In addition to sketching out the borders of reasonable debate on climate change, Mike Hulme has written some intelligent things on scientific consensus, as embodied in the Intergovernmental Panel on Climate Change (IPCC) reports:
[T]he use of consensus is merely one (structured) way of distilling evidence – evidence which might be somewhat ambiguous, incomplete or contradictory or where there is latitude for genuine differences of interpretation – into an overall agreed statement on an issue of scientific or public importance.
He also quotes an intelligent comment from a volume by P.N. Edwards and S.H. Schneider:
We have discussed earlier… why it may often be necessary for science to use consensus processes as a way of consolidating knowledge so that it can be useful for policy. Consensus knowledge, by construction, will always allow experts to disagree, with knowledgeable opinion existing at either tail of the distribution of views… Such scientific consensus is not ultimate ‘truth’ and, on occasion, may turn out to be wrong. But the alternatives to the IPCC style of consensus-building are even less likely to command widespread authority within the worlds of science and policy. ‘Vastly better [than random solicitation of views] is the work of groups like the IPCC… which although slow, deliberative, sometimes elitist and occasionally dominated by strong personalities, are nonetheless the best representation of the scientific community’s current general opinion.’
The big problem, from a policy perspective, is the number of politically influential agents who either continue to deny that potentially dangerous anthropogenic climate change is taking place, or who argue for various reasons that nothing ought to be done about it. The fact that these people don’t have views that are reconcilable with the best available evidence doesn’t mean they aren’t able to influence the public policy debate.