Working Paper: NBER ID: w25413
Authors: Qingran Li; William A. Pizer
Abstract: The choice of discount rate has a significant impact on net benefit estimates when costs today have benefits over long time horizons. Standard U.S. government practice for cost–benefit analysis is to bound such analysis using two alternative rates. These rates are meant to represent the rate of return paid by capital investment and the rate received by consumers. Previous work has shown this approach legitimately bounds the analysis—but only when future benefits accrue directly to consumers either in a two-period model or as a perpetuity. We generalize to consider arbitrary patterns of future benefits, accruing either directly to consumers or indirectly through future investment. We derive an expression for the appropriate discount rate and show that it converges to the consumption rate for benefits increasingly far into the future. More generally, the bounding rates depend on the temporal pattern of the undiscounted dollars. As an application, we estimate the appropriate discount rate for climate change damages from carbon dioxide, finding it lies in a narrow range (+/- 0.5 percent) around the consumer rate of interest.
Keywords: discount rate; public policy; cost-benefit analysis; climate change
JEL Codes: D61; H43; Q54
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| discount rate (E43) | net benefit estimates (H43) |
| time horizon of benefits (D15) | discount rate (E43) |
| ratio of investment and consumer rates (E20) | shadow price of capital (D33) |
| consumer rate of interest (E43) | social discount rate for climate change damages (H43) |