Suppose that someone proposed using nuclear explosions to melt the Arctic ice cap*, with the aim of opening the Northwest passage and reducing shipping costs, and that this proposal was supported by an analysis showing that world GDP could be permanently increased by 1 per cent, or maybe 3 per cent, as a result.
On the face of it, this seems (to me, anyway) like a crazy idea. Should such a proposal be dismissed out of hand or taken seriously and subjected to benefit-cost analysis or ? And, if we did do a benefit-cost analysis, what would be the result?
In all probability, a benefit-cost analysis run on standard lines would come out in support of a proposal like this, assuming the numbers I’ve given. Since the ice is already floating, melting it won’t raise sea levels (remember the ice cubes in An Inconvenient Truth). Very few people live in the area, so direct effects on humans would be small. And while the effects on ecoysystems would be devastating, lots of benefit-cost analyses ignore such effects or impute very low costs to them.
Of course, no one is likely to use nuclear explosions in the way I’ve described, but the melting of the Arctic ice cap is a likely consequence of global warming, not in the distant future, but in the lifetime of most people now living (given a predicted date of 2040, I could well be around to see this). And plausible estimates of the cost of stabilising CO2 levels at 550 ppm (probably enough to stop complete melting, though this is not clear) range from 1 to 3 per cent of GDP.
So, it’s not necessary to look at effects occurring around 2100 and argue about discount rates in order to conclude that we ought to be reducing CO2 emissions drastically. Consequences like the melting of the Arctic ice cap and the destruction of the world’s coral reefs are already visibly under way, and will happen within the next few decades.
The problem is that standard benefit-cost analyses, don’t impute any real significant cost to this predictable outcome. For example, Nordhaus and Boyer (more or less arbitrarily) estimate the value of the entire climate-sensitive human and natural environmnent in the US at $500 billion (about equal to the market capitalization of ExxonMobil) and estimate that protecting it is worth $5 billion per year. This figure includes not only all natural ecoystems but impacts on humans in coastal areas like New Orleans, so the value accorded to the natural environment can’t be much more than $2.5 billion per yea (about equal to annual expenditure on chewing gum).
The Stern report doesn’t really tackle this issue, and implicitly accepts estimates like those of Nordhaus and Boyer. This is a big omission, more than offsetting other modelling choices that tend to produce a fovourable benefit-cost assessment of stabilisation.
At this point there are a range of possible responses. One is to accept the economic analysis and conclude that melting the polar ice caps isn’t really a problem, that the destruction of coral reefs is a problem for the tourist and fishing industries, but otherwise unimportant and so on.
The second is to try to extend economic analysis to include some of the factors that are obviously being left out here, for example, by assessing people’s willingness to pay for environmental protection. Of course, this is somewhat problematic with an issue like global warming, since people’s willingness to pay depends on their views on the issue which in turn are influenced by economic analyses.
A third response is to concede that some of these issues are not well handled by economics in our current state of knowledge. The intuition that allowing the Arctic ice cap to melt is likely to have negative consequences we haven’t considered yet is supported by past experience of large-scale human interventions of this kind. But there’s no good way, at present, of incorporating this intuition into discounted cash flows.**
This piece by Paul Baer is well worth reading as a critique of economic analysis on this issue.
* The idea of melting ice caps with nukes has been put forward seriously, but only, as far as I can tell, for other planets, where it is one of many schemes for terraforming.
** My work on the precautionary principle, still in progress, is aimed at addressing exactly this issue.