Any policy whose costs fall now and whose benefits arrive in eighty years requires a judgement about how much a future benefit is worth today, and that judgement is made by choosing a discount rate. The choice is not technical. At two per cent a benefit arriving in a century retains about an eighth of its value; at five per cent, less than a hundredth. Analyses of climate policy that reached opposite conclusions have differed on almost nothing else. Economists defending a higher rate point to the observed returns on capital; those defending a lower one argue that the welfare of a future person should not be discounted merely because they are future.