Skip to content

Here’s how much vaccination improves your odds of staying alive

I've been browsing around some COVID vaccination data and came across a couple that might be of interest. First off, here is how much getting vaccinated protects you:

In the 50-64 age group, for example, the annualized mortality rate for the unvaccinated is currently 270 per 100,000, while the rate for the vaccinated is 10.4 per 100,000. That's a difference of 26x, or 2,500%. Get vaccinated!

Next, here's an oddball chart that shows how many people who have gotten one dose of the vaccine have gone on to become fully vaccinated:

In Minnesota, 92% of the people who have gotten one shot have gone back for a second. In New Hampshire, only 70% have gone back for a second shot.

Just eyeballing this chart, nothing special catches my eye. Is there some thread that connects states with high callback rates? It doesn't seem to be geography. Nor does it seem to have anything to do with red vs. blue states. Nor with average income. Maybe it's just random?

UPDATE: I have corrected the mortality rates for all age groups. I originally used numbers from both the CDC and Our World In Data, but that introduced some inconsistencies. The chart is now based entirely on CDC data from October (the most recent month available).

35 thoughts on “Here’s how much vaccination improves your odds of staying alive

  1. Steve Stein

    " Is there some thread that connects states with high callback rates? It doesn't seem to be geography. Nor does it seem to have anything to do with red vs. blue states. Nor with average income. Maybe it's just random?"

    Maybe it's because there has been no federally coordinated systems in place. Nothing to coordinate testing, vaccination, reporting. Nothing. It was all left to the states under the former guy, and nothing was put in place after the fire was raging. It's all ad-hoc.

  2. LostPorch

    How are they counting J&J, which only had a single dose initially? Possible regional differences in distribution between 1 dose and 2 dose vaccines?

  3. golack

    The % vaccinated data has some issues. You'll see jumps in some states, sometimes a few jumps.
    As far a PA is concerned, the state has a single vaccinated rate of 84% (CDC and CovidActNow numbers), yet the highest rate in PA's counties is only 79% (75% for the largest counties, from CovidActNow).
    The New Hampshire state site has 61.8% with at least 1 dose, 55.6% fully vaccinated, so 89% went back to get the second shot.
    https://www.covid19.nh.gov/

  4. kaleberg

    What accounts for the high vaccine protection for 18-29 year olds? They have a very low COVID death rate, but young people who do get vaccinated seem to have a much, much lower death rate. Is it because only a handful of vaccinated young people have died of COVID or is it because young people who get vaccinated are less likely to die in car crashes or commit suicide?

  5. rational thought

    Kevin was sort of playing games and being deceptive giving that 18-29 number.

    I found the data I think Kevin relied on for the chart . The numbers match exactly with kevin except for that age range . The table shows death rate from covid for week of October 12 ( not current week although that is not really a major issue) . The reported death rate per 100,000 for 18-29 was .19 for unvaccinated and 0.00 for unvaccinated! Yes zero. Now looks like they round to .0x so could be anything less than .005. But I think that would be either zero or one actual deaths . Two should get .01.

    So really the 18-29 age group just has so few deaths in a week that numbers too small to come to any valid conclusion.

    Not sure if Kevin just decided to assume that actual rate was .005 and use that ( you get his number that way) or the actual number was just below .005 AND the .19 was just almost exact. Almost surely the former as the latter would be quite a coincidence. To avoid showing an infinity percentage which would be a red flag something ain't right ( actual comparison for only one week might be infinity).

    What really makes this manipulation troublesome is that the table has another age category or 12-17, which somehow is missing from Kevin's table. Suspiciously. Because the relative rates there were .02 and .05 which gets only 150% effectiveness. Not so good .

    But, in reality, the weekly rates for that young age groups are just too small to make any comparison valid. Both the low 150% for 12-17 that Kevin ignores and the super high 3700% for 18-29 he has should both have been left off as too small to be valid.

    And you do not understand that this is just covid death rates, not all death rates. Do you really think covid vaccination would reduce death rate from everything that much ? Then it would be a semi immortality drug.

