CGD in the News

Letters to The Editor: Puzzle of Stern's Calculations on Global Warming Damage Explained (Financial Times)

January 04, 2010

Sir, The Stern review on climate economics may well achieve a long-overdue shift in the burden of proof towards those who argue that it is uneconomic to cut the rising path of carbon emissions by anything more than trivial amounts ("Benefits of climate action outweigh costs", November 7 - subscription required). Despite its emphasis on new scientific evidence of greater risks, it does so primarily by adopting essentially the discounting method I proposed in my book The Economics of Global Warming.

In this approach, which the Intergovernmental Panel on Climate Change has called "prescriptive", the rate of "pure time preference", or discounting solely for impatience, is set at zero. Future consumption is discounted only because of rising income levels. Only in this way do distant future generations have a meaningful chance to be taken into account.

In my earlier analysis, I limited the horizon to 300 years, because by then CO2 concentrations would begin to fall again as a consequence of mixing back into the deep ocean, as pointed out by oceanographer Eric Sundquist in 1990. In contrast, the Stern review adopts an infinite horizon.

It obtains finite effects only by applying a pure time preference rate slightly above zero, set at 0.1 per cent against the risk that humanity will self-destruct. Nonetheless, 93 per cent of the present value of all future economic welfare arises after the 2000-2200 period explicitly analysed in the review.

The review estimates that global warming damage reaches 1 to 4 per cent of gross domestic product by 2100, and 5 to 20 per cent only by 2200. The reader is thus surprised that the damage is equivalent to 5 to 20 per cent of GDP "now and forever". The puzzle is explained by the fact that the 2200 rate is frozen and continued for the infinite future thereafter. It is the overwhelming weight of this period, not risk-based attention to the band of uncertainty before 2200, that generates the summary all-time loss of 5 to 20 per cent of GDP.

Although I agree with the review's bottom line that strong abatement measures have a favourable benefit-cost ratio, this ratio may not be as high as it suggests once the partial reversal of global warming after about 2300 is taken into account.