MICROCAPITAL PAPER WRAP-UP: Insurance, Credit, and Technology Adoption: Field Experimental Evidence From Malawi, by Xavier Gine and Dean Yang

Written by Xavier Gine and Dean Yang. Published by the Journal of Development Economics, Volume 89, Pgs 1-11, 2009. Available at: http://siteresources.worldbank.org/INTFR/Resources/GineYang-InsuranceMalawi.pdf (35 pgs)

In “Insurance, Credit and Technology Adoption: Field Experimental Evidence From Malawi,” Xavier Gine and Dean Yang determine whether farmers who are insured against production risk have a greater demand for loans in order to invest in new hybrid seed technology than farmers who are uninsured against the failure of the hybrid seeds. This study examines maize and groundnut farmers in Malawi , where the major source of crop failure is the level of rainfall. All the farmers were offered loans to purchase high-yield hybrid maize seeds or improved groundnut seeds. Farmers in 16 areas were also required to purchase weather insurance that would forgive the loan should there be insufficient rainfall. The uninsured loan did contain an implicit limited liability constraint that allowed the lender to seize a portion of the yield rather than the collateral presented by the borrower in case of low yield. 33 percent of farmers offered the uninsured loan accepted the loan. The farmers who were required to buy insurance were 39.4 percent less likely to accept a loan to purchase hybrid seeds.

Because the farmers in this study lived in 32 different areas and the authors preferred not to disclose that there were two loan packages being offered, farmers were divided into two groups each consisting of people from 16 areas. Farmers were not offered the choice between two types of loans. Rather, one set of farmers was offered an uninsured loan package and other, an insured package. Farmers were surveyed on assets, knowledge of credit, risk aversion, and other characteristics. Both groups of farmers were nearly identical save for the average of 0.84 fewer years of education among the group offered the insured loan. The authors claim that this did not have much of an effect on results after they controlled for this variable as well as others in their regression equation.

For both loan packages, a 12.5 percent deposit was required upon purchase of a quantity of seeds sufficient for one acre of land. Farmers could opt to purchase either improved groundnut seeds or improved groundnut along with hybrid maize seeds. For the improved groundnut seeds, farmers were to repay USD 33.51, including an interest rate of 27.5 percent, in ten months. In contrast, the insured loan package involved a repayment of USD 36.23 to USD 38.34, also including insurance. The uninsured combined loan package for both groundnut and maize required a repayment of USD 35.52. The hybrid maize seeds had been proven to outperform traditional seeds by 12.7 percent. Since the success of these crops primarily depended upon rainfall, the insurance policy was to pay a fixed proportion of the loan amount based on millimeters of rainfall. The insurance policy did not account for any cause of crop failure other than rainfall. In a survey, 70 percent of farmers reported that rainfall as the primary cause of crop failure. The price of the insurance package varied between the regions in the study due to differences in recorded rainfall. Factoring in the insurance premium, the interest rate for the insured groundnut loan varied from 37.8 percent to 44.4 percent according to the location of the farmer.

A mathematical model was used to measure the value attached by farmers to the insurance that came at an extra cost versus the implicit insurance in the uninsured loan contract. A key point to note is that when the hybrid seed yield is less than the amount to be repaid, the uninsured loan calls for the seizure of a portion of the yield rather than the collateral. When the hybrid seed yield is high, the farmers who paid for an insured loan will have paid more than the farmers who accepted an uninsured loan. Therefore, when the yield is low enough, the consequences of purchasing an uninsured loan or an insured loan are similar. When the yield is high, the insured loan would reduce consumption because the cost of the loan would be higher.

The authors use a regression equation to determine the influence of the insurance package as well as characteristics of the farmer on acceptance of the loan. For both the uninsured loan and the insured loan, an increase in self-reported risk aversion by the farmer led to a decrease in acceptance of the loan. This correlation was less pronounced in the case of the insured loan. In the group offered the insured loan, the authors found that farmers with higher education, wealth, and income were more likely to accept the loan. One reason put forth by the authors is that wealthier farmers may have been more inclined to protect their collateral because they would have held more assets the lender would be able to seize in case of a default. To account for the relative unpopularity of the insured loan, the authors hypothesize that the attachment of an insurance policy raised warning signals about the viability of the seeds. The presence of an insurance policy may also have increased the pressure to produce a successful crop by emphasizing that there would be consequences if the loan were defaulted on. Ultimately, the presence of an insurance package did not persuade farmers to take the risk of employing new seed technology.

Similar Posts: