Utility, Risk, and Demand for Incomplete Insurance: Lab Experiments with Guatemalan Cooperatives

We play a series of incentivized laboratory games with risk-exposed cooperative- based coffee farmers in Guatemala to understand the demand for index-based rainfall insurance. We show that insurance demand goes up as increasingly severe risk makes insurance payouts more partial (payouts are smaller than losses), but demand is ad- versely effected by more complex risk structures in which payouts are probabilistic (it is possible that a shock occurs with no payout). We use numerical techniques to esti- mate a flexible utility function for each player and consequently can put exact dollar values on the magnitude of the behavioral response triggered by probabilistic insur- ance. Exploiting the group structure of the cooperative, we investigate the possibility of using group loss adjustment to smooth idiosyncratic risk. Our results suggest that consumers value probabilistic insurance using a prospect-style utility function that is concave both in probabilities and in income, and that group insurance mechanisms are unlikely to solve the issues of low demand that have bedeviled index insurance markets.