Growing up in the 80s and 90s in Chevy Chase, Maryland, an inside-the-Beltway suburb, I only learned one thing about fossil fuels: they were causing global warming. That is, the CO2 my parents’ SUV was producing was making the Earth a lot hotter and that would make a lot of things worse. Oh, and one more thing: that this was a matter of scientific consensus. Looking into the issue a bit, I found that there were professionals in climate science, such as Richard Lindzen of MIT, and Patrick Michaels of the University of Virginia, who said that global warming wasn’t the big deal it was made out to be. But they seemed to be very much in the minority. Who was right? Of course, I knew the majority isn’t always right—but it certainly isn’t always wrong. What was I supposed to make of all this? I think this is a predicament most of us experience. On the one hand, there is something authoritarian about calls to obey “consensus” such as John Kerry’s recent “When 97 percent of scientists agree on anything, we need to listen, and we need to respond.” On the other hand, there is something anti-science about the militant skepticism of some critics of the “climate change consensus.” For instance, ExxonMobil CEO Rex Tillerson says: “The term scientific consensus is an oxymoron in itself.” Not true. How can we possibly function in a complex division-of-labor society if we don’t consult experts—which includes learning about what there is consensus on (and what there isn’t) among the experts in different fields? Scientific consensuses are an important part of any modern society—they tell us the general state of agreement in a field, not so we can blindly obey the experts in question (experts and consensuses can be wrong) but so that we can understand and critically think about those experts’ views. For example, if you are thinking about nutrition, it is a valuable starting point to know where there is general agreement, where there isn’t, and why. If I read a book endorsing a controversial diet, I can’t really have a responsible opinion until I know what most experts in the field think about the issues—including whether they have powerful arguments against the book’s claims that I couldn’t have thought of myself. Thus, statements of scientific consensus can be extremely valuable tools. But they are only valuable, and only scientific, if that are explained clearly to the public. We need to know exactly who agrees with what for what reasons, and just as importantly, where there is disagreement within the consensus and for what reasons. For example, it makes a big difference if there is a consensus that there is some global warming vs. a consensus that there will be catastrophic global warming. It makes a big difference if the consensus is based on issues that the experts have expertise on, such as climate records, vs. issues that they do not have expertise on, such as the economics of fossil fuels vs. solar and wind. Most consensus statements, however, are very unclear on who agrees with what and why. They are unscientific consensuses—misrepresentations of the state of scientific opinion designed to further a political agenda. Take the consensus statement of the American Geophysical Union, which can be found in its entirety here. Like most consensus documents, it starts with something there is definitely a consensus on: “Extensive, independent observations confirm the reality of global warming.” But then, with equal certainty, it cites dramatic predictions of climate models that, even the Intergovernmental Panel on Climate Change reluctantly acknowledged, demonstrably failed to predict the climate of the past two decades. And still, with equal certainty, it calls for “urgent” political action to reduce fossil fuel use—with no acknowledgment of the cost of doing so. Are observations, dramatic model predictions, and complex political decisions really all on the same scientific footing? No—but this kind of statement makes it seem as if they are all a matter of expert consensus. I have spent quite a bit of time querying experts on this issue, and in my understanding the actual consensus in the field is something like the following.