Vectors 6: The Body Count
VECTORS: HOW HEALTH MISINFORMATION KILLS
Piece 6: The Body Count
Reader Navigation ⇢ This piece stands apart from the others in the series. There are no profiles here, no ecosystem analysis, no framework to apply. What follows is evidence, compiled and presented as plainly as the authors of the underlying studies presented it. The question this piece answers is the one every preceding piece has been building toward: how many people died because the misinformation worked?
Numbers this large resist comprehension. The mind that can feel the weight of a single death — a neighbor's father, a child's grandmother, a name on a memorial — does not naturally scale that feeling by six figures. This is one reason the body count from COVID-19 vaccine refusal has not landed with the force it deserves. The other reason is that the body count has been contested, muddied, and in some quarters actively suppressed. This piece attempts to cut through both problems: the psychological and the political.
The evidence is not ambiguous.
The Baseline: What the Pandemic Actually Cost
Before isolating the deaths attributable to vaccine refusal specifically, it is necessary to establish the total pandemic mortality picture against which all downstream calculations are made.
Official U.S. COVID-19 death records through the end of 2022 tallied approximately 1.09 million deaths coded to the virus as an underlying or contributing cause.[1] But official counts are, by methodological necessity, undercounts. Death certificates require a certifier to make a judgment call; the pandemic disrupted healthcare access and overwhelmed reporting systems, particularly early on. The more reliable measure — and the one epidemiologists have consistently preferred — is excess mortality: the number of deaths above the baseline that would have been expected in a given period absent the pandemic.
By that measure, the United States recorded an estimated 1.23 million excess deaths through the end of 2022, a figure that exceeds the official COVID tally by more than 11 percent.[2] And even that figure represents only deaths that occurred — not deaths that were prevented by the vaccination campaign, nor deaths that were preventable but were not prevented because vaccine uptake fell short.
The two-year county-level analysis published in PMC offers additional granularity: an estimated 620,872 excess deaths in the pandemic's first year and 538,708 in its second — approximately 1.16 million excess deaths across the first two years alone.[3] The pandemic was not a single event with a single death toll. It was a sustained catastrophe that killed differently in different places and among different populations, for reasons that became increasingly explicable as the vaccination campaign unfolded.
Methodological note ⊢ "Excess deaths" and "official COVID deaths" are not the same figure and should not be used interchangeably. Official counts reflect deaths where COVID-19 appeared on the death certificate. Excess mortality reflects the difference between total observed deaths and total expected deaths based on historical trends. The gap between them includes deaths that may have been caused by COVID but miscertified, deaths from delayed or disrupted care for other conditions, and deaths that resulted indirectly from pandemic conditions. Both figures are valid for different analytical purposes; this piece uses both where appropriate.
What the Vaccines Prevented
To understand the deaths attributable to vaccine refusal, you must first understand what vaccination accomplished. Two major modeling studies produced figures that are by now widely cited, and they bear restating in full.
The Lancet Infectious Diseases published a global modeling study in June 2022, led by Oliver Watson and colleagues, covering the first year of the vaccine campaign — December 8, 2020, through December 8, 2021 — across 185 countries and territories. Their estimate: without vaccination, approximately 31.4 million COVID-related deaths would have occurred during that window globally. The vaccination campaign averted an estimated 19.8 million of them.[4] In the United States specifically, the same period's vaccination campaign is estimated, in a Lancet-published analysis, to have prevented approximately 1.9 million deaths.[5]
The Commonwealth Fund, in collaboration with Yale School of Public Health, extended the analysis through November 2022. Their findings, published in December of that year: the U.S. vaccination program prevented more than 3.2 million additional deaths and 18.5 million additional hospitalizations from December 2020 through November 2022.[6] Without vaccination, the U.S. would have experienced 4.1 times as many deaths as it did. At the height of the Delta wave, absent the vaccine, daily deaths would have exceeded 24,000 — a figure that staggers because it did not happen, and it did not happen because enough people got vaccinated.
That last clause is the pivot point. Enough people.
The Preventable Deaths
The vaccines were available. The vaccines were effective. The vaccines were free. And hundreds of thousands of people refused them, and died.
The most rigorous U.S.-focused estimate of preventable deaths from vaccine refusal comes from a study published in the European Journal of Epidemiology in 2023, co-authored by researchers at Harvard and the CDC.[7] Using CDC data on COVID-associated death rates by vaccination status, the study estimated that at least 232,000 deaths among adults aged 18 and older were preventable had unvaccinated individuals received a primary vaccine series — and that figure covered only the fifteen-month window from May 30, 2021, through September 3, 2022.[8] The study's authors explicitly noted that this was a conservative floor: deaths in the period before and after the study window were excluded, and the laboratory-confirmed dataset they used likely missed some deaths entirely.
