Robert F. Kennedy Jr. on the Joe Rogan Experience #1999, fact-checked

What the evidence shows: Kennedy asserted a precise, universal multiplier: that mercury is 1,000 times more neurotoxic than lead. No toxicological consensus figure of this kind exists; relative neurotoxic potency between metals depends heavily on chemical form (elemental, inorganic, methylmercury, ethylmercury), dose, exposure route, and developmental timing, and researchers do not express cross-metal neurotoxicity as a single fixed ratio. On the specific mercury compound at issue, thimerosal is metabolized to ethylmercury, which the FDA and peer-reviewed toxicology reviews distinguish from methylmercury (the form found in fish and the most-studied neurotoxic mercury compound): ethylmercury is cleared from the body more quickly and has a different toxicokinetic profile, though a 2013 review in the Journal of Applied Toxicology found some in vitro measures of ethylmercury and methylmercury toxicity to cells were comparable and that combined exposures could theoretically produce enhanced neurotoxic effects, an area it called incompletely studied. The FDA's own review found that thimerosal-containing vaccines exposed some infants to cumulative mercury levels exceeding EPA guidelines for methylmercury, which is why thimerosal was removed as a precaution from childhood vaccines starting in 1999, but the agency states that a robust body of peer-reviewed research has found no established health harm, including neurodevelopmental harm, from thimerosal at vaccine-preservative concentrations. No source identified in this review supports a specific, quantified "1,000 times more neurotoxic than lead" comparison; the claim as stated is an unsupported exaggeration rather than an established toxicological finding, even though legitimate scientific debate exists over ethylmercury's neurotoxic risk relative to methylmercury.

What the evidence shows: Whole-cell DTP vaccine was used in the United States from the 1940s through the 1990s; U.S. health authorities (the Institute of Medicine, per CDC/ACIP) did conclude the whole-cell pertussis component was causally linked to rare acute encephalopathy, and CDC confirms no whole-cell DTP vaccine is currently licensed in the U.S., having been fully replaced by the acellular DTaP vaccine in the 1990s because of safety concerns. However, the transition was a phased shift to a lower-reactogenicity product as it became available, not a recall or ban triggered by a documented 1-in-300 death/brain-damage rate; that specific ratio does not appear in CDC, WHO, or peer-reviewed estimates of serious whole-cell DTP adverse events. CDC's own cited postlicensure surveillance (Japan, 1970-1974) found acute encephalopathy/encephalitis (including deaths) at a rate of about 7.6 cases per 10 million doses under whole-cell DTP, roughly 1 in 1.3 million, several orders of magnitude rarer than 1 in 300. Whole-cell DTP-containing vaccines remain WHO-prequalified and are still used in many low- and middle-income countries, including through Gavi, the Vaccine Alliance (which the Gates Foundation helps fund), because international health bodies continue to judge their benefit-risk profile favorable given pertussis's high fatality risk in unvaccinated infants; this reflects ongoing global vaccine policy, not a product banned for a proven safety reason and then covertly redirected to Africa. The claim mixes an accurate underlying fact (the U.S. did move away from whole-cell DTP) with a fabricated injury rate and an inaccurate characterization of why and how the transition occurred and of current WHO/Gavi-supported immunization programs.

What the evidence shows: Kennedy claimed that not one of the 72 vaccines on the U.S. childhood schedule has ever been tested pre-licensure against a true placebo, comparing vaccinated to unvaccinated children on health outcomes. FDA licensing records and peer-reviewed publications document at least one clear counterexample: the rotavirus vaccine RotaTeq was evaluated pre-licensure in 3 placebo-controlled clinical trials totaling 71,725 infants (36,165 vaccine recipients vs. 35,560 placebo recipients), with parents/guardians contacted after each dose to track intussusception, other serious adverse events, and deaths as outcomes, per RotaTeq's FDA-approved prescribing information. A separate large randomized, double-blind, placebo-controlled trial of another rotavirus vaccine (Ruiz-Palacios et al., New England Journal of Medicine, 2006) enrolled 63,225 infants roughly split between vaccine and placebo arms and measured severe gastroenteritis, hospitalization, and intussusception as health outcomes. Because Kennedy's claim is an absolute ('not one... ever'), even one documented pre-licensure placebo-controlled trial with health outcomes is sufficient to make the claim as stated false, regardless of how the broader vaccine schedule's testing history is characterized.

