There are so many empty slogans out there I wish we could tackle all of them at once. But the “97% of scientists agree” is surely the elephant in the room. Lots of people have tried to rebut it by dismissing the notion of consensus itself, or by praising the historical examples of renegade scientists who went against a prevailing consensus and turned out to be right. But that unnecessarily concedes the major claim itself, which the evidence shows is simply not true. I hope you enjoy the video, and that you’ll share it widely.
The claim that 97% of the world’s scientists agree is pretty much the ace of trumps in the whole climate debate. After all, who’s going to argue against a consensus that strong, backed by so many experts. But what exactly are they supposed to agree on? If you look behind the curtain, no one seems sure what the experts actually said. Or who they are. Or… anything.
At first glance it seems straightforward enough. In 2013 President Barack Obama famously tweeted that “Ninety-seven percent of scientists agree: #climate change is real, manmade and dangerous.”
In 2014, his Secretary of State John Kerry said 97% of “the world’s scientists tell us this is urgent.” And that same year, CNN said “97% of scientists agree that climate change is happening now, that it’s damaging the planet and that it’s manmade.”
That’s pretty much what most people think when they hear the 97% slogan: Every scientist believes man-made climate change is an urgent crisis.
But there are millions of scientists in the world. How many exactly were surveyed? When were they surveyed? Who did it? And what exactly did they agree on?
Let’s find out. I’m John Robson and this is a Climate Discussion Nexus Fact Check on the 97 percent consensus slogan.
To begin with, there are some ideas that pretty much all scientists accept. For instance that birds are descended from dinosaurs, though that idea was once dismissed as highly eccentric. And when it comes to climate, you don’t need a poll to tell you that carbon dioxide is a greenhouse gas, meaning it likely has some overall warming effect. That’s been known since the mid-1800s. And if you did do a survey, you would find overwhelming scientific agreement on that point.
Also, there are lots of indications that the world is somewhat warmer now than it was in the mid-1800s, the end of a natural cooling period called the Little Ice Age.
Finally, virtually nobody disputes that humans have changed the environment of our planet, by releasing emissions into the air, changing the land surface, putting things in the water, and so forth.
These aren’t controversial ideas, and they’re accepted even by most climate skeptics. What we don’t accept is that any of them prove that humans are the only cause of global warming, or that climate change is a dangerous threat.
If 97% of scientists believed that, it would be troubling. Though even so, we’d still have to find some plan whose benefits outweighed its costs. In any event, that level of consensus that the problem was manmade and urgent would certainly be noteworthy. But the thing is, they don’t agree on that.
A close look at what survey data we have, and there isn’t much, tells us, yes, there is a great deal of agreement that CO2 is a greenhouse gas to some degree, that the Earth has warmed in the last 160 years, and that humans affect their surroundings. But that survey data also tell us there’s far less agreement on everything else including whether we face a crisis.
So where did this 97% claim come from and why is it so widely repeated?
The 97% claim seems to have begun with a historian of science named Naomi Oreskes who, in 2004, claimed she’d looked at 928 articles about climate change in scientific journals, that 75% of them endorsed the “consensus view” that “Earth’s climate is being affected by human activities” and that none directly disputed it.
By 2006, in Al Gore’s An Inconvenient Truth, this finding had somehow morphed into “a massive study of every scientific article in a peer-reviewed journal written on global warming for the last 10 years and they took a big sample of 10%, 928 articles, and you know the number of those that disagreed with the scientific consensus that we’re causing global warming and that it’s a serious problem? Out of the 928, zero.”
That was a fib. Gore took a study that found 75% endorsed the idea that humans have some effect on climate and turned it into proof that 100% of scientists believe it’s a serious problem. It does no such thing.
And nor do the handful of other surveys on the subject. For instance five years later, in 2009, a pair of researchers at the University of Illinois sent an online survey to over 10,000 Earth scientists asking two simple questions: Do you agree that global temperatures have generally risen since the pre-1800s? and Do you think that human activity is a significant contributing factor? [Note: They asked some other questions too, but didn’t report the questions or results in the publication.]
They didn’t single out greenhouse gases, they didn’t explain what the term “significant” meant and they didn't refer to danger or crisis. So what was the result?
Of the 3,146 responses they received, 90 percent said yes to the first question, that global temperatures had risen since the Little Ice Age, and only 82 percent said yes to the second, that human activity was a significant contributing factor.
Interestingly, among meteorologists only 64 percent said yes to the second, meaning a third of the experts in the study of weather patterns who replied didn’t think humans play a significant role in global warming, let alone a dominant one.
What got the most media attention was that among the 77 respondents who described themselves as climate experts, 75 said yes to the second question. 75 out of 77 is 97%.
OK, it didn’t get any media attention that they took 77 out of 3,146 responses. But that’s the key statistical trick. They found a 97 percent consensus among 2 percent of the survey respondents. And even so it was only that there’d been some warming since the 1800s, which virtually nobody denies, and that humans are partly responsible. These experts didn’t say it was dangerous or urgent, because they weren’t asked. [Note: or as noted above, if they were the results weren’t reported.]
So far the claim that 97% of “world scientists” are saying there’s a climate crisis is pure fiction. But wait, you say. There must be more. Yes, there is. But not much.
Another survey appeared in 2013, by Australian researcher John Cook and his coauthors, in which they claimed to have examined about 12,000 scientific papers related to climate change, and found that 97% endorsed the consensus view that greenhouse gases were at least partly responsible for global warming. This study generated headlines around the world, and it was the one to which Obama’s tweet was referring.
