Academic journal article Economic Inquiry

A Semi-Parametric Model for Asymmetric Information in Health Plan Markets

Academic journal article Economic Inquiry

A Semi-Parametric Model for Asymmetric Information in Health Plan Markets

Article excerpt

I. INTRODUCTION

There are two potential sources of asymmetric information that induce market failure in the health plan market. The first, adverse selection, is analyzed by Rothschild and Stiglitz (1976). In their study, individuals are exogenously assigned one of two possible illness probabilities. Only the individuals know their true illness probability. Under these conditions, low-risk consumers underinsure in environments where equilibrium exists. Adverse selection also occurs when plans can observe characteristics that affect risk but cannot adjust premiums to reflect these risks. This type of adverse selection is discussed by Cutler and Reeber (1998). Here plan premiums are constrained to be the same across all policyholders, and therefore one cannot risk adjust on observable variables such as age or previous medical history. Similar to the first type of adverse selection, the low- and high-risk policyholders sort themselves into different plans, or plans attempt to "cream-skim" by adjusting reimbursements that are more favorable to the profitable low-risk policyholders. Death spirals can occur where the low-risk consumers leave the more generous plans, causing premiums of these plans to grow, which in turn induces more members to leave these plans. As the premiums grow, more individuals leave the generous plan until it is no longer viable.

The second source of information asymmetry comes from the individual's choice of preventive effort. Much recent publicity has shown how individuals can reduce their probability of illnesses or accidents. Choices of eating, drinking, exercising, smoking, unprotected sexual activity, driving behavior, recreational drug use, and sunscreen use are a few items from the long lists of decisions that affect the probability of illness. Health insurers cannot contract policyholders to exert a certain level of preventive effort because there is no mechanism to observe these efforts. Auto insurers can contract indirectly for effort by accessing public driving records and charging premium surcharges for traffic violations. There are no counterpart records for health plans. Arnott and Stiglitz (1988) and Pauly (1974) discuss the welfare effects when preventive effort can reduce illness probabilities, but the level of these efforts cannot be monitored and therefore cannot be written into a contract. When this condition occurs, full insurance is no longer feasible because no effort would be exerted under full insurance. Because health plans are not compensating individuals for the cost savings that come from preventive effort, the individual's choice of effort is usually less than optimal. Therefore, we call this information asymmetry moral hazard. This moral hazard results in underinsurance for the entire population.

Stewart (1994) looks at an economy where both adverse selection and moral hazard are present. He finds that the welfare losses coming from both sources are subadditive (i.e., the total welfare loss is less than the sum of the losses coming from both). Yet when both moral hazard and adverse selection are present, the economy is still worse off by having both sources of asymmetric information instead of one of them. In most of his simulations, the low-risk types substitute preventive effort for income protection, although this is not a theoretically proven result. His study shows that it is difficult to derive general results on the misallocation of effort and risk sharing when both sources are present because these misallocations depend on the individual's willingness to substitute the income savings from effort for the income protection that comes from insurance.

Setting good health care policy requires knowing about the presence of these informational asymmetries. Unfortunately, to date there is no consensus about the presence of informational asymmetries that come from either adverse selection or preventive effort. Most studies have attempted to determine only if adverse selection is present while ignoring the possible existence of preventive effort. …

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

Oops!

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.