Abstract
This paper reviews the evidence on external costs of risky behaviors in the U.S. and provides a framework for estimating them. External costs arise when a person does not bear all the costs of his or her behavior. They provide one of the strongest rationales for government interventions. Although the earlier estimates of external costs no longer have policy relevance, they demonstrated that the existence of external costs was an empirical question. We recommend that the estimates of external costs be updated as insurance structures, environments, and knowledge about these behaviors change. The general aspects of external costs may apply to countries other than the U.S. after taking into account differences in institutional, policy and epidemiological characteristics.