What the evidence shows: Kennedy has repeated this specific 1,135% figure in multiple interviews, including this one (FactCheck.org quotes this exact exchange with Rogan), attributing it to an early run of Thomas Verstraeten's Vaccine Safety Datalink analysis for the CDC. There is no public record of Verstraeten presenting a relative risk of 11.35 (which converts to roughly 1,035%, not 1,135%); the closest known figure, a relative risk range of 7.62 to 11.35, comes from a 2004 slide presentation by SafeMinds, an anti-vaccine advocacy group, which says it obtained the underlying statistical tables via a Freedom of Information Act request, without releasing the original documents or confidence intervals, and which describes the exposure as thimerosal in the highest category rather than specifically hepatitis B vaccine timing. Verstraeten's own published, peer-reviewed study of thimerosal-containing vaccines (Pediatrics, November 2003), based on more than 124,000 infants across health maintenance organizations, found no significant association between thimerosal exposure and autism in either phase of the analysis, though an early phase found some other associations (tics, language delay) that did not hold up in the second phase. Verstraeten has rejected the narrative that his findings were diluted or covered up. Subsequent larger and more rigorous studies have also found no causal link between thimerosal-containing or hepatitis B vaccines and autism. The specific 1,135% statistic therefore traces to an unpublished, uncorroborated secondary source rather than any validated CDC finding, and it does not match the peer-reviewed results of the study it is attributed to.

What the evidence shows: Salon and Rolling Stone simultaneously published Kennedy's "Deadly Immunity" article in July 2005, and both outlets ran corrections in the following weeks (Kennedy himself describes "four or five corrections" in this same interview) before Salon later retracted the piece entirely. Documented issues included a misstated infant mercury exposure figure (corrected from "187 times" to "40 percent" greater than the EPA daily limit), a false claim that a scientist held a patent on the measles vaccine, and quotations from the 2000 Simpsonwood transcript, including remarks by developmental biologist Robert Brent and CDC official Robert Chen, that were spliced or edited in ways that reversed their original meaning. Journalist Seth Mnookin documented these specific manipulations in a Scientific American account, and Kennedy continued repeating already-corrected figures (the mercury exposure claim) in television interviews just days after the correction ran. The assertion that no mistakes were ever identified is contradicted by this public record of itemized corrections and documented quote manipulations. The specific framing that the retraction was driven by pharmaceutical industry pressure, and its exact timing, are not independently confirmed by the sources reviewed and remain disputed by Kennedy and his allies.

What the evidence shows: Kennedy has repeatedly cited a jump in childhood chronic disease rates (in various speeches giving figures such as 6% to 60%, or a version pegging current rates near 54%) as evidence that the expanded vaccine schedule is harming children, but fact-checkers who traced his numbers found no single, consistent dataset supporting the specific figures or timeline he uses. FactCheck.org reported that experts, including a UCSF children's health policy researcher, called Kennedy's higher estimates unrealistic overestimates, and that no comparable measurement of childhood chronic conditions exists spanning from the early 1960s to today because definitions, diagnostic criteria, and survey methods have changed repeatedly. A likely source of the higher current-day figures is a 2011 study in which 43% of children had at least one of 20 tracked conditions, rising to 54% only under a broad definition that included being overweight, obese, or "at risk" for developmental delay, a methodology incompatible with any 1960s-era estimate. Separately, a peer-reviewed 1986 analysis of National Health Interview Survey data found that activity-limiting chronic conditions in children roughly doubled between 1960 and 1981, from 1.8% to 3.8%, with researchers attributing much of that rise to survey-design changes and population aging rather than a disease epidemic; this data set predates and does not corroborate Kennedy's 1986 (11.8%) or present-day (54%) figures. No epidemiological study establishes a causal link between the 1986 vaccine schedule and rising childhood chronic disease rates, and a separate PolitiFact review found the 1986 National Childhood Vaccine Injury Act did not itself drive a dramatic jump in the number of recommended vaccines. The claim is rated misleading: rising diagnosis rates for some pediatric conditions are real, but Kennedy's specific percentages compare incompatible data sources and the implied vaccine causation is not supported by the evidence.

