"Follow the Science"

Both sides claim science is on their side. Both sides abandon it the moment it isn't.

"Follow the science." Three words meant to end an argument. The trouble is that the people who say them loudest follow the science only as far as it agrees with them, and the instant it turns, they discover that the science is uncertain, the institutions are flawed, and more debate is suddenly, urgently needed. Both sides do this. Today we catch both of them at it, in their own words.

Watch the phrase itself do its work. "Follow the science" quietly smuggles in three claims: that there is a single thing called "the science," that it speaks with one voice, and that it points to exactly one policy. None of that is how science works. Science is a method for being less wrong over time: it argues with itself, revises, and lives in uncertainty. The slogan converts that messy, self-correcting process into a finished verdict you are not allowed to question.

Notice also what the phrase is really doing in a political fight: it reframes a values disagreement as a knowledge disagreement. Whether to close schools is not a question science can answer: it is a question about how to weigh a child's education against an elderly person's risk. But if you call your answer "the science," then anyone who weighs those values differently is not just disagreeing with you. They are "denying science." The framing turns an opponent into a heretic.

Here is that case at full strength. Through the pandemic, contested judgment calls were sold to the public as settled science, and dissent was treated not as disagreement but as a threat to be crushed. The clearest receipts are in the institutions' own emails.

On October 4, 2020, three epidemiologists, Martin Kulldorff of Harvard, Jay Bhattacharya of Stanford, and Sunetra Gupta of Oxford, published the Great Barrington Declaration, arguing for focused protection of the most vulnerable rather than blanket lockdowns. Four days later, on October 8, NIH Director Francis Collins emailed top infectious-disease official Anthony Fauci. The signatories, Collins wrote, were "fringe epidemiologists," and "there needs to be a quick and devastating takedown of its premises." Not a debate. A takedown. The emails were later released through FOIA and obtained by the House Select Subcommittee on the Coronavirus Crisis.

Then the lab-leak question. The possibility that COVID-19 escaped a lab in Wuhan was, in 2020, widely dismissed across mainstream outlets as a debunked conspiracy, not merely doubted but mocked. A May 2020 Washington Post fact-check ran under the headline "Was the new coronavirus accidentally released from a Wuhan lab? It's doubtful." Five years later, in an August 2025 interview on Ira Stoll's "The Editors" Substack, the fact-checker who wrote it, Glenn Kessler, admitted he had been "completely wrong" and expressed his "infinite regret." When Stoll observed that "people only remember the headline," Kessler agreed: "That's on me."

By 2023, the same theory once branded a conspiracy had been formally reassessed by parts of the U.S. intelligence community. The Department of Energy concluded with "low confidence" that COVID-19 "likely" originated from a lab; FBI Director Christopher Wray, in a February 2023 interview, confirmed his agency had "moderate confidence" the virus came from "a potential lab incident in Wuhan." Other intelligence agencies still leaned toward natural origin with low confidence: the community remains split. The lab-leak theory is not proven. That is not the point. The point is that a genuinely open question was declared closed, and the people asking it were smeared, which is not following the science. It is using "the science" as a club.

Now the other side, with equal force, because the same skepticism that was right about the lab leak went lethally wrong elsewhere.

"Manufactured doubt" is a documented strategy with a body count. For decades, tobacco companies funded the message that the science linking smoking to cancer "wasn't settled." They never needed to win the argument, only to make it look unresolved, so action stalled. Reflexive distrust of consensus is not always brave truth-telling. Sometimes it is the trap built for you.

The pandemic grew its own version, and the cost was not theoretical. As trust cracked, prime-time hosts and senators promoted unproven COVID treatments, hydroxychloroquine, then ivermectin, a drug mostly used to deworm livestock, against the explicit warnings of the FDA, CDC, WHO, and ivermectin's own manufacturer, Merck. A high-profile pro-ivermectin study was later retracted. The fixation had a cost beyond the drugs themselves: a 2020 analysis by Stat News estimated that at the pandemic's critical early stage, the hydroxychloroquine craze pulled roughly a third of all clinical-trial volunteers into studies of a drug that did not work, starving research into treatments that might have worked.

