The relevance of uncertainty in the climate change policy debate is without doubt. Surprisingly, there have been few attempts to examine the direct policy consequences of including uncertainty in an integrated climate-economy framework. This paper presents results concerning optimal policy stringency and instrument choice when economic and climate parameters assume distributions rather than single values. Uncertainty is found to raise the optimal level of emission reductions relative to an optimization based on one set of central parameter estimates. Much of this effect can be related to economic rather than climate uncertainty. Uncertainty also leads to a preference for taxes over quantity controls.
Previous studies of uncertainty in the climate change context have used a small number of states to measure the value of earlier information, learning and adaptation. This paper attempts to refocus attention on the more basic question of whether, in the absence of new information and learning, the inclusion of uncertainty yields significantly different policy conclusions. For policymakers confronting the problem of climate change today, this is the more relevant question.