What the evidence shows: Wi-Fi and cell phones both emit radiofrequency (RF) radiation, a form of non-ionizing radiation with too little energy to damage DNA directly, unlike higher-energy ionizing radiation such as x-rays. The National Cancer Institute (part of NIH) reports that four large epidemiologic studies of cell phone use and cancer, including the Interphone case-control study, the Danish cohort, the Million Women Study, and the COSMOS cohort, found mixed but overall no consistent association between RF exposure and glioma, meningioma, or acoustic neuroma, including among long-term or heavy users. Some individual analyses reported small statistically significant increases in specific tumor types, but researchers considered these findings inconclusive due to recall bias and other methodological limitations. Animal studies conducted by the U.S. National Toxicology Program, along with a similar study from the Ramazzini Institute, found a small number of heart Schwann cell tumors in male rats exposed to high RF doses, but an independent panel (ICNIRP) concluded methodological weaknesses in both studies prevent drawing conclusions about whether RF exposure can cause cancer. The NCI's fact sheet addresses cell phone RF exposure specifically and does not separately assess Wi-Fi; Wi-Fi routers and devices transmit at substantially lower power than cell phones, so evidence directly linking Wi-Fi exposure to cancer risk is even more limited than the inconclusive cell phone research. As of the NCI's most recent review, current evidence does not establish that RF radiation from cell phones causes cancer in humans, undermining the stronger and more specific claim that Wi-Fi radiation causes cancer.

What the evidence shows: Kennedy claimed Wi-Fi radiation opens the blood-brain barrier. A 2026 scoping review in Physics in Medicine & Biology examined 35 experimental and clinical studies on radiofrequency electromagnetic field (RF-EMF) exposure and blood-brain barrier permeability, spanning roughly 900 MHz to higher frequencies used in wireless systems; seventeen studies reported increased permeability or molecular changes while eighteen found no effect, and studies with the strictest temperature controls and exposure measurement tended to find no effect. The review's authors concluded that current evidence does not establish a causal link between RF-EMF exposure and blood-brain barrier disruption, and that human data remain sparse and do not clearly separate heating effects from other mechanisms. The U.S. National Cancer Institute states that radiofrequency radiation from devices such as cell phones, which fall in the same non-ionizing frequency range used by Wi-Fi equipment, has too little energy to damage DNA, and that the only consistently established biological effect in humans is mild localized heating, with no other clearly established dangerous health effects. Wi-Fi transmitters operate at power levels far below those used in most of the reviewed animal and cell-culture studies. As of mid-2026, the claim that Wi-Fi-level radiofrequency exposure opens the human blood-brain barrier is not supported by the weight of current evidence, which remains mixed only at higher, non-Wi-Fi exposure levels and inconclusive in humans.

