On August 12, Sen. Ted Cruz told his followers on X that he had the receipts.

"Bombshell Data: Fauci Covered up that Covid Vaccine Produced 82% Miscarriage Rate," the post read, promoting an episode of his podcast. "We now have the receipts: Dr. Fauci and Pfizer knew the second dose of the COVID vaccine carried an 82% risk of miscarriage for women in their first trimester."

By Thursday evening, the post was gone. So was the podcast episode. Cruz had deleted both after the claim collected a community note, a wave of criticism from across the political spectrum, and a segment on CNN. By CNN's count, the post reached more than 600,000 people before it disappeared.

The number was not true, nor was it bombshell data. It was a five-year-old error that fact-checkers first took apart in July 2021.

The numbers, side by side

Miscarriage is common. Roughly one in ten recognized pregnancies ends in miscarriage whether or not a vaccine is anywhere in the picture, and that has always been true. So the only question worth asking is whether vaccinated women miscarried more often than everybody else.

They did not. Here are the rates.

  • Women who never got the vaccine: about 10%, the ordinary background rate, per the American College of Obstetricians and Gynecologists
  • Vaccinated women in the study Cruz was citing: 12.6%, out of 827 pregnancies
  • Vaccinated women across 21 pooled studies: 9%, out of 123,185 pregnancies
  • Vaccinated women in CDC's final count: 10.8%, out of 12,097 pregnancies
  • What Cruz told 600,000 people: 82%

One of those numbers is not like the others.

The cleanest test is a study that tracked both groups at the same time and counted them the same way. Boston University researchers followed 1,815 women trying to conceive between December 2020 and November 2022.

  • Women who were not vaccinated: 26.6% miscarried
  • Women vaccinated before conceiving: 23.9% miscarried

Both figures run high because that study signed women up before they got pregnant and caught very early losses most studies never see. That is exactly the point. The yardstick was identical for both groups, and the vaccinated women came out slightly lower.

Now the comparison Cruz did not post. Something does raise the odds of losing a pregnancy, and it is the thing the vaccine was designed to prevent.

  • Miscarriage after a COVID infection: two to three times the odds, across 25,058 pregnancies tracked in Texas
  • Stillbirth when the mother had COVID at delivery: 1.26%
  • Stillbirth when the mother did not: 0.64%
  • Stillbirth during the Delta wave, mothers with COVID: four times the risk

Those last three come from CDC data covering 1.2 million American deliveries.

The vaccine did not move the number. The virus did.

Where 82% comes from

The figure traces back to a real study: preliminary findings on mRNA COVID-19 vaccine safety in pregnant people, published in the New England Journal of Medicine in April 2021 by Shimabukuro and colleagues using CDC surveillance data.

That study tracked 827 people who were vaccinated during pregnancy and whose pregnancies had already concluded by the end of March 2021. Among those 827 completed pregnancies, 104 ended in miscarriage, a rate of 12.6%, which the authors noted was in line with the background rate in the general population.

To get to 82%, you have to do something the study never did. You throw out the 700 participants who were vaccinated in the third trimester, keep all 104 miscarriages in the numerator, and divide by the 127 people who remain.

104 divided by 127 is 81.9%.

Why that calculation is broken

The error is not subtle, and it has a name in epidemiology: you have changed the population you are measuring after seeing the outcomes.

Consider who could possibly appear in that 827-person group. The study had a March 2021 cutoff. A woman vaccinated in her third trimester would deliver within weeks and land in the completed-pregnancy column. A woman vaccinated in her first trimester had roughly six months of pregnancy left, she could only show up in that column by the cutoff if her pregnancy ended early.

So the early-vaccination group was, almost by construction, made up mostly of pregnancies that had ended early. Deleting the third-trimester participants and keeping the losses doesn't isolate a first-trimester risk. It manufactures one.

