Claim: 97% consensus found for “many decades”

One has to be on one’s guard when John Cook is presenting the result of his findings. Previous post was about Cook illustrating his research with exaggerated claims from three politicians. This post will be about his first sentence in that video, just after the statements of the politicians:

For many decades, study after study have found that 97% of publishing climate scientists agree that humans are causing global warming.

That specific claim drew my attention. It seems to suggest that many studies found that 97% result over a long time frame. From what I learned about the consensus papers, that didn’t agree very well with reality.

Let’s take it part by part and see where we get.

“For many decades”

That seemed odd. As far as I know, the first “study” was published in 2004, which is confirmed in the consensus handbook:

Naomi Oreskes was the first to quantify the level of expert agreement on human-caused global warming in 2004.

When we now look at the publishing dates of the Oreskes paper (December 3, 2004) and the Cook 2016 paper (April 13, 2016), then the difference amounts to almost 11.5 years. That is somewhat more than one decade, by no means “many decades”. There should at least two, and preferably more, decades in order to state “many decades”.

Then where does this “many decades” claim comes from?

There is a table of estimates of consensus in the Cook 2016 paper. It lists polls, surveys and scientific papers. Of the 14 entries in total, there are 13 entries between 2004 (Oreskes paper) and 2015 (last paper described in Cook 2016). Before 2004 there is … only 1 entry, a poll of AMS/AGU members in 1991. From 1991 until 2015 is 25 years (2.5 decades).

So far, so good.

Okay, but what is his definition of a “study”? That brings us to the next part of the statement:

“study after study”

That first “study” is a poll of the members of an organization, so to get to the “many decades” claim, he had to count all instances of polls, surveys and scientific papers. Apparently, the definitions used by Cook are:

• to calculate the consensus time frame: scientific papers, surveys and polls
• everywhere else: only scientific papers.

The redefinition of the time frame and what is a “study” may seem insignificant, but it adds almost 1.5 decades to the time frame, therefor justifying the “for many decades, study after study” claim. If he would have used the same definition, then this time frame would be cut in half…

It gets more interesting in the next part of the statement. Watch carefully how the pea is moving from one shell to another:

“have found that 97%”

Keep in mind that to get to get to this point, Cook counted all instances of polls, surveys and scientific papers. Now he adds the 97% figure, therefor suggesting that this 97% figure was found over the entire period.

The pea has arrived in a new shell…

The Cook 2016 paper shows a different story. The 1991 poll resulted in a 67% figure and there was no data until 2004. That would exclude the first half of the consensus period, undermining the base of the “many decades” claim.

Also, not all “studies” result in a 97% figure. Five have 97% or close to that. One was higher (Oreskes). All the rest get a result lower than 97%. So most of the “studies” get a result lower than the 97%.

Those who are new to the subject would understand from this statement that the 97% consensus figure was consistently found over many decades, which is not what was found in his own paper that he is describing in the video. It was found in only half of the period and in a majority of the papers found less than 97%. Which is not something one would get from hearing this statement.

He does something similar in the next part of the statement:

“of publishing climate scientists”

This is not exactly true. For example, the first “study” in 1991 was a poll of AMS/AGU members. If we have to believe that there is a (97%) consensus for “many decades”, then logically this claim could only be true if all AMS/AGU members were publishing “climate scientists” in 1991. I don’t believe for one moment that there were 400 publishing climate scientists in 1991.

As far as I know, not all the papers that get a result of 97% or higher investigated the agreement of publishing scientists. For example, the Oreskes paper didn’t look at the qualifications of authors. If only papers with 97% or higher investigating the opinions of publishing climate scientists, then the consensus period would shrink well below 1 decade.

But then, does it really matter whether this 97% consensus is there for 2.5 decades or for only less than a decade? This brings us to the heart of the debate:

“agree that humans are causing global warming”

What is his exact definition of “humans causing global warming”? Throughout the video and the consensus handbook, the same vague words are used (“humans causing global warming” or “human-caused global warming”) without definition.

The Cook 2016 paper was a collection of several papers on consensus, so there were several definitions used. There is one common denominator though: they all attempt to quantify the agreement on the anthropogenic nature of the warming. But none of those papers went further than that. None specifically investigated for example impacts or whether political action would be needed.

For example, these are the three top categories with the example that was used in the Cook own 2013 paper (and were the basis of the 97% finding):

1. Explicit endorsement with quantification
   (explicitly states that humans are the primary cause of recent global warming):
   Example: “The global warming during the 20th century is caused mainly by increasing greenhouse gas concentration especially since the late 1980s”
2. Explicit endorsement without quantification
   (explicitly states humans are causing global warming or refers to anthropogenic global warming/climate change as a known fact):
   Example: “Emissions of a broad range of greenhouse gases of varying lifetimes contribute to global climate change”
3. Implicit endorsement
   (implies humans are causing global warming. E.g., research assumes greenhouse gas emissions cause warming without explicitly stating humans are the cause):
   Example: “..carbon sequestration in soil is important for mitigating global climate change”.

According to the definitions and examples above, my position would land in the second category, but I really doubt that Cook and the authors of the handbook would classify me like that (I guess they would classify me as a contrarian, a denialist or something of that like).

From earlier papers authored by Cook, I remember that he made it no secret that the importance of communicating the consensus is to facilitate political action. This is not different in the handbook. For example, the paragraph on “The effectiveness of consensus messaging” is revealing:

A number of studies show that consensus messaging is a powerful communication tool (see page 14). Simply communicating the current state of scientific agreement (97%) not only raises perceived consensus, it also has a positive influence on acceptance that global warming is real, human-caused, and is a serious problem. Most importantly, it increases support for climate policy. The 97% consensus offers a lot of bang for one’s communication buck.

This is clearly written by someone who goes waaaay beyond (his own) category 1.

There is a big difference in style. The definitions of “human-caused global warming” used by Cook are:

• To quantify the 97% consensus in a scientific paper: warming caused by our emissions of greenhouse gases
• To communicate the 97% consensus to the public: dangerous warming caused by our emissions.

There surely is a consensus on the first definition, but not on the second.

That is not honest. If an investigator found an agreement on statement A (which has been researched), but in communication about that agreement claims to have found agreement on a more far-fetching statement B (which has not been researched), then he is deceiving his audience.

The goal of this video should have been to illustrate the Cook 2016 paper, but all these tricks morphed the actual findings into something completely unrecognizable. It now suggest that this 97% figure was found in all/most/many papers over the complete range of those “many decades”. Viewers will certainly be impressed by this long-term agreement, but it is not what was found. Wasn’t the intention of the video to explain the findings of the paper?

What he basically did was:

• By the fluid use of the word “study”, he increased the period of consensus from 12 to 24 years
• By using the 97% number in the same sentence as “many decades, study after study”, he suggested that these were spread over the entire period
• By using publishing climate scientists in the same sentence as “many decades, study after study”, he suggested that these were all or at least most of the ones that were surveyed over the entire period.

These tricks allowed him to exaggerate the period and level of 97% agreement. He also doesn’t mention that he researched the agreement on only the most trivial meaning of causing global warming (the Cook 2016 paper only investigated the anthropogenic aspect of the warming).

When it comes to communication, that might be a very smart move, but it is not very truthful…

Advertisement