What the evidence shows: Joe Rogan stated that a recent study found glyphosate in the blood of roughly 90 percent of Americans tested. The most comprehensive biomonitoring data available comes from the CDC's National Health and Nutrition Examination Survey (NHANES) for 2013-2014; peer-reviewed analyses of that dataset report glyphosate was detectable in the urine, not the blood, of about 81 to 82 percent of the U.S. population aged 6 and older (n=2,310). Glyphosate has a short biological half-life, on the order of hours to a few days, and is cleared mainly through urine, which is why biomonitoring programs measure urinary rather than blood concentrations. Detection at these trace levels indicates exposure, not necessarily harm, and the NHANES analyses do not by themselves establish adverse health effects at the levels measured in the general population. Rogan's approximate percentage is in the neighborhood of the nationally representative NHANES figure, but he attributes the finding to blood testing when the underlying biomonitoring measured urine; smaller, non-representative convenience-sample projects have reported somewhat higher percentages, adding to the confusion around the exact figure.

What the evidence shows: Kennedy's claim misstates both the timing/mechanism of the U.S. DDT ban and Monsanto's role as a DDT manufacturer. The ban took effect in June 1972 through an administrative cancellation order issued by EPA Administrator William Ruckelshaus, not through a law Congress passed in 1973. EPA's own economic review of the ban ("DDT: A Review of Scientific and Economic Aspects of the Decision to Ban Its Use as a Pesticide," EPA-540/1-75-022) states that of roughly a dozen U.S. companies making DDT in the early 1950s, the last major producers besides Montrose (Geigy in 1966, Allied Chemical and Olin in 1969, Diamond Shamrock in 1970, Lebanon Chemicals in 1971) had already exited before the ban; by 1971 hearing records describe Montrose Chemical Corporation, not Monsanto, as the "sole remaining manufacturer of the basic DDT chemical," and the report contains no mention of Monsanto as a DDT producer at all. Monsanto's major chemical-industry legacy in this era was tied to other compounds, such as PCBs, of which it was the sole U.S. producer from the 1930s until 1977, rather than DDT. The claim's date (1973), its mechanism (a new law), and its characterization of DDT as Monsanto's flagship product are all inaccurate.

What the evidence shows: Peer-reviewed data show celiac disease prevalence in the United States rose gradually over roughly 50 years, not in a discrete 2006 spike: a Gastroenterology study comparing stored blood from 1948-1954 with samples from 2006-2008 found undiagnosed celiac disease was 4 to 4.5 times more common in the modern cohorts, but the authors explicitly note their cross-sectional data cannot pinpoint when the change occurred and describe a decades-long trend rather than a single-year jump. A separate review similarly documents prevalence climbing from about 0.03% in the 1970s to 0.5%-1.26% by the early 2010s, again indicating a gradual, multi-decade rise predating 2006 rather than an event tied to that year. The specific mechanistic link between glyphosate and celiac disease traces to a 2013 paper by Samsel and Seneff in Interdisciplinary Toxicology, which is a hypothesis and literature-review article proposing biological pathways and correlating glyphosate usage trends with rising hospital discharge rates, not an epidemiological study demonstrating that glyphosate causes celiac disease or that a 2006 spike occurred. No cited source establishes a 2006 'explosion' in celiac disease coincident with glyphosate desiccation, and the literature attributes the multi-decade rise to a mix of environmental and dietary factors rather than a single causal exposure beginning in one year.

What the evidence shows: The scientific record on glyphosate and non-Hodgkin lymphoma (NHL) is mixed, not settled as "caused by," as Kennedy asserts. In 2015, IARC classified glyphosate as "probably carcinogenic to humans" (Group 2A), citing "limited" evidence in humans (based largely on elevated NHL rates among farmers and pesticide applicators), "sufficient" evidence in animals, and strong evidence of genotoxicity. The U.S. EPA reached the opposite conclusion, stating glyphosate "is not likely to be carcinogenic to humans" based on a larger set of animal studies than IARC reviewed, a position shared by regulators in Canada, Australia, the EU, Japan, and New Zealand. The largest prospective U.S. cohort, the NIH-funded Agricultural Health Study (Andreotti et al., 2018, following 54,251 pesticide applicators, 82.8% of whom used glyphosate), found "no association apparent between glyphosate and any solid tumors or lymphoid malignancies overall, including NHL and its subtypes," though it flagged a non-significant trend toward acute myeloid leukemia in the highest-exposure group. Despite this regulatory and scientific split, U.S. juries have repeatedly sided with plaintiffs, and Bayer (which acquired Monsanto) agreed in 2020 to pay more than $10 billion to resolve roughly 125,000 Roundup claims alleging the herbicide caused cancer. Kennedy was the plaintiffs' attorney who won an early landmark Roundup verdict against Monsanto, giving him a direct financial and professional stake in promoting the causation claim.

