Any comparison of the costs and benefits of climate change is dominated by the chosen discount rate. But, although the Stern Review emphasises the ethical nature of the parameters entering into its choice of a relatively low discount rate, its discussion of the ‘pure time preference’ parameter is unbalanced. In particular, no consideration is given to the role of ‘agent-relative ethics’, which (i) has a wellestablished philosophical pedigree going back to David Hume; (ii) is likely to correspond closely to world-wide public attitudes towards intergenerational welfare; and (iii) would entail discounting a unit of welfare accruing to future generations compared to an equal unit accruing to people alive today at a positive rate. The authors also discuss the other ethical parameter upon which the discount rate depends, namely the elasticity of marginal utility with respect to consumption. In the conventional model, this simultaneously reflects different aspects of inequality aversion as well as risk aversion, which complicates its interpretation. Finally, they discuss the divergence between market rates of discount and the low rate chosen in the Review, and the limitations—on the one hand—on the normative significance of market rates, as well as the danger—on the other hand—of relying on rates chosen by elites or philosopher kings.

One of the most controversial areas of economics is the practice of discounting: attaching a lower weight to future costs and benefits than present costs and benefits. Discounting appears to offend notions of sustainable development and the interests of future generations. Recent advances in the theory of discounting hold out strong hope that the ‘tyranny of discounting’ can be avoided through the use of time varying discount rates (TVDRs). This paper reviews the recent rationales for TVDRs and applies the results to issues such as nuclear power and global warming control.