Discussion Paper
No. 2007-44 | September 19, 2007
Richard S. J. Tol
The Social Cost of Carbon: Trends, Outliers and Catastrophes


211 estimates of the social cost of carbon are included in a meta-analysis. The results confirm that a lower discount rate implies a higher estimate; and that higher estimates are found in the gray literature. It is also found that there is a downward trend in the economic impact estimates of the climate; that the Stern Review’s estimates of the social cost of carbon is an outlier; and that the right tail of the distribution is fat. There is a fair chance that the annual climate liability exceeds the annual income of many people.

JEL Classification:



Cite As

[Please cite the corresponding journal article] Richard S. J. Tol (2007). The Social Cost of Carbon: Trends, Outliers and Catastrophes. Economics Discussion Papers, No 2007-44, Kiel Institute for the World Economy. http://www.economics-ejournal.org/economics/discussionpapers/2007-44

Comments and Questions

Radmilo Pesic - Richard S.J. Tol: SCC paper
September 20, 2007 - 13:49 | Author's Homepage
An excellent paper. Presents a good example of properly conducted meta analysis. Right methodology choice with a broad and representative data base. Firm and correct conclusions. By pointing on the climate policy’s equity implications the author opens a wide field for future studies. A challenging text! RATE 5

Richard S.J. Tol - Pesic's comments
November 26, 2007 - 12:44
Thank you

Gary W. Yohe - Invited Reader Comment
November 09, 2007 - 11:58
see attached file

Richard S.J. Tol - Yohe's comments
November 26, 2007 - 12:43
Thanks to Gary Yohe for great comments, as usual. On the IPCC: I may have overstated my case. While the published estimates seem to have come down slightly between 2001 and 2007 (in flat contradiction to the IPCC 4AR), it does not follow that they will continue to decline. Then again, the Schneider/Smith chapter of the IPCC seems to speculate on the results of future research. On the carbon tax: My calculations make two assumptions. First, the carbon tax does not affect emissions. Second, the carbon tax is not recycled. That is why I refer to this as a liability, rather than a tax. Relaxing these assumptions would require a different analytical set-up, that is beyond the paper. On the tails: Agreed. The analysis is readily redone for specific rates of pure time preference.

Anonymous - Referee Report
November 20, 2007 - 08:12
see attached file

Anonymous - Anonymous comments
November 26, 2007 - 12:55
On splitting the sample by quality: This is easily done, although one should be careful about splitting a sample of 200 in too many ways. On the per capita payment: The calculations use the national average per capita income and the national average per capita emissions. While income per decile can be found for a large number of countries, this is not true for emissions per decile. Even if the data were available, I am not sure that the referee is right, as poor people tend to use energy very inefficiently.

Cameron Hepburn - Referee Report
November 26, 2007 - 09:51
see attached file

Richard S.J. Tol - Hepburn's comments
November 26, 2007 - 13:17
Great comments, as usual, by Cameron Hepburn. What is new: Note that many of the new estimates are by others than me and my co-authors. It would be interesting to do a sample split FUND/non-FUND. Note that FUND generated the highest and the lowest estimate in the sample. />Omissions: The literature on the social cost of carbon indeed adopts a very specific view on ethics and decision making. The next version of the paper will make this explicit. Inflation rates are mercifully low, and small compared to the uncertainty about climate change. Independence: I would be happy to make the spreadsheets available, so that others can replicate my work, and change assumptions at will. I did report a sensitivity analysis on the "quality weights" in the 2005 papers, and the conclusions are robust. Trends: Good point. The best guesses show a downward trend. The mean in current paper is higher than the mean in the 2005 paper because the former accounts for the fat tail, and the latter does not. This is muddled in the paper. Stern: Most of the estimates of the social costs of carbon are along a business as usual path, and Stern is an outlier in that sample. Note that Stern is still on the high side in the sample with similar discount rates. Stern's claim, repeated by Hepburn, that earlier studies exclude risk and non-market impacts, is simply not true. Climate policy: Indeed. This is a dumb error. Climate policy outside the EU is too weak. EU climate policy does not pass a cost-benefit test (unless one cooks the books).

Anonymous - Policy to reduce the probability of bankruptcy?
January 25, 2008 - 11:46
This is a fascinating paper that raises a number of important issues. More detail on the calculations would be welcome. In particular, the assumptions underpinning the analysis of Table 2 need to be spelt out in greater detail. I presume that some assumptions are being made about the incidence of carbon taxes - yet these assumptions are not made explicit. If a proportion of the population of a country were to face 'bankruptcy' as a result of the imposition of such taxes at high levels, presumably something would be done in the tax/benefit system to mitigate this. It strikes me that this latter action should be modelled - or at least it should be made clear that it is not modelled. Otherwise the 'bankruptcy' predictions are upwardly biased. In sum, this is a case where ceteris paribus isn't a terribly useful assumption.