Process Wins One: Federal Judicial Center Deletes Climate Chapter from Judicial Manual

Charles Rotter

Every so often, an institution pauses, looks at what it has just done, and quietly backs away from the edge. That appears to be what has happened at the Federal Judicial Center (FJC), and it is worth marking the moment.

As confirmed in a February 6, 2026 letter from FJC Director Judge Robin L. Rosenberg to West Virginia Attorney General John B. McCuskey, the Federal Judicial Center has removed the climate science chapter from the Reference Manual on Scientific Evidence, Fourth Edition. That single administrative act—“has omitted the climate science chapter”—represents a rare and welcome course correction.

The Fourth Edition of the Reference Manual is no ordinary publication. Courts have cited earlier editions more than a thousand times. Judges rely on it not to tell them what to think, but to help them understand how to evaluate expert claims without becoming surrogate scientists. Its authority flows precisely from its reputation for restraint.

The now-removed climate chapter threatened that reputation.

The problem was never that climate science exists or that climate changes. The problem was that the chapter treated deeply disputed questions—especially climate attribution—as if they were settled background facts. In doing so, it blurred the line between education and advocacy, effectively supplying judges with a pre-framed lens through which ongoing and future climate litigation could be viewed.

That did not go unnoticed. In late January, a coalition of state attorneys general, led by West Virginia, formally objected. Their letter argued that the chapter “places the judiciary firmly on one side” of unresolved scientific and legal disputes and risks turning the Manual into something resembling a quasi–amicus brief for climate plaintiffs. That is strong language, but the underlying concern was straightforward: courts are supposed to test claims, not absorb them prepackaged.

The timing made the issue even more sensitive. Climate litigation increasingly hinges on attribution—attempts to link specific storms, floods, or heatwaves to global emissions, and then to particular companies or states. These are not minor technical questions. They sit at the intersection of complex models, sparse data, and cascading assumptions. If attribution were genuinely settled science, there would be little need for protracted expert battles in court. The fact that those battles dominate climate cases tells its own story.

Against that backdrop, a judiciary-support institution publishing what looks like an endorsed framework for attribution is not neutral. It risks tilting the playing field before a single expert is sworn in.

Congress had already begun asking uncomfortable questions about climate-focused judicial education and whether advocacy organizations were exerting undue influence through “training” and “guidance.” The FJC’s climate chapter landed squarely in that line of fire. Faced with mounting scrutiny from both state officials and Congress, the Center chose not to defend the chapter but to remove it.

That decision deserves cautious applause.

This was not a scientific retraction in the journal sense. No datasets were withdrawn. No equations corrected. What was retracted was something more important in this context: institutional overconfidence. The FJC implicitly acknowledged that publishing a climate chapter in its Reference Manual carried risks that outweighed any supposed benefits.

In doing so, it reaffirmed a basic but increasingly neglected principle: disputed scientific claims belong in court, subjected to cross-examination, competing experts, and explicit discussion of uncertainty. They do not belong embedded in an “authoritative” guide that lawyers can cite as if it were a neutral referee.

A genuinely skeptical climate guide for judges would be relentlessly unglamorous. It would dwell on uncertainty, model sensitivity, data limitations, and the difference between scenario exploration and causal proof. It would emphasize how little global averages say about local outcomes and how fragile attribution chains can be. Most of all, it would warn judges against mistaking institutional consensus for empirical certainty.

The removed chapter appears to have done the opposite. And when that became impossible to ignore, the FJC stepped back.

That is the encouraging part. Institutions rarely retreat unless pushed, and rarer still is a retreat that restores proper boundaries rather than inventing new justifications. For a brief moment, process triumphed over presumption.

This does not mean the issue is settled. The incentives that produced the chapter remain. Climate litigation remains attractive precisely because it bypasses legislatures and voters. Efforts to “educate” judges toward preferred narratives will continue, likely in subtler forms. Chapters can be rewritten. Language can be softened. Pressure can be reapplied.

But for now, the judiciary’s support arm remembered what it is not supposed to be: a vehicle for resolving contested policy questions by administrative suggestion.

That is not a victory for any particular scientific theory. It is a victory for skepticism in its proper sense—for restraint, for adversarial testing, and for the idea that courts should decide cases based on evidence presented in court, not conclusions embedded in advance.

Sometimes progress looks dramatic. Sometimes it looks like a chapter quietly disappearing from a manual.

This time, the disappearance matters.