Assessing scientific opinion on the relative roles of natural and human-caused factors in recent climate change


In recent years, a number of studies have found that 90-95% of scientists agree that the climate changes and/or that global temperatures are warmer now than in the 19th century. This result seems to be quite well replicated.

However, these same studies are widely used by political activists and advocacy groups to imply that 90-95% of scientists also agree with the specific claim that recent climate change is mostly human-caused, and that anybody who questions this claim is anti-science. This has worrying implications for the scientific community, as it actively discourages scientific inquiry into the important scientific problem of “climate change attribution”.

Moreover, a careful inspection of the results of these surveys reveals that it is false. For instance, below we describe the results of three of these surveys:

As can be seen from the figure, all three of these surveys confirm that at least 90-95% of scientists agree that “climate changes” and/or that there has been some “global warming”.



But, when the respondents are asked on whether this climate change/global warming is mostly human-caused or mostly natural, there are a wide range of opinions



In a 2015 study in the journal Science & Education, one of us carried out a detailed reanalysis of another study, Cook et al. (2013), which has been widely cited as proof of a “97% scientific consensus” that recent global warming is “mostly human-caused”.

In the Cook et al. (2013) study, the authors examined nearly 12,000 abstracts of papers containing the keywords “global climate change” or “global warming”. They sorted the abstracts into 7 categories depending on what position the abstract implied on the human contribution. Cook et al. (2013) implied that 97.1% of the abstracts agreed “that human activity is very likely causing most of the current [global warming]”.

However, in Legates et al. (2015), we re-analysed the Cook et al. (2013) results and showed that Cook et al. had only found 0.5% of the abstracts to have explicitly made that claim. They had found that two thirds of the abstracts had provided no position on whether global warming is mostly human-caused or mostly natural, and only 8% of the abstracts had explicitly stated any opinion on this issue:



Of the 8% of abstracts which made an explicit claim on the human contribution, the vast majority (91%) apparently did not offer any opinion on whether recent global warming is mostly human-caused or mostly natural. Instead, they just stated that at least some of the global warming is human-caused.

According to the Cook et al. (2013) results, only 88 of the 11,944 abstracts explicitly stated whether global warming was mostly human-caused or mostly natural. 64 claimed it was mostly human-caused and 24 claimed it was mostly natural.



Moreover, when we re-analysed the abstracts themselves we found that even these figures were unreliable. We found that only 41 of the 64 abstracts which Cook et al. had rated as “mostly human” had explicitly made that claim. We also identified several abstracts which Cook et al. had mistakenly not included in the “mostly natural” categories.

Nonetheless, our main conclusion was that Cook et al. (2013)’s widely-cited claims of “a 97% scientific consensus” on climate change attribution are completely unjustified.


For more information, see:

D. R. Legates, W. Soon, W.M. Briggs and C. Monckton of Brenchley, 2015. “Climate Consensus and ‘Misinformation’: A Rejoinder to Agnotology, Scientific Consensus, and the Teaching and Learning of Climate Change”. Science and Education. Vol. 24, pp299-318, Link to journal abstact ; Link to pre-print pdf.