The climate science consensus is 78 – 84 percent (Updated)

I think I've been negligent in not making the actual climate science consensus known to people. In one of my projects it occurred to me that no one has made the information readily available, at least not to my knowledge. I'll save the details for the journal article, but I think a quick snapshot of the consensus might be helpful.

First, there is no 97%.

There is not a single survey of climate scientists that reports 97% agreement with the proposition that most of the observed warming was caused by human activity.

There are three recent and high quality studies that surveyed climate scientists on this question or its semantic equivalent. The scholars who performed this work are qualified, competent survey researchers, and in one case the study was published in an esteemed international journal of survey research. Here are the results:

• 84%*
  
• 81%
  
• 78%
  

The researchers, respectively:

Farnsworth, S. J., & Lichter, S. R. (2012). The structure of scientific opinion on climate change. International Journal of Public Opinion Research, 24(1), 93-103. Link

Bray and von Storch. "A survey of the perceptions of climate scientists 2013." Helmholtz-Zentrum Geesthacht, Geesthacht. (2014). Link

Stenhouse, Neil, et al. "Meteorologists' Views About Global Warming: A Survey of American Meteorological Society Professional Members." Bulletin of the American Meteorological Society 95.7 (2014): 1029-1040. Link

NOTE: I am not aware of any other studies published in the last five years that satisfy the requirements here – direct survey of climate scientists on the key attribution question. Other studies are either not direct surveys, do not restrict their samples to climate scientists, or do not ask whether humans are responsible for most of the warming. (There are earlier versions of the Bray and von Storch and Stenhouse et al. studies – I report the most recent surveys.) For example, Doran and Zimmerman, Anderegg et al., and Verheggen et al. do not satisfy the criteria. For more on the Verheggen study, see my paper in Environmental Science & Technology.

* It may be more accurate to say 78-81%, since Farnsworth and Lichter ask a more ambiguous question – whether "human-induced greenhouse warming is now occurring", not explicitly whether most of the warming is human-caused. The explicit surveys give us 78-81%.

The 97% meme is scam that will likely be a formal topic of study for future historians and other scholars. It arose chiefly from a fraudulent and invalid study – Cook et al (2013). This was not a survey of climate scientists, but rather a study where a team of activists read academic paper abstracts and decided what they mean. Setting aside for a moment the fraud that was later revealed, the study was based on a faulty search of broad academic literature using casual English terms like "global warming", which missed lots of climate science papers but included lots of non-climate-science papers that mentioned climate change – social science papers, surveys of the general public, surveys of cooking stove use, the economics of a carbon tax, and scientific papers from non-climate science fields that studied impacts and mitigation. The team seemed to have no idea how to search scientific literature and unfamiliar with meta-analysis techniques**.

The team of activists wanted to deliver a high consensus figure to advance their political cause – an impossible conflict of interest. The paper includes repeated lies about their methods, and there are no valid findings from the study. No estimate of a consensus can be computed from their data by any method known to science. The journal editor who published and promoted the paper is Obama advisor and campaign donor Daniel Kammen, which created a massive conflict of interest when the fraud was disclosed to him and the journal – Obama had already famously cited and promoted the false finding, as it served his policy priorities. Kammen and the journal have done nothing to manage that conflict of interest, and have yet to retract the paper.

Both scientists and the media have almost exclusively cited the junk studies conducted by unqualified political people like Cook and Oreskes, rather than qualified researchers. The junk studies generated the higher estimates, which is probably why they were cited more. The valid scientific studies, performed by trained researchers, have largely been ignored. This hints at a larger problem, may be an example of something like Gresham's Law, and will be more thoroughly explored in peer-reviewed literature.