Dr. Jake Scott, an infectious disease physician at Stanford Medicine, put it plainly: "It's like calculating a school's dropout rate in October by looking only at the students who have already left."

Dr. Kevin Ault, a gynecologist and former member of the CDC's Advisory Committee on Immunization Practices, called the claim "an incoherent story" on CNN. "This is a well-studied area; we knew this in 2021, that there wasn't an increased risk, and now we have five more years of data," he said. "It's hard to believe we would miss a factor of eight increase in miscarriages."

There is a companion statistic, an "81% rate" drawn from Pfizer's post-authorization adverse event report, that circulates alongside the 82%. It is built on the same error, dividing 28 reported losses by the 32 pregnancies with known outcomes while ignoring 238 with no outcome reported. Measured against all 270 pregnancies in that report, the 28 losses come to about 10.4%, right back in the ordinary range. The inflated version also fails on its own terms: 28 divided by 32 is 87.5%, not 81%. The number doesn't even match the calculation it claims to come from.

Who else has counted

Three more studies compared vaccinated and unvaccinated women directly, and none of them found a gap.

In September 2021, researchers examined 105,446 pregnancies across eight U.S. health systems and found that women who miscarried were no more likely to have been vaccinated in the previous 28 days than women whose pregnancies continued.

That October, a Norwegian registry study in the New England Journal of Medicine compared 4,521 first-trimester miscarriages against 13,956 ongoing pregnancies and found the same thing, with the vaccinated group coming in slightly lower.

A 2023 review in Human Reproduction pooled 21 studies covering 123,185 vaccinated pregnancies. Miscarriage rate: 9%. Live birth and ongoing pregnancy rates: identical to the unvaccinated.

The fair objection is that most of this evidence is observational, and women who seek out vaccination differ from women who don't in ways no statistical adjustment fully captures. That is a real limitation. It is not one capable of hiding a sixfold difference. An effect that large would have to stay invisible across eight American health systems, a national Norwegian registry, 21 pooled studies, and five years of follow-up, all at once.

What the texts actually say

Cruz was reacting to messages released August 10 by Sens. Rand Paul and Ron Johnson, pulled from Fauci's government phone and dating to January 2021, one month after the first vaccines received emergency authorization, when the CDC and major medical organizations had not yet issued a recommendation for pregnancy either way.

The exchange begins with then-Surgeon General Vivek Murthy asking Fauci and then-CDC Director Rochelle Walensky whether there was any reason to prefer vaccinating early versus late in pregnancy. Fauci initially said no. Later that day he added: "I asked around a bit more and another issue came up that you need to be aware of. Since many people have significant cytokines storm and fever after the 2nd dose, this theoretically could be associated with miscarriage in the 1st trimester."

That is the line the senators highlighted. Here is the line from the next day that they did not:

"In the vaccine trials, pregnant women were not allowed but following the EUA, a large number (more than 10,000) pregnant women have been vaccinated. No issues have arisen."

Fauci went on to note that animal toxicity studies "did not raise any red flags." His private assessment and his public statements in early 2021, that the data showed no red flags, that pregnant women had not been in the trials, that the risks had to be weighed against the real dangers of COVID during pregnancy, track each other closely.

To generate the appearance of a cover-up, the Paul and Johnson presentation spliced Fauci's cautious February 2021 remarks between comments he made roughly six months later, in August 2021, after the CDC had formally recommended vaccination during pregnancy and after considerably more data had accumulated. A scientist saying "we need more data" in February and "it's very clear it's safe" in August is not a contradiction. It is what it looks like when evidence arrives.

Victoria Male, an associate professor of reproductive immunology at Imperial College London, told FactCheck.org that caution in January 2021 was "sensible," and that "subsequent evidence found no increased risk of pregnancy loss." Her summary of where things stand: "The current data continue to show no increased risk of miscarriage (or any other pregnancy complication) associated with COVID-19 vaccination."

