The Representative Heuristic: Impact on the Demand for Catastrophic Insurance

Previous evidence shows that individuals do not purchase insurance for disasters prior to one occurring, and many do not buy insurance even when policies are subsidized. After a disaster occurs, however, individuals tend to demand more insurance and often over-insure. We develop a model which allows investors to be subject to the representative heuristic, which causes individuals to underweight prior probabilities and overweight posterior probabilities. Such deviations from Bayes Rule in updating probabilities lead to lower demand for insurance when a disaster has not occurred recently, even if a policy is subsidized; indeed people may optimally select zero coverage, even for a sub-sidized policy. The representative heuristic also causes individuals to overreact after a disaster occurs and demand more insurance than a conventionally rational individual would demand, even if the policies are unfairly priced. Fur-thermore, we depict the subsidization needed prior to a disaster to induce individuals to purchase insurance. We also show the level of unfair pricing that can be tolerated post-disaster