Tips for being a good science consumer and science writer. When you see an estimate of the climate science consensus:

• Make sure it's a direct survey of climate scientists. Climate scientists have full speech faculties and reading comprehension. Anyone wishing to know their views can fruitfully ask them. Also, be alert to the inclusion of people outside of climate science.
• Make sure that the researchers are actual, qualified professionals. You would think you could take this for granted in a study published in a peer-reviewed journal, but sadly this is simply not the case when it comes to climate consensus research. They'll publish anything with high estimates.
• Be wary of researchers who are political activists. Their conflicts of interest will be at least as strong as that of an oil company that had produced a consensus study – moral and ideological identity is incredibly powerful, and is often a larger concern than money.
• In general, do not trust methods that rest on intermediaries or interpreters, like people reviewing the climate science literature. Thus far, such work has been dominated by untrained amateurs motivated by political agendas.
• Be mindful of the exact questions asked. The wording of a survey is everything.
• Be cautious about papers published in climate science journals, or really in any journal that is not a survey research journal. Our experience with the ERL fraud illustrated that climate science journals may not be able to properly review consensus studies, since the methods (surveys or subjective coding of text) are outside their domains of expertise. The risk of junk science is even greater if the journal is run by political interests and is motivated to publish inflated estimates. For example, I would advise strong skepticism of anything published by Environmental Research Letters on the consensus – they're run by political people like Kammen.

Outlets like Chris Mooney, Scientific American, DeSmogBlog, ClimateWire, and the misnamed ScienceBlogs site are not alert to fraud and junk science if it promotes their political agenda. Channeling von Clausewitz, for those people science is just politics by other means. They'll cite and promote this stuff, and they won't cite actual scientific research – they've not reported the professional surveys above. None of them have yet retracted or corrected their promotion of the Cook fraud. When media and science writers start reporting the Cook fraud, outlets like Mooney and SciAm will probably be the very last to acknowledge the fraud, if ever. Our civilization is not in good shape in terms of how we manage the effects of politics on science, but we'll get better.


** In an earlier version of this post, I said that the Cook researchers were not scientists (I meant to say climate scientists.) Dana Nuccitelli, second-author of the Cook paper, objected to that claim. With this study, there are two groups of people who might be termed the researchers – the raters who conducted the study, and the authors of the paper, groups that only partially overlap. The raters were not generally scientists, including in their ranks a luggage entrepreneur and a blogger for whom English is a second language (Jokimaki, who displayed notable scientific promise on the fraud-revealing rater forum.) There are some scientists among the authors, e.g. Sarah Green, but not climate scientists. I've removed that clause. Only climate scientists would be qualified to interpret climate science abstracts, and even then they wouldn't understand some of them and would not be blind to the work of colleagues and rivals. This vague subjective rating method is not promising.

Addendum:

After I first posted, Georgia Tech climate scientist Judith Curry argued that the relevant figure for the Stenhouse et al study is much less than 78%. For the self-identified professional field category "Meteorology and Atmospheric Science", the consensus is 61%.

The 78% figure I cited is from the professional category "Climate Science." I automatically chose the highest estimate to be conservative in my report. To laypeople or even scientific outsiders, the difference between atmospheric science and climate science is unclear, but it's quite common for similar-sounding terminology to carry major distinctions among researchers. Curry's intuition is that the "climate science" people likely work on climate impacts, and that she would have chosen "atmospheric science" to classify herself had she participated, even though she is a climate scientist in the common use of the term. She further reports that the atmospheric scientists are the experts on attribution, and therefore their agreement carries more weight than the self-identified climate scientists in this study.

This raises an important issue – who are the real experts, and how do we identify them? You might think that anyone who is a climate/atmospheric scientist is an expert on the human contribution to global warming, but I suspect that most climate and atmospheric scientists would disagree with that. It's 2015, and science is very specialized. I'm not sure there are more than 200 experts on climate change attribution, or specifically atmospheric warming attribution. There might not even be 100.

For now, I'm leaving the 78% figure, for at least one reason. In their discussion, Stenhouse et al. report that they asked respondents about warming over the last 150 years. Six respondents e-mailed them to say that their answers would have been different had asked only about the last 50 years. They're not explicit, but I take the implication to be that answers would have changed from less attribution or confidence in human forcing to more attribution. I don't know how many other respondents' answers were shaped by the wording of that question, so for now I go with the highest estimate, 78% from the "climate science" category rather than the 61% from "meteorology and atmospheric science."