It is good to see that Let them eat carbon is starting an interesting debate about the issues raised in the book. Guy Shrubshole looks at the green taxes result that was covered in the media earlier this week for the website Carbon Brief, and makes two criticisms. He argues that:
- The concept of the social cost of carbon is flawed as a means of appraising policy, on the basis of a short paper from Paul Ekins in 2005. The Government now prefers a marginal abatement cost approach.
- The Stern Review gave a much higher Social Cost of Carbon.
Ekins is the head of the Green Fiscal Commission and I have debated these issues with him before. Some of the arguments he makes are spurious. For example, this argument that a marginal concept such as the social cost of carbon is inapplicable to climate change:
SCC is what economists call a ‘marginal’ concept. Marginal analysis works best when a small change in one variable, for example emissions (e.g. 1t of carbon), produces a relatively small change in impact. It works least well, and is generally thought to be inappropriate, when largescale impacts and whole-system changes are involved, especially when these might be triggered by crossing thresholds related to the variable concerned. Yet, as TD05 (p.64) reports, in respect of climate change “Large scale impacts, such as migration, can be triggered by relatively modest climate changes in vulnerable regions."
He has elided changes in emissions and temperature there. Whether or not small changes in temperature have a significant effect, which is debatable, there is no question of a small change in emissions doing so. One tonne of carbon dioxide makes an absolutely negligible difference to the climate.
Yes the social cost may not be linear, but that is why most studies find it changes over time including the one I used for our release. To the extent that there could be unknown tipping points around the corner, that would cause problem for any attempt to balance the costs and benefits of climate change policy - not just the social cost of carbon. Ekins suggests that we need to make a "precautionary determination" to limit emissions and avoid dangerous climate change, but we can't junk the need to balance the costs and benefits in that way.
The most robust example of the kind of argument Ekins is making has come from Martin Weitzman. He has argued - to put a complicated argument about long tailed risk in simple terms - that even a small chance of a potential catastrophe is a good reason to take radical action to curb emissions. That doesn't just mean risk - which is embedded within many calculations of the social cost of carbon already - but as yet unknown risks. As we can't know that there isn't a small chance of a complete catastrophe, we need to take action. I consider Weitzman's argument in more detail in the book but, to put it simply, the problem with it is that there are lots of unlikely, unquantifiable and potentially catastrophic risks. It would be wrong and utterly impractical to base major economic policies on them.
He also talks about the impact on the poor, and the need for appropriate equity weighting. But given that the overwhelming majority of the harms expected from climate change - particularly in high estimates of the cost of climate change like the Stern Review - affect people living some time in the future and much more prosperous than we are today, it is hard to make that an argument for implementing radical and expensive climate change policy now.
The marginal abatement cost approach appeals to politicians. It basically says that we should accept after some high level analysis, like the Stern Review, the targets are a good idea, and that there is a cost-effective means of achieving them out there. Then just argue about whether or not each individual policy is a relatively expensive or inexpensive way of reducing emissions compared to other policies. The problem is that means entirely neglecting the question of whether actual climate policies are proportionate, whether the harms they are addressing justify the cost they impose on families. I don't think we should write a blank cheque like that.
Yes the Stern Review did contain a higher estimate of the social cost of carbon. But it wasn't intended for use in policy appraisal and researchers at DEFRA therefore adjusted it to the Shadow Price of Carbon that I've used. And anyway, surely we shouldn't just be looking at one study? We should take the IPCC approach and look across the academic literature.
Fortunately, and as I report in the book, Richard Tol has done that. This is what he found:
The best available knowledge – which is not very good – [suggests a] government that uses the same 3 per cent discount rate for climate change as for other decisions should levy a carbon tax of $25 per metric ton of carbon (modal value) to $50/tC (mean value).
That estimate uses the social cost per tonne of carbon, which can be converted to the social cost per tonne of CO2 by a ratio of 100:27.29 (1 tonne of CO2 contains 0.2729 tonnes of carbon). The mean value of $50/t C is therefore equivalent to $13.65 (roughly £7.70) /t CO2. In other words, most of the literature finds that the social cost of carbon is lower than that in the study which I used. And as was reported on Monday even that high estimate of the social cost implied green taxes are excessive by £500 a family.