Should we beware of the Precautionary Principle?

How should society deal with risks when there is scientific uncertainty about the size of these risks? There has been much recent discussion of the Precautionary Principle, which states that lack of full scientific knowledge should not be used as a reason to postpone cost-effective preventive measures. We show in this paper that the Precautionary Principle contradicts one important intuition about the right way to act in the face of risk, namely the principle of ‘looking before you leap’. When we expect to learn more about the future, the effectiveness of our preventive measures will be greater if we learn before we act. However, a number of other ways of taking uncertainty into account are consistent with a reasonable interpretation of the Precautionary Principle. First, postponing preventive measures may increase our vulnerability to damage, which induces a precautionary motive for risk-prevention, similar to the precautionary savings motive. Secondly, stronger preventive actions often yield more flexibility for the future, so that acting early has an option value. Thirdly, when better information comes from a process of learning-by-doing, the risk associated with early events is amplified by the information they yield about the future. This plays a role analogous to that of an increase in risk aversion, making us more cautious. Fourthly, because imperfect knowledge of the risk makes it difficult to insure, the social cost of risk should include a risk premium. Finally, uncertainty about the economic environment enjoyed by future generations should be taken into account. This raises the benefit of acting early to prevent long-term risks. If the Precautionary Principle sometimes gives good and sometimes gives bad advice, there is no escape from the need to undertake a careful cost-benefit analysis. We show that standard cost-benefit analysis can be refined to take account of scientific uncertainty, in ways that balance the Precautionary Prin... [ABSTRACT FROM AUTHOR]