A Failed Climate Coup in the Courts
The Federal Judicial Center tried to pass off one-sided propaganda as ‘settled science.’
Feb. 8, 2026
Progressive climate orthodoxy hasn’t disappeared in the Trump era, and the latest evidence comes in a strange tale from a surprising place—the bureaucracy of the federal courts.
In December the Federal Judicial Center, the research and education agency for the federal courts, published its latest edition of the Reference Manual on Scientific Evidence. A joint product with the National Academies of Sciences, Engineering, and Medicine, the manual is billed as a resource for judges deciding complex scientific cases. But instead of a cure for insomnia, the new edition included some political pamphleteering.
The manual begins with a foreword by Supreme Court Justice Elena Kagan, who writes that while judges are “generalists” who often learn about scientific issues “through the adversary process, . . . sometimes it also helps to have a dispassionate guide.” Judges will do their job better when they learn from the technical expertise of scientists, she writes, and “the law will become stronger as it further reflects sound science.”
She’s right about that, but this manual included some unsound science. The guide advises judges on how to handle a case of disagreement between scientific experts. Judges should consider, “Is this a case of truly unsettled science?” the guide poses, “Or is it a case of relatively settled science, in which one party has recruited experts who interpret the evidence differently than most of the community?”
In case judges didn’t get the hint, the paragraph concludes this “sometimes occurs as a result of strategic manipulation from stakeholders who stand to be harmed if the public were to understand the true state of scientific consensus” such as in the cases of “tobacco, ozone depletion, and climate change.”
That trio gives the game away: Anyone who disagrees on climate is akin to those who deny that cigarettes cause cancer, A footnote to the section refers readers to “Michael E. Mann, The New Climate War: The Fight to Take Back our Planet” and “Naomi Oreskes, The Scientific Consensus on Climate Change.” These are polemical texts.
The manual’s Reference Guide on Climate Science was written by Jessica Wentz and Radley Horton of Columbia University. Ms. Wentz is at the Sabin Center for Climate Change Law, whose “core mission” is to “develop and promulgate legal techniques to combat the climate crisis and advance climate justice.” She also works with the Environmental Law Institute’s Climate Judiciary Project, which says it helps judges “understand the scientific and technical evidence before them.”
The good news is that this one-sided advocacy caught the attention of House Judiciary Committee members Jim Jordan (R., Ohio) and Darrell Issa (R., Calif.), who wrote to the Judicial Conference of the United States. The “scientific” contributions to the manual “appear to have the underlying goal of predisposing federal judges in favor of plaintiffs alleging injuries from the manufacturing, marketing, use or sale of fossil fuel products,” they wrote. Why are taxpayers funding an exercise in judicial indoctrination?
Twenty-seven state Attorneys General—including from West Virginia, Florida, Idaho, Montana, New Hampshire, Texas and Wyoming—also objected to the manual’s bias and asked that the chapter be withdrawn. Article III of the Constitution “guarantees every litigant . . . the right to an independent and impartial tribunal,” the AGs write.
And, what do you know, on Friday Judge Robin Rosenberg, director of the Federal Judicial Center, wrote to West Virginia Attorney General John McCoskey saying “the climate science chapter” has been omitted from the revised Reference Manual. The letter offered no explanation, but you can guess.
This might be a case of all’s well that ends well, except that someone on the Judicial Center was either asleep or tried to slip ideology into what should be “a dispassionate guide,” to borrow Justice Kagan’s words. A public accounting of how that happened would be useful.
