Pollution Control in a Stochastic Environment

This paper sets up a model using stochastic dynamic programming to analyze pollution control decisions. Environmental decision making is characterized by both uncertainty and irreversibility: Uncertainty might be related to future costs and benefits of adopting a particular control policy. Irreversibility may arise when investments in abatement capital are irreversible, or when environmental damage becomes irreversible at a certain point in time. Under such circumstances, the question of the optimal timing of pollution control arises. This paper shows that: (1) Increases in the uncertainty of future control benefits lead to a lower current degree of pollution control if investment in control facilities is irreversible, (2) ignoring the possibility of new information in the future leads to a higher level of control than it is optimal, and (3) the possibility of a major catastrophe may result in a lower optimal degree of pollution control.