Uncertainty about risks to health virtually requires that people have health insurance. But how is the cost of premiums determined? Should rates vary according to some indicators of risk? How much do premiums vary with risk? Do the young and the healthy actually subsidize the old and the unhealthy?

Choosing economically optimal health insurance coverage involves a trade-off between risk reduction and the overuse of health care. The economic purpose of insurance is to reduce financial uncertainty or risk - the more health insurance lowers the risk, the greater will be the increase in social well-being. But increases in health insurance also increase the amount of medical care demand, because insurance lowers the out-of-pocket cost of health care - the larger the demand response of medical care to cost sharing, the greater the decrease in social well-being, due to the purchase of too much health care. This study examines this trade-off empirically by estimating both the demand for health insurance and the demand for health services. It relies on data from a randomized controlled trial of the cost sharing's effects on the use of health services and on the health status for a general, nonelderly (under age 65) population.

Choosing optimal health insurance coverage involves a trade-off between the gain from risk reduction and the deadweight loss from moral hazard. This paper examines this trade-off empirically by estimating both the demand for health insurance and the demand for health services. It relies on data from a randomized controlled trial of cost-sharing's effects on the use of health services and on the health status for a general, non-elderly population.