profits, is mitigated. Take-up rates of 92% and 86% were measured for the treatment and control groups, respectively, which are statistically undistinguishable. Notably, farmers who believed that prices would fall were the least likely to accept the loan that mitigated that risk. This is contrary to the expected outcome of the experiment.

The authors attribute this result to pessimism, toward crop prices and the ‘new’ loan product.

Secondly, by isolating the impact of the indemnified loan on investment in agricultural inputs, it is found that a greater share of investment is directed towards primary crops. Although this result is not particularly noteworthy on its own, it becomes interesting when paired with the finding that farmers are shifting from maize cultivation toward eggplant, to a magnitude of 17.5%. Since eggplant is the riskier crop due to perishability, the results suggest that its cultivation incurs the greatest relative reduction in risk.

Finally, an interesting result is found in the marketing of agricultural production. Treatment farmers are found to be 15-20% more likely to sell their goods to market traders than to “farm gate buyers who come to them and pick up the crop”. There is case-based evidence in the rural markets literature that although farm gate prices are guaranteed, they tend to be lower. The findings suggest that farmers are more willing to bear the risks of bringing goods to the market, in order to reap the reward of higher prices.

The research design involved consultation with farmers where concerns over crop prices were identified as a cause for loan default, and a reason against taking bank loans. Bank records and baseline data also supported this view. Yet results are not entirely consistent with the expectation that risk mitigation would lead to greater use of credit. The authors maintain that the loan indemnity, which is a form of insurance, is a product that makes economic sense, claiming that a longer time frame and more strict lending criteria would provide better data for analysis.