What the evidence shows: The claim matches Tyrone Hayes et al.'s 2010 PNAS study, which exposed 40 genetically male (ZZ) Xenopus laevis frogs to atrazine at 2.5 parts per billion for three years, a concentration the paper itself notes is below the EPA's human drinking-water standard of 3 ppb (0.003 mg/L) for atrazine, so that part of Kennedy's comparison is accurate. Kennedy's figure of 4 frogs turning into fertile females also matches the paper exactly (4 of 40, or 10%, underwent complete sex reversal to fertile females). His figure of 30 frogs being chemically castrated is directionally correct but understates the paper's actual finding that the remaining 36 of 40 males (90%) were demasculinized (chemically castrated, with reduced testosterone and impaired reproductive function). His description that only 30 of the frogs were double Z super males is also muddled: the study's design used an all-male, genetically ZZ population of 40 frogs, not a subset of 30. More broadly, EPA's drinking-water standard is a human health benchmark based on chronic ingestion, not an aquatic no-effect threshold for amphibian gonadal development; EPA's own Scientific Advisory Panel formally reviewed the amphibian gonadal-development question in 2003 and 2007 precisely because the frog findings were scientifically contested, and EPA has continued to register atrazine. Comparing the ppb figures alone therefore overstates how directly the frog findings translate into a human safety concern.

What the evidence shows: The frog study Kennedy describes is real: Hayes et al. (2010, Proceedings of the National Academy of Sciences) found that 10% of genetically male African clawed frogs (Xenopus laevis) exposed to 2.5 parts per billion (ppb) atrazine throughout larval development became functional females that mated with unexposed males and produced viable, fertile eggs; exposed males also showed reduced testosterone and fertility. That exposure concentration is close to, though below, the U.S. Environmental Protection Agency's legal drinking-water limit for atrazine of 0.003 mg/L (3 ppb). However, the EPA's human health risk assessment, which reviewed dietary and drinking-water exposure, found no risks of concern for the general population, including children, from atrazine at regulated levels; the agency identified elevated risk only for children who crawl or play on recently treated lawns and for workers who mix or apply the herbicide. Extrapolating amphibian endocrine-disruption findings directly to routine childhood exposure risk in humans is not supported by current regulatory risk assessments, even though the underlying frog research and the 10% figure are accurately stated. Amphibian ecological risk from atrazine remains an active, separately tracked regulatory concern distinct from the human dietary risk assessment.

What the evidence shows: The strong sales of 'The Real Anthony Fauci' (2021) are well-documented: publisher marketing and bestseller lists credit it with over a million copies sold within months. The specific claim here, that Barnes & Noble and most independent booksellers pulled or refused to carry the book, is not corroborated by independent, non-Kennedy-affiliated reporting; that allegation traces mainly to Children's Health Defense, an organization Kennedy founded and chairs, rather than neutral third parties. Independent reporting on a later, related Kennedy book found that some individual independent bookstores did decline to stock his work, but the owners quoted cited business reasons, such as waning reader interest in pandemic-themed titles, or disagreement with his content, not a coordinated boycott, and that book remained available online through major retailers including Amazon and Barnes & Noble. No mainstream fact-checking organization or wire service has substantiated a claim of large-scale, coordinated removal of 'The Real Anthony Fauci' from most Barnes & Noble or independent bookstore locations.