Two hundred and thirty-two thousand deaths. In fifteen months. Each one a person who was alive when the vaccine was available and did not receive it.
Scope ◎ The 232,000 figure is the most methodologically defensible estimate for a specific, bounded time window using individual-level mortality data. It is not a ceiling. Earlier Commonwealth Fund and Lancet modeling, working from different study periods and methodologies, produce estimates in the same order of magnitude from different angles. No credible peer-reviewed estimate finds the number of preventable deaths from vaccine refusal to be small.
The Geography of Refusal
The preventable deaths were not distributed evenly. They concentrated where vaccine refusal concentrated. And vaccine refusal concentrated along a political geography that, by the delta wave of 2021, had become one of the clearest predictors of who died.
A study published in JAMA Internal Medicine in 2023, authored by researchers at Yale, examined 538,159 deaths of individuals 25 and older in Florida and Ohio between March 2020 and December 2021, cross-referenced with voter registration data from 2017.[9] In the period before vaccines became widely available, excess death rates for Republican and Democratic voters tracked closely. After May 1, 2021 — when vaccines became eligible for all adults — the gap opened and widened. By the end of the study period, the excess death rate among registered Republican voters was 43 percent higher than the excess death rate among registered Democratic voters.[10] The disparity was largest in counties with the lowest vaccination rates.
The causal mechanism the researchers identified was vaccination uptake. In counties where at least half the population was vaccinated, there was no meaningful gap in excess deaths between Republicans and Democrats. The vaccine, where accepted, protected across party lines. Where it was refused, it could not.
Pew Research Center's county-level analysis of COVID deaths across the full pandemic confirmed the same pattern from a different angle: during the Delta and Omicron waves, death rates in counties with less than 50 percent adult vaccination were roughly twice what they were in counties with 80 percent or more vaccination coverage.[11] Counties that voted for Trump at higher margins had substantially lower vaccination rates and substantially higher death rates during every wave in which vaccines were widely available.[12]
The Lancet's own cross-state analysis, published in 2023, found that states which deployed science-based interventions including vaccination and targeted vaccine mandates achieved COVID death rates comparable to the best-performing high-income nations.[13] States that did not achieved death rates that reflected the choice.
Accountability framing ⚖ The correlation between political identity and vaccine refusal is documented. The correlation between vaccine refusal and preventable death is documented. These two documented correlations do not establish that Republican politicians caused the deaths — but they do establish that the information environment those politicians cultivated, and the vaccine hesitancy they promoted or tolerated, had a calculable body count. The series has profiled the architects of that information environment. This piece supplies the arithmetic.
What the Numbers Mean
This is not a close call. The evidentiary record is not contested at the scientific level, whatever noise persists in the political sphere. COVID vaccines reduced death. People who chose not to receive them died at higher rates than people who did. The shortfall in vaccination coverage — driven by a misinformation ecosystem this series has documented in preceding pieces — produced a preventable death toll in the hundreds of thousands.
The minimum defensible estimate: 232,000 preventable deaths in fifteen months from vaccine refusal alone, among American adults, using conservative methodology on CDC data, cross-validated by researchers at Harvard and the CDC itself.[14]
The broader modeling picture: the vaccination program as actually implemented, against the counterfactual of no vaccination, prevented 3.2 million deaths through November 2022. The gap between what the vaccine accomplished and what it could have accomplished — had uptake been higher, had the information environment been cleaner — is itself a death toll, though one that cannot be assigned a single number.
Synthesis ⊕ The three evidentiary strands converge on a single finding. First: excess mortality data establishes the true scale of pandemic death, beyond what official counts capture. Second: modeling studies establish what the vaccines accomplished — and what additional deaths would have occurred without them. Third: refusal-specific analyses establish the deaths that occurred not because the pandemic could not be stopped, but because the tools to stop it were rejected. Taken together, the evidence supports a conclusion the series has been building toward from the beginning: misinformation about COVID vaccines did not merely spread falsehoods. It killed people. We can count some of them.
Series bridge — The evidence assembled in this piece does not require the ecosystem analysis of Piece 1, the radicalization dynamics of Piece 2, or the profiles of Pieces 3 and 5 to stand on its own. It stands. But for readers who have arrived here through the series, the data lands differently: the 232,000 preventable deaths are not a statistic produced by an abstraction. They are the output of a system whose architects, mechanisms, and institutional failures the preceding pieces have traced in detail. Piece 7 turns to what comes next — and why the stakes of getting it right have not diminished.
