1000 Jumbo Jet Crashes

Vaccine refusal has caused more than 350,000 deaths in the US — equal to 1000 jumbo jet crashes killing everyone on board

8 min readOct 11, 2022

Summary

In early April of 2021, people in the United States were showing up at vaccines centers more than 4 million times per day. This included first vaccinations and those finishing their primary series of two shots. By late April, that number had plummeted to 2.5 million per day, by June less than 1 million, and by July less than 1/2 million. The people who wanted vaccines got them quickly, and then demand for the vaccines fell off.

If the early-April pace of vaccinations had continued, every adult in the US would have been fully vaccinated by mid-July 2021. The fact that this did not happen, and the continued low uptake of vaccinations since, has caused a great many unnecessary deaths — at least 350,000 through March 2023 by my conservative estimate.

This article details the data and calculations that lead to this conclusion and presents a map that brings the unnecessary loss of life down to the county level.

(This article was originally published in October 2022 and has been updated with data through March 2023.)

Background / Data

The rollout of the COVID-19 vaccines was a fantastic achievement by thousands of scientists, lab technicians, manufacturing personnel, logistics planners, truck drivers, and many others. It surely ranks as one of the greatest scientific accomplishments of all time. Just 11 months after the SARS-CoV-2 genome was first sequenced, healthcare workers were being vaccinated against it.

In early April 2021, there were 4 million vaccine doses per day administered in the United States, with about 2 million people per day becoming fully vaccinated. By April 15, about 90 million people had completed the primary series of two shots. If this pace of vaccination had continued through July 15, every adult in the US (260 million) would have been vaccinated. Unfortunately, the rate of vaccination fell off dramatically after its peak on April 9.

This data project looked at the question:

If the peak rate of vaccination had continued since early April 2021, and all adults in the United States became fully vaccination by July 15, how many lives would have been saved?

The CDC publishes several datasets and graphs that are useful to answer this question:

Overall vaccination rates, which show the peak and decline cited above.
Mortality rates over time for vaccinated and unvaccinated portions of the population.
Daily COVID-19 deaths over time for all people.

Given a hypothetical ramp-up from April 2021 to full country-wide vaccination on July 15, there are two phases of this data analysis:

The 90 days from approximately April 15 to July 15 when the US could have continued vaccinations at the rate of 2 million completions per day. During this time, not everyone in the US would have been fully vaccinated, but many more would have than actually happened. How many of the people who died during this time, and were not vaccinated, could have realistically received the vaccine and been saved by it?
The period July 15 2021 to March 15 2023, when every adult in the US could have been fully vaccinated, but only some were. How many of the COVID-19 deaths during this time did not have to happen?

These questions were answered with a standard spreadsheet, and then the avoidable-mortality result entered to a Python/pandas program to create the mapping data file.

(A note about the data for children… The analysis of COVID-19 mortality for children under age 18, and how many of them could have been saved by prompt vaccination, is somewhat complicated. The CDC statistics for mortality include a mix of deaths “at any age”, “age 5+ years”, and “age 6+ months”. Overlaying this is the fact that children became eligible for the vaccines and boosters at various ages and various dates throughout the pandemic. But these wrinkles do not significantly affect this analysis because the total mortality for children age 0–17 throughout the COVID-19 pandemic is about 2200 or 0.19% of all deaths. Each of these deaths is tragic, but the relatively small number does not alter the overall conclusions here.)

Results

Data about COVID-19 vaccination rates, effectiveness and deaths was entered into a MacOS Numbers spreadsheet (Numbers, Excel, PDF) with three sheets of calculations.

Initial Vaccine Ramp-Up

The first sheet covers the period April 15 to July 15 in 2021 and examines people who could have been vaccinated but were not.

A key figure is on Line 6 showing that 167 million people were actually vaccinated, even though all 260 million adults could have been. There were 93 million people who had less virus protection than was available to them. To estimate unnecessary deaths for this group, I found:

The average number of people in this group who could have been vaccinated at the start (zero) and people who could have been vaccinated at the end (all). This is 46.5 million.
The unvaccinated and vaccinated mortality rates for this period by averaging the CDC rates on the first/middle/last days. These rates are 3.86 and 0.24.

I assumed that a constant 46.5 million people were unvaccinated throughout the period and found the difference between their unvaccinated and vaccinated mortalities. This is about 20,000 avoidable deaths.

Ongoing Full Vaccination

The second sheet covers the period 7/15/21 to 4/15/23 and looks at how the country would have fared with full, ongoing vaccination.

The key analysis problem for this period is that unvaccinated and vaccinated mortality rates (Columns 2 and 3) varied widely throughout the months. There were waves of high and low deaths due to virus variants that were more or less transmissible and virulent. Maybe a vaccine does not work equally well at all times. It is plausible that the vaccinated death rate was low only during times when a less deadly variant was circulating.

To solve this issue, I sampled the CDC mortality rates approximately every 60 days and weighted each rate by the fraction of sampled deaths in that time slice. The weighting factors are in Column 5. The overall, weighted vaccinated mortality is at the bottom of Column 6, or 0.645 deaths per week per 100k population.

The lower three rows show the actual number of COVID-19 deaths in the period, the estimated number of deaths if all 313 million people (age 5+) had been vaccinated, and the difference of about 338,000 preventable deaths.

The calculation of expected, vaccinated deaths is:

(313,000,000 / 100,000) * 0.645 * (52+35) weeks = ~175,000

Total Unnecessary Deaths

The third sheet combines the two time periods and compares estimated deaths (with full vaccination) to actual deaths.

The result is a shocking number — 358,000 people in the US died unnecessarily in 20 months. This is equivalent to 1000 jumbo jet crashes that kill everyone on board.

This count of avoidable deaths is conservative and almost certainly low. I included no adjustment for reduced virus spread from a highly vaccinated population — herd immunity. While the exact amount of herd immunity and its threshold is debated by researchers, it would certainly have offered additional protection to the country.

Since the spring of 2021, some people legitimately have had trouble reaching a vaccine center because of poverty, rural isolation, or disability. In some ways, the federal and state governments could have done a better job with the vaccine rollout. But the overwhelming reason for the tragic loss of so many people was (and still is) vaccine refusal — poor acceptance and understanding of science and an inability to read basic information about the vaccines. Are the vaccines perfect? No. Are they completely without side-effects or risk? Nothing is. But the immediate and guaranteed benefits of vaccination far outweigh the slight risks, and this has been obvious from many news sources.

Present Day

In March 2023, there are about 260 COVID-19 deaths per day in the US, yet only 17% of the population is fully vaccinated with all the boosters recommended for their age. Based on the latest vaccinated mortality, about 170 of these daily deaths could be avoided. These people are dying for no good reason, and the loss of life is equivalent to a new 9–11 attack every 18 days.

Mapping the Result

To show lost lives at an immediate, local level, I created a map by US county. The map is interactive, allowing the reader to hover over any county and see actual deaths along with an estimate of preventable fatalities. The estimates apply the overall US-wide ratio of unneeded deaths (assuming full vaccination) to total deaths.

To create the map, I used the excellent dataset from CovidActNow and wrote some Python/pandas code to process it. The program:

Finds the start/end deaths for each county for this time period and subtracts to find deaths during the period.
Multiplies by the fraction (from the spreadsheet) of those deaths that likely would have been prevented by full population vaccination.
Emits a TSV data file that serves as input to a Flourish choropleth map.

The number of avoidable deaths for each county is an estimate based on overall US data. Some county figures shown will be higher than the true value and some will be lower. But bear in mind that my calculations are conservative. It is likely that more people than estimated died unnecessarily in each county.