Lab experiment finds no moral hazard in disaster risk insurance

Insurance could be an important way to deal with the growing damages resulting from natural hazards. But how to design disaster risk insurance most effectively?

12/17/2020 | 11:27 AM

Individuals are also part of the puzzle, as they can limit potential flood damage by investing in protective measures (self-insurance). One potential problem with insurance is that it may lead to moral hazard: people may become less careful (e.g. by not investing in self-insurance) when they expect an insurance reimbursement. The problem of moral hazard has been documented in many contexts, but evidence on moral hazard in disaster insurance markets is scarce. 

A new article published in the Journal of Economic Behavior and Organization, by IVM researchers Jantsje Mol, Julia Blasch and Wouter Botzen, addresses the moral hazard problem in disaster risk insurance with experimental economics methods. They developed an economic lab experiment to measure investments in self-insurance in response to different treatments. Moral hazard was found in the scenarios where the probability of loss was high (15%) but not when the probability of loss was low 3%). 

CalculatorThe experiment further examined the impact of different financial incentives on self-insurance under different types of insurance coverage (mandatory/public versus voluntary/private). The traditional financial incentive from the insurance industry is increasing the deductible level, which will decrease coverage. The idea is that under reduced coverage, a policyholder will be more motivated to take care of the risk, seeing that a larger part of the damage has to be paid out-of-pocket. The experiment tested different deductible levels and the results are in line with theoretical predictions: increasing the deductible leads to slightly higher investments in damage-reducing measures. 

Another financial incentive to stimulate investments is to give policyholders a discount on their premium, based on the expected value of damage-reduction. Such a premium discount is already common practice in health insurance, but had not yet been tested in the context of disaster risk insurance. The experimental results show that a premium discount can indeed increase investments in individual damage-reducing measures to the same extent as an increase in probability from 1% to 5%. The premium discount is effective both in the mandatory and in the voluntary insurance markets.

The results have implications for the design of insurance schemes to cope with increasing natural disaster risks. The article is published in a special issue on Behavioral Insurance in the Journal of Economic Behavior and Organization

The following link provides free access to the article for 50 days: paper at publisher.