1 Irreversibility , Pollutant Stock Uncertainty , and the Timing of Implementation of Emission Limits

Using real options, we analyze the timing of implementing emission limits for a decaying pollutant when random environmental effects affect its stock. Two types of irreversibility are present: the first type results from long-term environmental damage, and the second from sunk costs in pollution abatement. We assume that social damage from pollution is a power function of the stock of pollutant. With reference to the deterministic case, we find that a small level of uncertainty may either delay or advance an investment designed to cut pollutant emissions, depending on the cost of reducing emissions relative to expected social costs. When environmental uncertainty increases, environmental damages end up prevailing and pollutant emissions should be curbed immediately. More generally, this paper shows that a deterministic model is not always appropriate for dealing with stochastic environmental problems, even when uncertainty is small.