VIP Very important piece.
The ’97 per cent’ figure is supported by three different published articles, with a forerunner by Naomi Oreskes…In 2004 she looked at 928 abstracts of articles in the climate science field. According to her, 75 per cent supported the view that human activities were responsible for most of the warming in the last fifty years…She had not interviewed them. Instead, she had looked at the abstracts of their articles, and come to a view about what their authors must have thought. Why those 928? Well, they were the papers in the ISI database from 1993 to 2003 that had the words ‘climate change’ as a tag. Ms Oreskes seemed somehow to have excluded articles by scientists such as Christy, Lindzen, Michaels and Idso, all of them sceptics, and somewhat to their surprise. What was the method of evaluation? She divided the papers into six groups and found that 75 per cent of them either explicitly or implicitly accepted the ‘consensus view’. What was that? In her words: the evidence for human modification of climate is compelling. Is that a bad thing? That seems not to have been part of her survey.
In 2009 Zimmerman and Doran asked scientists two questions: did they think that temperatures had risen and whether humans were significantly responsible. Again, no mention of dangerous consequences, but at least the authors did actually ask some scientists what they thought. But then the methodology gets very sloppy, and I’ll summarise it like this. They used an online survey of 10,257 members of the American Geophysical Union, whose membership is around 60,000. The respondents seemed to be the right ones to interview, given their fields of interest, but only 3,146 actually replied. Now they excluded nearly all of those who had replied, for one reason and another, to produce 79 scientists who said they were climate scientists and had published more than half of their work on ‘climate change’. Of them 77 both thought that temperatures had risen and that humans were significantly responsible. The fraction 77/79 gives you 97 per cent.
I read that paper when it first appeared and could not understand how it was published. I was young and naïve in those days.
The crème de la crème comes with the work (if that is right term for it) of John Cook, occasionally aided by Stefan Lewandowsky…In 2013 Cook et al and a team of volunteers looked at more than 12,000 abstracts, rated them according to whether or not they implicitly or explicitly endorsed the view that human activity had caused (wait for it) some of the warming, and again found the magic 97 per cent. See — it’s true! Surely those three separate ratings of 97 per cent have something going for them.
On the face of it, no. Unfortunately for Cook, Legates and others later in the same year published a rebuttal. They found that only 41 papers – 0.3% of all 11,944 abstracts or 1.0% of the 4,014 expressing an opinion, and not 97.1% – had been found to endorse the claim that human activity is causing most of the current warming. Elsewhere, Craig Idso, Nicola Scafetta, Nir J. Shaviv and Nils-Axel Morner and other climate scientists protested that Mr. Cook ignored or misrepresented their work. Cook has been trying to defend his results ever since, but more and more scorn has, in my view quite rightly, been poured on the work. You can read some of the objections here, here and here, for starters. As I have said before, this is terrible stuff methodologically, the worst I’ve ever seen in a peer-reviewed journal.
I read that paper as well, very closely, looking at the primary data. It was interesting that two thirds of the abstracts, at the first reading, yielded no opinion on the consensus they were looking for.