Probability Weighting: An Incomplete Solution for Insurance Demand Puzzles
51 Pages Posted: 10 Feb 2019 Last revised: 4 Mar 2020
Date Written: February 25, 2020
Probability weighting is often used to explain insurance choices that conflict with expected utility (EU) preferences. We derive new theoretical results on the effects of probability weighting in the context of common insurance demand puzzles. We identify decreasing relative overweighting (DRO) as a useful condition on the probability weighting function for comparative statics. In a binary-risk model, probability weighting predicts higher demand than EU alone, explaining commonly-observed overinsurance of modest risks. On the contrary, probability weighting does not explain underinsurance for low-probability high-impact risks or for insurance contracts exposed to nonperformance risk.
Keywords: Insurance Demand, Probability Weighting, Non-expected Utility, Comparative Statics
JEL Classification: D11, D81, G22
Suggested Citation: Suggested Citation