Notes
U.S. observed COVID-19 deaths from January 2020 through December 2022: n = 1,094,230. Source: Lindeman M et al., "Excess Death Estimates Compared with State-Reported and Observed COVID-19 Deaths, New Jersey and the United States, 2020–2022," Frontiers in Public Health, December 2024. https://doi.org/10.3389/fpubh.2024.1338579 ↩︎
U.S. NCHS estimated excess deaths, January 2020 through December 2022: n = 1,233,366 — exceeding the observed COVID death count by more than 11 percent. Source: ibid. ↩︎
County-level estimates across 3,127 counties: 620,872 excess deaths in the first pandemic year (March 2020–February 2021) and 538,708 in the second (March 2021–February 2022), totaling approximately 1.16 million excess deaths. Woolf SH et al., "Monthly Excess Mortality Across Counties in the United States During the Covid-19 Pandemic, March 2020 to February 2022," PNAS, May 2022. https://pmc.ncbi.nlm.nih.gov/articles/PMC9094106/ ↩︎
Watson OJ, Barnsley G, Toor J, et al., "Global Impact of the First Year of COVID-19 Vaccination: A Mathematical Modelling Study," The Lancet Infectious Diseases, June 23, 2022. https://doi.org/10.1016/S1473-3099(22)00320-6. Estimated 31.4 million deaths would have occurred without vaccination; 19.8 million averted. ↩︎
"According to a June 2022 study published in The Lancet, COVID-19 vaccination in the United States prevented an additional 1.9 million deaths from December 8, 2020, to December 8, 2021." See Wikipedia, "COVID-19 Pandemic in the United States," citing Watson et al. (2022). ↩︎
Fitzpatrick MC et al., "Two Years of U.S. COVID-19 Vaccines Have Prevented Millions of Hospitalizations and Deaths," Commonwealth Fund / Yale Center for Infectious Disease Modeling and Analysis, December 13, 2022. https://www.commonwealthfund.org/blog/2022/two-years-covid-vaccines-prevented-millions-deaths-hospitalizations ↩︎
Jia KM, Hanage WP, Lipsitch M, Johnson AG, Amin AB, Ali AR, Scobie HM, Swerdlow DL. "Estimated Preventable COVID-19-Associated Deaths Due to Non-Vaccination in the United States," European Journal of Epidemiology, November 2023. https://pubmed.ncbi.nlm.nih.gov/37093505/ ↩︎
The study estimated at least 158,000 preventable deaths in the 30 jurisdictions with complete data, extrapolated to approximately 232,000 (95% CI: 214,000–250,000) for the full U.S. adult population. The authors explicitly note this is an underestimate, as deaths before and after the study window are excluded. Ibid. ↩︎
Wallace J, Schwartz JL, Goldsmith-Pinkham P. "Excess Death Rates for Republican and Democratic Registered Voters in Florida and Ohio During the COVID-19 Pandemic," JAMA Internal Medicine, July 2023. https://pmc.ncbi.nlm.nih.gov/articles/PMC10366951/ ↩︎
Adjusted analysis: excess death rate among Republican voters was 43 percent higher than among Democratic voters after vaccine eligibility opened to all adults (May 1, 2021 through December 2021). See Wallace et al. (2023), ibid.; corroborated by independent NBER working paper (Goldsmith-Pinkham et al., 2022) finding a 76 percent gap in excess death rates in their broader dataset. See also Yale School of Public Health press release, November 3, 2022. https://ysph.yale.edu/news-article/study-finds-large-gap-in-excess-deaths-along-partisan-lines-after-covid-19-vaccines-introduced/ ↩︎
Pew Research Center, "The Changing Political Geography of COVID-19 Over the Last Two Years," March 3, 2022. Counties with less than 50 percent adult vaccination had death rates roughly twice those of counties with 80 percent or higher vaccination coverage during the Omicron wave. https://www.pewresearch.org/politics/2022/03/03/the-changing-political-geography-of-covid-19-over-the-last-two-years/ ↩︎
Ibid. Counties supporting Trump by the highest margins in 2020 had substantially lower vaccination rates and substantially higher death rates during vaccine-era pandemic waves, compared to counties supporting Biden by the largest margins. ↩︎
Bollyky TJ et al., "Assessing COVID-19 Pandemic Policies and Behaviours and Their Economic and Educational Trade-offs Across US States from Jan 1, 2020, to July 31, 2022: An Observational Analysis," The Lancet, March 2023. https://doi.org/10.1016/S0140-6736(23)00461-0 ↩︎
The 232,000 figure is the floor established by individual-level mortality data. Readers should note that it covers only the primary vaccine series among adults 18+ during one 15-month window, and explicitly excludes deaths in the period before or after that window, deaths among the incompletely vaccinated, and any pediatric deaths. The true number of deaths attributable to vaccine refusal across the full pandemic is higher. How much higher is a matter of ongoing research; no peer-reviewed estimate suggests the number is lower. ↩︎