Not just Cruz

The claim moved through the Republican caucus and conservative media in a matter of days. Rep. Lisa McClain asserted on Fox Business on August 11 that "The New England Journal of Medicine said 82% of women who got the vaccine in their first or second trimester had a miscarriage"; the host issued a clarification the next day. Johnson said on X that "we will never know how many babies were miscarried because of Fauci's lies." NIH Director Jay Bhattacharya told Fox News on August 11 that it was "certainly possible" the vaccines caused miscarriages and that the scientific debate "rages on", a characterization FactCheck.org found unsupported.

Notably, when asked to defend the 82% figure, Johnson's communications director did not. "The question isn't what the actual percentage of miscarriage risk is," she said, pivoting instead to a broader argument about safety surveillance and informed consent.

That is a revealing answer. The statistic did the work of getting attention. Once it was challenged, it turned out to be disposable.

The criticism came from the right, too

The sharpest reactions to Cruz's post did not come from his usual opponents.

"Deeply, dangerously dishonest. He knows better," wrote David French, a conservative columnist at the New York Times. Norman Ornstein called it "an utter dishonest and dangerous disgrace." Former Republican Rep. Joe Walsh wrote: "This is a bald faced lie." Brad Polumbo pointed out that the underlying data "was NEVER 'covered up' by Dr. Fauci. It's been public since its 2021 publication in NEJM."

That last point deserves emphasis. The alleged cover-up concerns numbers published in one of the most widely read medical journals in the world, available to anyone with an internet connection for five years. Cruz's claim was that Fauci concealed data that Cruz's own staff found in a public journal article.

Cruz is a former Texas solicitor general who has argued cases before the U.S. Supreme Court. Parsing statistics and reading source documents carefully is, quite literally, the skill set. He deleted the post within a day and appended a correction to his podcast, while maintaining that his broader point about Fauci still holds. The Daily Beast reported that it sought comment from his office.

Should Fauci sue?

It is a natural question, and the honest answer is that he almost certainly would not win.

Fauci is a public figure many times over. Under New York Times v. Sullivan, he would have to prove not merely that Cruz's statement was false, but that Cruz made it knowing it was false or with reckless disregard for whether it was true, by clear and convincing evidence. That is a demanding standard, and it is demanding on purpose.

The Constitution's Speech or Debate Clause would not shield Cruz here. Hutchinson v. Proxmire established in 1979 that a senator's statements in press releases and newsletters fall outside that protection, and a podcast promotion on X is not a legislative act. But Cruz has other defenses: he was characterizing a document, he arguably framed parts of the claim as interpretation, and he retracted within roughly a day, which cuts against recklessness and limits damages. Fauci would also face the practical problem of proving concrete harm to a reputation already subject to years of public attack.

A lawsuit would hand Cruz a platform, a fundraising line, and a years-long discovery fight. That is not a recommendation against accountability. It is a recognition that defamation law is a poor instrument for it, deliberately so, because the same rules that protect Cruz protect every journalist and critic who gets something wrong about a powerful person.

Why it matters anyway

The remedy that is available is the one that already happened, and it is worth noticing that it worked slowly and incompletely. A community note appeared. Fact-checkers responded. A television segment ran. Cruz deleted the post.

But the post was live for about a day and reached over 600,000 people. Corrections do not travel the way accusations do, and the people who saw "82% miscarriage rate" from a sitting U.S. senator largely did not see the retraction. Some fraction of them are now carrying a belief that will inform a real decision, whether to get vaccinated while pregnant, whether to trust a doctor's advice, whether public health guidance can be believed at all.

That belief is false, and the cost of believing it falls on the women who act on it. Vaccinated women miscarried at about 10%. Unvaccinated women miscarried at about 10%. Women who caught COVID while pregnant lost pregnancies at two to three times that rate and delivered stillbirths at twice the rate, four times during Delta. Those are the numbers that bear on a decision, and they got buried under one that never existed.