Measurement of Economic Uncertainty in Public Water Resource Development: An Extension

Abstract

Renewed interest in incorporating measures of uncertainty in public investment decisions is a welcome addition to applied benefit-cost analysis. Taylor and North's (T-N) recent effort in this Journal to incorporate the consideration of uncertainty in benefit-cost analysis follows a technique that we have endorsed for some time (Mercer and Morgan 1975a, b, and 1976; Stacy and Buxbaum). This paper logically follows T-N by illustrating a more general procedure and demonstrates that inclusion of uncertainty in benefit-cost analysis need not be limited to the triangular distribution. The probability distribution used here is the Weibull. The major advantage of the Weibull distribution is its generality. It does not depend on fixed (absolute) estimates of endpoints (range) and thus offers increased capability to test for sensitivity to uncertainty. Use of the Weibull distribution is illustrated here by application to the Spewrell Bluff project and comparison of the results to those of T-N.