There is a (small) chance that the consensus on which they base their recommendations is wrong. As Chris Mooney acknowledges today on openDemocracy:
"While a scientific consensus on global warming...exists, that fact alone does not automatically confer upon the consensus view the status of eternal truth. After all, when the IPCC stated in 2001 that 'most of the observed warming over the last 50 years is likely to have been due to the increase in greenhouse gas concentrations,' the word 'likely' has a precise meaning: a 66-90% chance that the result was true. That also leaves some possibility that it might be untrue".
But this possibility should not drive policy, says Chris Mooney. Rather, he suggests that Immanuel Kant's categorical imperative (“act only according to that maxim by which you can at the same time will that it should become a universal law”) should inform decisions:
"When it comes to science policy decision-making, global-warming sceptics demonstrably fail to do [apply the imperative]. Or, if they do, then they must believe the following things: one, On every issue where scientific dissent exists, politicians should side with that dissent rather than the mainstream view; and two, on every issue where some scientific uncertainty exists, such uncertainty justifies political inaction". (full text here)
The National Academies of Science - which include all of those from the G8 nations, including the US, plus Brazil and China, but not (apparently) India - will recommend cost-effective actions to cut greenhouse gases (see here). This looks like a classic "no-regrets" type recommendation: make the cuts that save money by making energy use more efficient. Will there be no language in it that will meet the more radical suggestions made by Fred Pearce in his memo to the G8 first published on this site last week?
[PS, 8 May: I was wrong to write that the Indian Academy of Science was not a signatory. This error arose from reading a BBC report which does not mention India. The statement is available on. e.g., the web site of the (British) Royal Society here. The language is open-ended: "scientifically-informed targets", "cost-effective steps", "leadership" etc. etc.]