Non-Linearities in the cost of reducing emissions

One of the pressing issues in climate change is estimating the cost of reducing emissions. While there is a great deal of uncertainty and disagreement about the cost of abatement in general, there is relatively better knowledge about the cost of small reductions in emissions. As larger reductions in emissions are considered, larger deviations from the status quo must be imagined, thus leading to greater uncertainty. Uncertainty about the cost of small reductions can be interpreted as uncertainty about the level of costs, while uncertainty about large reductions reflect uncertainty about the non-linearity of costs. While there has been some discussion of the impact of the convexity of the abatement cost curve on the costs of meeting the Kyoto protocol [4] and on optimal instrument choice, there has been little discussion on the impact of nonlinearities on optimal emissions trajectories. This paper considers the impact of the convexity of the abatement cost curve on optimal emissions trajectories, considering both deterministic and stochastic climate damages. In a multi-period model with discounting, it is optimal to gradually increase abatement