What the evidence shows: Kennedy has repeatedly cited a Cleveland Clinic study of its own employees (Shrestha et al., Open Forum Infectious Diseases, 2023) to claim the vaccinated face several times the COVID-19 risk of the unvaccinated; he made an identical claim, including the same 3.5-times figure, on this June 2023 Joe Rogan Experience episode. The study, which measured bivalent booster effectiveness in roughly 51,000 healthcare workers, did find a statistical association between a higher number of prior vaccine doses and higher subsequent infection risk during the Omicron/XBB period, and its authors wrote this unexpected finding needed further study. However, the study was not designed to isolate vaccine dose count as a cause of infection risk, did not control for behavioral or exposure differences between groups, and its lead author, Dr. Nabin Shrestha, told fact-checkers the association does not establish causation and that determining whether more vaccine doses cause greater susceptibility 'wasn't the point of the study.' Multiple fact-checking and expert reviews (FactCheck.org, PolitiFact) concluded the finding likely reflects confounding factors such as immune imprinting, differential exposure, or the health-seeking behavior of frequently vaccinated healthcare workers rather than a direct causal effect of vaccination, and cautioned the healthcare-worker cohort's results should not be generalized to the public. A subsequent Cleveland Clinic follow-up study by the same lead author, using the same employee cohort, found being 'up-to-date' on vaccination was not associated with lower or higher COVID-19 risk during the XBB period (adjusted HR 1.05, 95% CI 0.88-1.25). The claim is considered a misleading interpretation of a real but limited and unreplicated observational finding.

What the evidence shows: Kennedy claimed the 1918 Spanish flu deaths were vaccine-induced rather than caused by influenza. No influenza vaccine existed in 1918; the first licensed flu vaccine was not approved for widespread use until 1945, decades after the pandemic. Genetic material recovered from the preserved tissue of multiple 1918 victims, first sequenced starting in 1997, has been identified as an H1N1 influenza A virus, establishing the pathogen behind the pandemic. A 2008 study by NIH researchers Morens, Taubenberger, and Fauci, examining pathology and bacteriology data from over 8,000 autopsies, found that most 1918 deaths involved secondary bacterial pneumonia, a well-documented complication that occurs after influenza virus damages the respiratory tract and allows bacterial invasion, not evidence that victims lacked influenza. A fact-check of this specific claim (including this Rogan podcast appearance) found no supporting evidence for the vaccine-induced-pandemic theory.

What the evidence shows: NIH awarded EcoHealth Alliance a grant (about $3.7 million over six years) to study bat coronavirus spillover risk, and EcoHealth subcontracted roughly $600,000 to the Wuhan Institute of Virology for collaborative work, including experiments that combined genetic elements of different bat coronaviruses. An October 2021 letter from NIH Principal Deputy Director Lawrence Tabak disclosed that one such experiment produced an unexpected result, mice infected with a modified virus became sicker, which EcoHealth failed to report promptly, a grant violation; the letter also stated the viruses used could not have produced SARS-CoV-2. Whether this work meets the specific U.S. government technical definition of "gain-of-function research of concern" is genuinely disputed: NIH, EcoHealth Alliance and the Wuhan-based lead researcher maintain it does not, because the bat viruses studied were not already known to be highly transmissible or virulent in humans, while critics such as Rutgers microbiologist Richard Ebright argue the experiments unequivocally qualify. In 2024 congressional testimony, Tabak told lawmakers "it depends on your definition of gain-of-function research" when pressed on the question. NIH funding of the EcoHealth-WIV bat coronavirus research is confirmed, but characterizing that research as gain-of-function "of concern" remains an open dispute between government officials and outside virologists, not the settled fact Rogan's statement implies.

