Certain and Uncertain Utility and Insurance Demand: Results From a Framed Field Experiment in Burkina Faso

In this paper, we argue that discontinuous preference over certain and uncertain outcomes (as in Andreoni and Sprenger, 2009; 2012) have a dampening effect on the demand for insurance. The intuition is that if agents exhibit a disproportionate preference for certain outcomes, they would undervalue uncertain insurance indemnity payments compared to certain premium cost and exhibit lower demand for insurance compared to a classic expected utility maximizer. Inspired by the seminal work of Andreoni and Sprenger, we design games to identify agents with a disproportionate preference for certain outcomes and play them with 571 cotton farmers in Western Burkina-Faso. We then provide experimental evidence that this is a powerful framework to understand demand for micro-insurance. Specifically we show that agents with discontinuous preference respond positively to an alternative presentation of a classic insurance contract: they are willing to pay more for a given contract if the premium cost is artificially made uncertain by being directly deducted from indemnity payments. We also explore alternative behavioral arguments such as loss aversion but argue that they offer less appealing framework to understand the full set of our results. Our results have practical implications for the design of insurance contracts.