Abstract
The precautionary principle appears to be directly applicable to the prospect that human activities will cause substantial changes to global climate. The magnitude and characteristics of the change in climate and its effects on human activities and unmanaged ecosystems are highly uncertain, potentially catastrophic, and nearly irreversible on human time scales. Nevertheless, the precautionary principle adds little to benefit cost analysis for evaluating climate policies. Benefit cost analysis can incorporate aversion to uncertainty about future outcomes, and also provides a method for balancing this concern against concern for uncertain near term policy costs. The policies favored by setting constraints on future outcomes, such as atmospheric stabilization and tolerable windows, may be less precautionary than the policies favored by benefit cost analysis, as benefit cost analysis can more readily accommodate concerns for moderate as well as severe harms. Concerns about possible climate catastrophes can also be incorporated in benefit cost analysis and, in any event, may have limited implications for policy choice.