Today, Simon Dietz and Nick Stern (D&S) of the LSE released a new working paper. They make a few simple changes to Bill Nordhaus’s DICE model. The most important are that they allow for a range of climate sensitivities, and the possibility that climate change could cause catastrophic damages if the Earth’s temperature were to rise by another 4 to 6 degC.
The Social Cost of CO2 (SCCO2) from the model in 2015 goes up to between $32 and $103 per tonne of CO2. For those who don’t know, the SCCO2 is the extra impact caused by emitting one more tonne of CO2 in a specific year. The impact happens in many countries and over many years, as the emissions spread out across the world and take decades to be removed from the atmosphere.
This higher SCCO2 may not sound like a big deal. But it is. These new estimates use Nordhaus’s own discount rate assumptions, which are much higher than those Nick Stern has previously used. Higher discount rates make impacts that appear in the future seem less worrying. The discount rate used in this paper is 4.5% per year.
In the press release accompanying the paper, Nick Stern says “I hope our paper will prompt other economists to strive for much better models which will help policy-makers and the public to recognise the immensity the potential risks of unmanaged climate change. Models that assume that catastrophic damages are not possible fail to take account of the magnitude of the issues and the implications of the science.” It is hard to disagree with that.
There is another well-known economic model of climate change, called PAGE, which D&S have used in the past, and which they mention in this new paper. It already includes a range of climate sensitivities, and the possibility of catastrophic damages if the global temperature rise were to exceed an uncertain threshold. The current default version, PAGE09, also uses a range of discount rates. But it is easy enough to run the model with the single higher discount rate used by D&S in DICE. In fact, that is what I did today. It gives a range for the SCCO2 in 2015 of $13 – 117 per tonne of CO2 (10% – 90% confidence interval), with a mean value of $66 per tonne of CO2. The range is slightly wider than that found by D&S because PAGE09 contains another 30 or so uncertain inputs in addition to the two that they were able to insert into DICE.
The similarity of this result to that of D&S suggests that PAGE09 could be a fruitful starting point for the better economic models that Nick Stern craves. PAGE09 is freely available to researchers. Why not contact me for details?