On the Role of Good Faith in Insurance Contracting

The law of insurance contracts provides that if the policyholder is shown to have knowingly misrepresented material facts about his risks in his application, the insurer can cancel the contract ex post facto and refuse to pay any claims. This good faith principle is widespread, but implemented with unequal strictness, under common law or statute law. In this paper, we analyze the role of good faith in insurance application, when policyholders are imperfectly informed about their risk type. We extend the Rothschild-Stiglitz (1976) model of an insurance market with adverse selection to the situation where individuals only receive a signal of their risk type and where a costly verification of the individuals' risk type and/or signal is possible. We characterize the optimal investigation strategy of the insurer, and the insurance indemnity that should be paid contingent on the result of the investigation, when the insurance market is at a competitive equilibrium. We show that the high-risk types get full, fair insurance without any investigation. The contract intended for the low-risk types involves probabilistic investigation, either of the signal directly, or of the risk type and then of the signal if a high risk type is revealed, depending on the costs of the two types of investigation and the posterior probability of the signal. In either case, the equilibrium is Pareto superior to that in the original Rothschild-Stiglitz model, and exists for a larger range of the population proportions of the two risk types. We also analyze the issue of the onus of the proof when intentional misrepresentation of risk is alleged by the insurer, and find the dependence of the optimal choice of the legislative rule depends on the rival parties' costs of proving good or bad faith.