What the evidence shows: The 2015 Nobel Prize in Physiology or Medicine recognized the discovery of avermectin (whose derivative ivermectin is used against parasitic roundworm diseases such as river blindness), a discovery made decades before COVID-19 existed; it did not address any antiviral or COVID-19-related application. Separately, a WHO guideline panel reviewed pooled data from 16 randomized controlled trials (2,407 patients) on ivermectin for COVID-19 and found the evidence on whether it reduces mortality, mechanical ventilation, hospitalization, or time to clinical improvement to be of very low certainty, and in 2021 WHO recommended ivermectin be used for COVID-19 only within clinical trials. Kennedy's statement that ivermectin "won the Nobel Prize for efficacy in humans" is accurate as far as parasitic disease treatment goes, but raising it in this COVID-19 discussion implies the prize supports ivermectin as a COVID-19 treatment, which it does not, and which the rigorous trial evidence reviewed by WHO has not established.

What the evidence shows: The 1976 US swine flu immunization program, which vaccinated roughly 45 million people, was suspended in December 1976 primarily because of a sharply elevated incidence of Guillain-Barre syndrome (GBS), a rare paralytic neurological condition, among recipients, not because of a specific death count. Contemporary and later reviews documented several hundred GBS cases linked to the vaccine within weeks of vaccination, alongside the fact the anticipated pandemic never materialized. Fact-checking organizations have found no confirmed, officially verified count of deaths attributable to the vaccine; figures such as 25 or 32 deaths circulate in secondary accounts but are not authoritative, and some reported deaths could not be directly linked to the vaccine. The claim that the program was halted specifically "after 25 deaths reported" mischaracterizes the actual decision driver, a GBS safety signal affecting several hundred people, and cites a death figure that fact-checkers describe as unconfirmed.

What the evidence shows: The underreporting figure Kennedy cites traces to a 2010-2011 Harvard Pilgrim Health Care pilot report (led by Ross Lazarus) for the Agency for Healthcare Research and Quality, which found that fewer than 1% of vaccine adverse events were captured through traditional voluntary VAERS reporting compared with an experimental automated electronic-record screening tool; it is a preliminary, non-peer-reviewed pilot report, not a formal CDC study, and it did not isolate deaths or measure COVID-19 vaccines specifically. CDC and FDA, which co-manage VAERS, state on the system's own site that VAERS is a passive surveillance system anyone can submit a report to, that reports are not verified, and that VAERS "is not designed to determine if a vaccine caused a health problem"; HHS's data guide further notes underreporting varies widely by severity, with serious events far more likely to be reported than minor ones. VAERS does contain roughly the figures Kennedy cites for total reports following COVID-19 vaccination, but CDC, FDA, and independent fact-checkers have repeatedly said raw VAERS counts, including reported deaths, cannot be interpreted as vaccine-caused without further investigation. Fact-checking organizations have found that applying the Lazarus pilot's rough underreporting ratio to VAERS death or serious-injury tallies, as Kennedy and others have done, is not supported by the study's own scope and has been disputed by FDA as lacking evidentiary basis for such a large multiplier. Overall: the underlying data point exists but is applied well beyond what the source study measured and contradicts CDC/FDA's explicit guidance on how VAERS data should be interpreted.

What the evidence shows: Kennedy's claim that almost half of the FDA's budget comes from pharmaceutical companies refers to industry-paid user fees authorized under laws like the Prescription Drug User Fee Act (PDUFA), and the figure is broadly consistent with government data. A 2023 HHS/ASPE issue brief found that in fiscal year 2022, user fees accounted for 46% ($2.9 billion) of FDA's total budget of $6.2 billion, with the share varying by program (66% of the human drugs program budget, 43% for biologics, 35% for medical devices). FDA's own public explainer confirms that user fees supplement congressional appropriations, can legally be spent only on the specified review and safety activities Congress designates, and that approval decisions are made based on science rather than tied to fee payment. So "almost 50%" is a reasonable approximation of the total-agency figure, though the share is considerably higher for the drug-review programs specifically. Whether this funding structure amounts to regulatory "capture" is a matter of ongoing debate among health-policy researchers rather than a settled empirical fact.