Academic journal article Journal of Risk and Insurance

Insurance in a Market for Credence Goods

Academic journal article Journal of Risk and Insurance

Insurance in a Market for Credence Goods

Article excerpt


We study the impact of variations in the degree of insurance on the amount of fraud in a physician-patient relationship. In a market for credence goods, where prices are regulated by an authority, physicians act as experts. Due to their informational advantage, physicians have an incentive to cheat by pretending to perform inappropriately high treatment levels leading to overcharging patients. Our approach aims on the impact on changes in each, patients' and physicians' incentive structure when the proportional degree of coinsurance varies. It is shown that a higher coinsurance rate may lead to either less fraud in the market and a lower probability of patients searching for second opinions or more fraud and more searches.


This article deals with the provision of expert services in a physician-patient framework. Physicians act as experts when patients do not know the exact medical services they need, and physicians therefore determine how much and which type of service will be demanded. The literature labels such services or goods as credence goods (e.g., Darby and Karni, 1973). This feature can also be observed in markets for legal and repair services where customers ex post often cannot determine if they were served appropriately or not. A particular consequence in this situation with asymmetric information is supplier-induced demand, which is beside moral hazard, one of the main informational problems in the market for health services. It refers to situations where either physicians can and do treat more than what is medically necessary or charge for a more expensive treatment than the one they actually provide, as they have better information available than the patients. Moral hazard describes the case where patients demand more than necessary, due to the fact that they are insured and thus pay only a small fraction of the bill. In both cases it is argued that a higher degree of coinsurance weakens the problems created. If patients have to pay more, they consume less, which reduces their moral hazard. If patients have to pay more they also have a higher incentive to control their physician, which in turn reduces the potential for supplier-induced demand. We investigate this latter effect and show that this conclusion does not hold in general.

We study the impact of insurance arrangements on the degree of fraud in such a market. Our exclusive aim lies on the changes in patients' and physicians' behavior, since we are dealing with fraud within the patient-physician relationship. In particular, we ask whether a higher coinsurance rate on the side of the patients makes fraud in this market more or less likely. In our model, physicians diagnose patients. As patients do not have any information about the degree of illness, physicians can claim that the illness is very severe even if only a small treatment is necessary. The choice variable of the patients is to accept or reject the diagnosis, in which case they go to another physician. As a first intuition, one would suggest that an increase in the coinsurance rate will lead to more patients searching for a second opinion if they face a diagnosis for a large treatment. The reasoning would be that a higher coinsurance rate gives patients a larger financial incentive to search, as they have to pay a larger fraction of the medical bill. More patients searching for second opinions give physicians an incentive to diagnose more honestly. However, this intuition is only partially correct.

In particular, we show that: If the coinsurance rate increases, patients are ceteris paribus more willing to reject a high diagnosis and to search for a second opinion, which is in agreement with the intuition just given. However, in equilibrium, there are two possible consequences of an increase in the coinsurance rate: Either patients search less and physicians diagnose more honestly or patients search more and physicians diagnose less honestly. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.