A missing error term in benefit-cost analysis.

Area of Science:

• Environmental economics
• Policy analysis
• Statistical modeling

Background:

• Benefit-cost models are crucial for environmental policy but often neglect random error, leading to underestimated forecast variance.
• Ex-ante analyses lack historical data for traditional variance estimation, posing a challenge for accurate forecasting.

Purpose of the Study:

• To develop a novel estimator for random error variance in ex-ante benefit-cost models.
• To assess the impact of incorporating variance estimates on policy decision-making, particularly for risk-averse individuals.

Main Methods:

• Developed a new variance estimator using analysis of variance measures and the coefficient of determination (R²).
• Applied the estimator to a benefit-cost model for the Clean Air Act.
• Provided a framework for evaluating the utility of variance estimates versus the default zero error variance.

Main Results:

• The new method provides an estimate for omitted random error variance in ex-ante benefit-cost analyses.
• Application to the Clean Air Act model showed an increase in the probability of large net benefits.
• The probability of a negative net present value increased from 0.2% to 4.5% when incorporating the variance estimate.

Conclusions:

• The developed variance estimator corrects a common bias in benefit-cost models, offering more realistic uncertainty assessments.
• Accurate variance estimation can significantly influence policy choices, especially when decision-makers consider risk and uncertainty.
• The proposed framework aids in determining when incorporating variance estimates improves decision utility compared to assuming zero error.