Better Analysis of Uncertainty Needed for Water Resource Projects
PAD-78-67: Published: Jun 2, 1978. Publicly Released: Jun 2, 1978.
- Full Report:
When the Corps of Engineers estimates costs and benefits of proposed water resource projects, it encounters uncertainty. The two basic probability approaches to analyzing uncertainty are the "relative frequency" approach which relies on past events and the "subjective interpretation" approach which depends on whether an event will occur at all.
The Corps is more successful in analyses of costs and benefits which depend on the "relative frequency" approach. For example, if floods have occurred at a certain frequency in the past, they can be expected to occur at similar frequencies in the future. However, predictions of events such as future development in a flood-prone area are subjective. The Corps' treatment of this type of uncertainty could be improved. For example, when an estimate of future events is based on the judgment of a panel of experts, it should reflect the degree of agreement in the panel, not just the "average" opinion.
Recommendation for Executive Action
Comments: Please call 202/512-6100 for additional information.
Recommendation: The Corps should: explicitly recognize intangible benefits and costs and show what effect they have on judgments about a project, indicate the level of confidence surrounding estimates which involve uncertainty, make increased use of sensitivity analysis to show the potential impact of uncertainty on expected costs and benefits, adjust estimates to account for uncertainty, incorporate an "option" value in its analysis to reflect the fact that an irreversible action may preclude some other action whose future value could be greater than expected, and consider devoting more analysis to factors that are the greatest sources of uncertainty.