Averting catastrophes in a more complex world

It lies in the nature of a catastrophe that its cost and benefits are non-marginal. This makes policy decision making in the presence of catastrophes a complex task. Previous work has shown that when projects are non-marginal they should not be evaluated in isolation and that standard cost benefit analysis may provide biased results. This paper pursues this topic in three directions. First, I introduce the possibility that policy measures may be dependent. This allows the policy decision maker to exploit synergistic relationships and avoid unwanted and unintended consequences. Second, the existence of real world constraints may effect the optimal policy set. I evaluate the optimal policy under different constraints on model parameters such as cost, willingness to pay and risk. Third, I allow the likelihood of a catastrophe to increase with time. This provides a realistic framework for evaluating catastrophes characterized by accumulation or tipping points. I show how these extensions change the willingness to pay and the optimal policy set. This paper is largely theoretical, but provides policy decision makers with guidance on the existence and nature of possible biases through analytical results and examples.