On February 12, 2026, the EPA issued a final rule rescinding its 2009 determination that six greenhouse gases endanger public health and welfare. President Trump called it "the largest deregulatory action in American history." EPA Administrator Lee Zeldin, speaking at a Heartland Institute conference in April, urged climate skeptics to "celebrate vindication." Vindication of what, exactly? The atmospheric CO2 concentration did not drop. The physics of radiative forcing did not change. The 16 years of peer-reviewed literature linking greenhouse gas accumulation to extreme heat, respiratory illness, flooding, and wildfire did not retract itself.
Here is the distinction that matters, and it is a methodological one: the 2009 finding was a scientific determination, built on evidence that had been accumulating since at least the early 1990s. The Clean Air Act required EPA to make that determination before it could regulate. Once made, the finding became the legal foundation for vehicle emission standards, power plant rules, and the entire architecture of federal climate policy. Rescinding the finding does not touch the evidence. It pulls the keystone from the arch. The stones are still there.
What the Science Actually Says, and How We Know It
Think of the endangerment finding as a formal diagnosis. A doctor who diagnoses you with hypertension is not inventing the condition; she is translating a body of measurement into a clinical determination that triggers treatment. If a hospital administrator later voids that diagnosis for billing reasons, your blood pressure does not normalize. The EPA's rescission is the administrator voiding the paperwork.
Brian Lynk, senior attorney at the Environmental Law & Policy Center, put it plainly in the April 8 lawsuit filed in U.S. District Court: the finding "rests on nearly two decades of conclusive scientific evidence and long-settled law." That evidence includes replicated temperature records, attribution studies linking specific emission sources to measurable warming, and epidemiological data connecting heat events to excess mortality. This is not a single study. It is not a preprint. It is a body of knowledge built through the slow, grinding process of independent replication across dozens of research groups on multiple continents.
The EPA's own justification for the rescission is telling. The agency did not argue the science was wrong. It argued the causal chain between U.S. vehicle emissions and domestic public health harms was "too attenuated" under Section 202(a) of the Clean Air Act. That is a statutory interpretation argument, not a scientific one. Zeldin's "vindication" framing is therefore a category error, and a consequential one, because it lets a legal maneuver masquerade as empirical revision.
Why the Confusion Is Not Accidental
I will grant the administration one fair point: the $1.3 trillion in projected regulatory costs cited by supporters is a real number representing real economic pressure on industries and consumers. Regulatory cost is a legitimate policy consideration. But cost-benefit analysis is not the same as scientific falsification, and treating them as equivalent is how you end up with a cabinet official telling skeptics their doubts have been vindicated by a rulemaking document.
Transportation is still the nation's leading source of carbon emissions. The floods in Minnesota and the rising insurance costs along the Chesapeake are still happening. The lawsuits filed April 8 by environmental groups and Alaskan tribes will test whether the rescission survives judicial review, and that outcome matters enormously for what comes next. But whatever courts decide, the scientific record is not on trial. It never was.
Scientists, science communicators, and science journalists need to hold that line with precision: the finding is gone, the evidence is not. Those are different sentences, and conflating them is exactly what the "vindication" framing is designed to encourage.
