Should governments, in discounting the future benefits and costs of public projects, use a discount rate that declines over time? The argument for a declining discount rate is a simple one: if the discount rates that will be applied in the future are persistent, and if the analyst can assign probabilities to these discount rates, this will result in a declining schedule of certainty-equivalent discount rates. A growing empirical literature estimates models of long-term interest rates and uses them to forecast the declining discount rate schedule. I briefly review this literature, focusing on models for the United States. This literature has, however, been criticized for a lack of connection to the theory of project evaluation. In cost-benefit analysis, the net benefits of a project in year t (in consumption units) are to be discounted to the present at the rate at which society would trade consumption in year t for consumption in the present. With simplifying assumptions, this leads to the Ramsey discounting formula. The Ramsey formula results in a declining certainty-equivalent discount rate if the rate of growth in consumption is uncertain and if shocks to consumption are correlated over time. Using the extended Ramsey formula to estimate a numerical schedule of certainty-equivalent discount rates is, however, challenging.