But here again, appearances were deceiving.
Two-thirds of the papers that Cook and his colleagues examined expressed no view at all on the consensus. Of the remaining 34%, the authors claimed that 33% endorsed the consensus. Divide 33 by 34 and you get 97%. But this result is essentially meaningless, because they set the bar so low.
The survey authors didn’t ask if climate change was dangerous or “manmade”. They only asked if a given paper accepted that humans have some effect on the climate, which as already noted is uncontroversial. It could mean as little as accepting the “urban heat island” effect.
So a far better question would be: How many of the studies claimed that humans have caused most of the observed global warming? And oddly, we do know. Because buried in the authors’ data was the answer: A mere 64 out of nearly 12,000 papers! That’s not 97%, it’s one half of one percent. It’s one in 200.
And it gets worse. In a follow-up study, climatologist David Legates read those 64 papers and found that a third of them didn’t even say what Cook and his team claimed. Only 41 actually endorsed the view that global warming is mostly manmade. And we still haven’t got to it being “dangerous”. That part of the survey results was simply invented, by politicians and activists.
Other researchers have condemned the Cook study on other grounds too. For instance economist Richard Tol showed that over three-quarters of the papers counted as endorsing even the weak consensus actually said nothing at all on the subject. And evidence later emerged that the authors of the paper were drafting press releases about their findings before they even started doing the research, which indicates an alarming level not of warming or of consensus but of bias.
The reality is that neither this study, nor a handful of others like it, prove that 97% of scientists believe climate change is mostly manmade, let alone that it’s a crisis. The fact that people who claim to put such stock in “settled science” accept such obvious statistical hocus pocus is both astounding and disappointing.
So what do climate experts really think? The year before Obama sent out his tweet, the American Meteorological Society (AMS) surveyed its 7,000 members. They got about 1,800 responses. Of those, only 52% said they think global warming over the 20th century has happened and is mostly manmade. The remaining 48% either think it happened but is mostly natural, or it didn’t happen, or they don’t know. And while it’s possible that the three-quarters who didn’t answer split the same way as those who did, it’s also possible that committed alarmists are more likely to answer such surveys. In any case, it’s a small sample, even of AMS members, let alone of the world’s scientists.
There was one more survey a few years later by the Netherlands Environment Agency that claimed 66% of climate experts believed humans were mostly responsible for warming since 1950. Which falls far short of 97% even if it outperforms the other studies.
A social psychologist named Jose Duarte, who specializes in survey design, published an analysis of that one, pointing out that they diluted the sample by including large numbers of psychologists, philosophers, political scientists, and other non-experts, making their results meaningless as a measure of what scientists think. Just as you’ll find that the people who cite that 97% number are overwhelmingly not trained scientists, certainly not trained statisticians.
So we’re no farther ahead than when we began. Most experts agree on the basics, namely that carbon dioxide is a greenhouse gas and probably causes some warming and that humans have some impact on climate probably including some warming. But they actively debate the rest: How much warming will there be? Is it a problem? Should we try to stop it, or adapt, or wait and see? These are all important questions and we need good answers.
And there's the claim that many of the world’s national science academies, representing hundreds of thousands of scientists across the globe, have issued statements supporting the consensus about global warming and demanding government efforts to cut emissions. The problem is, not a single one of those societies took a survey of their members before issuing their statements in the name of their members. The statements were put out by a small number of activists using their committee positions to make it look as though their views are shared by all the world’s experts. But if they are, why didn’t these authors survey their members before publishing the statements?
There are a couple of other studies that claimed to prove a consensus. But they run into the same problems. All they show is wide agreement on the uncontroversial bits. They offer no information about whether a majority of scientists think global warming is a crisis. And then they’re spun wildly by non-scientists to tell us things they don’t begin to say, often about questions they didn’t even attempt to investigate .
The problem isn’t just that we don’t know what percentage of scientists agrees with this or that statement about global warming. It’s something much worse. All this talk of a 97% consensus amounts to a dishonest bullying campaign to stifle scientific debate just when we need it most because the question looms so large in public policy.
As physicist Richard Feynmann once said, “I would rather have questions that can't be answered than answers that can't be questioned.” And that’s especially true when we’re asked to take drastic action based on those answers.
Not long ago that survey expert I mentioned earlier, Jose Duarte, warned his fellow scientists about the negative consequences of claiming consensus. He said:
“It is ill advised to report a consensus as though it is an aggregation of independent judgments. Humans are an ultrasocial species, and dissent is far costlier than assent to a perceived majority… A scientist who contests the prevailing narrative on human-caused warming, or merely produces smaller estimates, will likely end up on a McCarthyite blacklist of ‘deniers’. Self-described mainstream climate scientists refer the public to such lists, implicitly endorsing the smearing of their colleagues. This is disturbing, and unheard of in other sciences.”
The unfortunate truth is that there is strong political pressure for climate experts not to question claims of impending doom. Those who do so face steep personal and professional costs, including a barrage of abuse that can be highly unpleasant for people who mostly wanted to devote their lives to the quiet pursuit of knowledge not to noisy polemics. And that means we should listen carefully to them when they feel compelled to speak out anyway.
Whether they represent 50%, or 10%, or 3% of experts, what matters is the evidence they bring and the quality of their arguments.
And on that, I would hope we have 100 percent agreement.
For the Climate Discussion Nexus, I’m John Robson.