Valuing climate damages

The US National Academy of Sciences (NAS) has just released a new report about the social cost of carbon dioxide (SCCO2), described by one economist as ‘the most important number that you’ve never heard of”.

The report is the result of many months of work by dozens of experts, and comes in at about 380 pages. So it will take some digesting. But even a first look shows that it is mainly recommending incremental improvements to the existing method used in the US, which gives a central value of $37 per tonne of CO2 in 2015, rising to $43 in 2020 (all in $2007 using a 3% discount rate).

The NAS report does not give a new value for the SCCO2. Instead it says that the SCCO2 should be estimated by combining socio-economic, climate, damage and discounting modules. Key uncertainties and sensitivities should be adequately identified and represented.

This needs to be set alongside the only way we have of estimating the global harm caused by CO2 emissions. Integrated assessment models, like my PAGE09 model which gives a mean value of about $150 per tonne of CO2 if valued by an average citizen in Europe, or $250 per tonne of CO2 if valued by an average US citizen.

It is encouraging that most of what the NAS report recommends is already standard practice in the PAGE09 model. It contains separate socio-economic, climate, damage and discounting modules. Key uncertainties and sensitivities are scrupulously identified, represented and carried through the calculations. So it is not surprising that most economists expect any new SCCO2 estimates to be higher than those currently in use in the US, perhaps approaching the mean values that come out of the default PAGE09 model.

Whether any of this will be of more than just academic interest remains to be seen, given that some of Trump’s advisers say the price should be zero. But the NAS report shows that, whatever Trump may think, there is plenty for developers of integrated assessment models, including the PAGE model, to do in the months ahead.

Leave a reply