Knowledge about insurance products, adverse selection and moral hazard

Abstract

Many individuals often lack knowledge about prominent features in insurance products. We expect well-informed individuals to 1) increase adverse selection by choosing insurance products more suitable to their risk profile and 2) increase ex-post moral hazard by reacting more to higher coverage, compared to less informed buyers. I first test these by looking at the impact of direct measure of knowledge about own insurance coverage on both adverse selection and moral hazard. Combining a randomised experiment with survey data for a vehicle insurance product, I find that 1) all adverse selection is driven by individuals with correct beliefs about their own coverage, while surprisingly 2) informed buyers react less to coverage than misinformed buyers. I then argue that in a standard model of insurance choice, it is hard to justify the second result, without allowing for selection on moral hazard.