Measuring fraud potential in the context of crop insurance: honesty among smallholder coffee farmers in Latin America

Purpose Honesty plays an important role in insurance fraud as a potential result of moral hazard. Knowing the honesty of policyholders can give insights into the risk of fraudulent behavior and costs of an insurance product. The purpose of this study is to measure honesty in a context inspired by an agricultural crop insurance case. Design/methodology/approach We conduct a lab-in-the-field experiment among coffee farmers in Colombia and Peru to elicit farmers' honesty. The experiment is a contextualized version of the die-roll task, in which numbers on the die are replaced by pictures of coffee trees. Findings On average, 3.8% of the participants lie in the experiment. Lying (reporting a higher payoff) is not frequent for pictures with high-yielding coffee trees, although these are associated with the lowest payoff. Potential reasons for the heterogeneity of honesty levels are the contextualization of the experiment, gender and culture. Research limitations/implications Our contextualized experiment is a step towards increased external validity of honesty research. Future research can build on our approach by designing the experiment even closer to an insurance situation or another real-world economic context. Social implications Effective insurance solutions are crucial for farmers to manage the multitude of risks they face. Estimating honesty in a contextualized insurance setting can help insurance companies and policy makers to design better and more effective insurance solutions for farmers. Originality/value To the best of our knowledge, our study is the first using a contextualized honesty task to measure potential dishonesty in a context inspired by an agricultural crop insurance case.