On March 20, 2026, a coalition of climate skeptics applied for intervenor status in a U.S. court case targeting the EPA's endangerment finding. Their argument: the scientific basis for the finding is "weaker than previously believed." That claim deserves scrutiny, not dismissal. But scrutiny is exactly what it cannot survive.
Here is my position, stated plainly: scientific consensus is a reliable guide, not a barrier to legitimate dissent. The confusion between those 2 things is doing real damage right now, and it is worth being precise about why.
What Consensus Actually Is
Consensus is not a vote. It is not a committee decision or a show of hands at a conference. Think of it like a river delta: hundreds of independent streams, each carrying different sediment from different terrain, converging on the same broad plain. When climate scientists at NOAA, the IPCC, the U.S. National Climate Assessment, and independent research groups across 68 countries reach the same conclusion through different methods, that convergence is the signal. A single tributary running the wrong direction is not evidence the delta is wrong. It is a question worth investigating.
Gregory Wrightstone, executive director of the skeptic coalition, argues that consensus itself is "antithetical to science." He is half right. Consensus enforced by authority, without replication or methodological transparency, would be antithetical to science. But consensus that emerges from independent replication across decades is the closest thing science has to confirmed knowledge. Those are not the same thing, and treating them as equivalent is the error that matters here.
The 2026 EPA rescission of its endangerment finding drew criticism from mainstream scientific institutions for leaning on selective, contrarian analysis while sidelining the IPCC and the U.S. National Climate Assessment. That is not dissent. That is curation: choosing which evidence counts based on the conclusion you want.
When Dissent Is Legitimate
I want to grant the skeptics one real point: peer review has gatekeeping problems. The City Journal evidence on scientific stagnation is worth taking seriously. Federal R&D spending is more than 30 times what it was in 1956, yet genuinely revolutionary findings are rarer. Many peer-reviewed results fail to replicate. Publish-or-perish incentives reward novelty over rigor. A proposed "renegade scientist" grant program that explicitly funds unconventional, interdisciplinary proposals is a good idea. Peer review can suppress heterodox thinking, and that is a real cost.
But that critique applies to the machinery of science, not to the conclusions that survive it. The replication crisis is an argument for better methodology, not for treating all conclusions as equally uncertain. When a finding has been replicated across independent labs, different datasets, and multiple methodological approaches, the burden of proof on the dissenter is high. Not impossible. High.
Legitimate dissent looks like Barry Marshall drinking H. pylori and then documenting the gastritis with biopsies, not like a coalition filing legal briefs citing studies that contradict the preponderance of evidence. One is a methodological challenge. The other is a political one wearing a lab coat.
The ongoing EPA litigation will test whether courts can distinguish between these 2 things. I am not optimistic. Judges are not trained to evaluate whether a contrarian study's sample size, funding source, or replication record justifies overturning a finding backed by decades of convergent evidence.
So here is what should change: scientific institutions need to publish clearer methodological audits of consensus claims, not to appease skeptics, but to make the evidentiary basis legible to courts and policymakers. Show the replication record. Name the independent datasets. Make the delta visible. Consensus that cannot explain itself is vulnerable to exactly the kind of legal challenge now unfolding, and that vulnerability is self-inflicted.