And it killed people. A 2023 study published in JAMA Health Forum (Perlis et al.), drawing on a 50-state survey of 13,438 American adults who had been infected with COVID, found that about 6% reported using ivermectin or hydroxychloroquine, and that the strongest predictors of doing so were endorsing COVID vaccine misinformation, distrust of physicians and scientists, and belief in conspiracy theories. Vaccine refusal, in turn, was concentrated along political lines, and multiple analyses have correlated it with measurably higher COVID death rates in the communities most steeped in that distrust. "Just asking questions" is sometimes genuine inquiry. Sometimes it is a costume worn by a conclusion that already decided to reject whatever the experts say, and the bill comes due in lives.

Equal scrutiny, not forced equivalence. So we name what each side refuses to see, and we do not pretend the two failures are identical, only that both are real.

What they won't say, the institutional side. It earned the distrust. When you put "quick and devastating takedown" in an email about Harvard and Stanford scientists, brand an open question a conspiracy, and are later forced to confess in plain English that you were "completely wrong," with "infinite regret," you do not get to act shocked that millions stopped believing you. The lab-leak fiasco was not a defeat for misinformation. It was a gift to it, wrapped and handed over by the very people who claimed to guard the truth.

What they won't say, the skeptic side. Being right about the lab leak does not make you right about everything, and the overreach has a morgue attached. The same instinct that correctly doubted a too-confident consensus also promoted horse dewormer as a cure and floated the claim that COVID was a deliberately engineered bioweapon, the least supported theory of all, since no government builds a plague to release on its own people with no antidote in hand. "The experts were wrong once" is not a method. It is a permission slip to believe whatever you already wanted to.

What neither side will say. "The science" almost never settles a policy by itself. Science can tell you how a virus spreads. It cannot tell you how much childhood to trade for how much safety, or how many livelihoods a lockdown is worth. That trade is a value judgment, yours, your neighbor's, your government's, and dressing it in a lab coat does not make it a fact. Both sides hide their values behind the word "science," then convict the other of denying it.

So here is the test. When a new study lands tomorrow, watch yourself. If it confirms what you already believed, will you wave it through as "the science"? And if it contradicts you, will you suddenly remember that studies can be flawed, institutions can be captured, and more research is needed?

That double standard is the whole game. The honest position is not "trust the experts" and it is not "do your own research." It is a single standard of evidence, applied the same way whether the answer flatters you or wounds you. Almost nobody manages it. The question we hand you is whether you will even try. You decide. We never will.

"Follow the science" · "fringe" (used to dismiss credentialed dissenters) · "takedown" · "conspiracy theory" · "debunked" · "doubtful" (the fact-check headline) · "misinformation" · "denier" · "do your own research" · "gain-of-function" · "zoonotic / natural origin" · "bioweapon" · "horse dewormer" (used to ridicule ivermectin users).

Figures and quotes can be checked against the original releases: The Great Barrington Declaration (published October 4, 2020; Martin Kulldorff of Harvard, Jay Bhattacharya of Stanford, Sunetra Gupta of Oxford) · Francis Collins to Anthony Fauci, October 8, 2020 FOIA-released email ("fringe epidemiologists" and "quick and devastating takedown"; obtained by the House Select Subcommittee on the Coronavirus Crisis) · Washington Post, "Was the new coronavirus accidentally released from a Wuhan lab? It's doubtful," May 2020 (the original fact-check by Glenn Kessler) · Ira Stoll, "The Editors," theeditors.com, August 14, 2025 (the Kessler interview, including "completely wrong," "infinite regret," and the "people only remember the headline" exchange) · NBC News, Department of Energy classified assessment coverage, February 2023 ("low confidence" the virus "likely" originated from a lab) · NPR, FBI Director Christopher Wray on COVID origins, February 28, 2023 ("moderate confidence" of "a potential lab incident in Wuhan") · Roy H. Perlis et al., "Misinformation, Trust, and Use of Ivermectin and Hydroxychloroquine for COVID-19," JAMA Health Forum, September 2023 (n=13,438 U.S. adults; about 6% used either drug).