    But interesting issue because another cdc study showed reduced death rates ( but not that much) from causes other than covid and by a very significant amount. Nobody really believes that but many are missing that the study is BAD news for vaccines. Because, as no way the vaccine itself prevents heart attacks and car accidents, all it indicates is that those who did not vaccinate are unhealthier and take more risks regardless of covid . They would have had a higher death rate even if they did vaccinate. The unvaccinated smoke more , are fatter, and do not mask or socially distance as much.

    So how much of the relative better death rate from covid for vaccinated is NOT due to the vaccine but just that they are less likely to die of covid regardless of vaccine? Probably a substantial amount and more in younger age groups .

    However that might be offset by another bias that understates the vaccine. Just comparing vaccinated and unvaccinated is not quite fair . A better comparison would be never infected vaccinated and unvacinated or infected vaccinated and unvaccinated. And I guarantee that a higher percentage of unvaccinated have been infected.

    So unvaccinated group is getting more protection from natural immunity than vaccinated group .

    In fact, possible, with omicron, and as the % of unvaccinated never infected shrinks, that the infection rate and even possible but unlikely death rate , will be HIGHER among vaccinated than unvaccinated. And that will be taken as proof the vaccines are not working at all. But it will not be.

    1. Kevin Drum

      I admire your thoroughness!

      For the two youngest age groups I went to the underlying CDC data. For 12-17, the number of deaths was tiny (5 or 6) and it was incomplete. It was obviously both unreliable and meaningless, so I didn't include it.

      I did the same for 18-29, but you're right: I made a mistake. I wrote down the vaxed death rate as 0.005, but it's actually 0.0065. I also left the unvaxed death rate as 0.19 (from the OWID data) but the CDC data is 0.17. This is October data in all cases. I have fixed the chart.

      UPDATE: In fact, I just revised the entire chart using CDC data so that everything is coming from the same place.

      Thanks for pointing this out.

    2. lawnorder

      I would be unsurprised if there is a correlation between death rate and vaccination status. It's a reasonable inference that vaccinated people are more cautious about covid than unvaccinated people. It doesn't seem like much of a stretch to jump from "more cautious about covid" to "more cautious overall", and from "more cautious overall" to "less likely to die of avoidable causes".

      As you say, this weakens the case for vaccines a bit, but not much, since the numbers relied on are for deaths from covid, not deaths from all causes. It's believable that anti-vaxxers may have been a bit more likely to die from covid before vaccines came along, just because their less cautious attitude made comorbidities more likely. However, it's not a matter of "a bit more likely". The numbers say the unvaxxed are A LOT more likely to die of covid than same age vaxxed.

  6. rational thought

    And do not want to sound too harsh with Kevin from my prior post although my criticism for including a likely somewhat faked and meaningless 18-29 number , while leaving off 12-17 which would show something he would prefer to avoid ( although it too is meaningless), is 100% valid.

    But the first chart outside of those age ranges is both correct in numbers ( I checked) and methodologically valid . And it is based on the most current week available ( 10/12) and not cherry picked. Although other weeks would vary significantly up or down, this does not appear to be a picked " best week" for vaccine. And Kevin even gets % right which is often messed up. But not sure why he uses % ages. For example the 1930% means a 20.3 times lower death rate, which people understand better. Many assume 1930% means 19.3 times. Admit it- some of you thought that. Better to just show it as x times.

    Frustrating that kevin , or whoever did the table if Kevin got it somewhere, just had to be manipulating for younger age ranges . Just exclude them both or include them both with explanation that numbers too small to be statistically valid. The good although imprecise numbers for other age groups demonstrates the point well enough.

    Never understand why anyone trying to make a valid point shown in actual statistics also has to skew things to make it appear exaggerated and even bigger. Just undercuts their argument when the distortion is found.

  7. rational thought

    I think most or all of the explanation for differences in first and second shots is that much of that reported difference is just not real.

    There is a significant number of second shots and boosters that are, mistakenly, being recorded as first shots because they are not matched up correctly with the first shot . Maybe some by deliberate deception - I know two people who wanted a Moderna booster when had Pfizer before, and not yet approved, who just " accidentally " asked for it as a first shot with a slight variation of their name, and recorded as first shot for that person. But mostly just due to record keeping errors. We have been pretty good at counting total shots but not as good at keeping track of which number shot .

    Some places some age groups numbers now show over 100% with first shot ..which obviously had to be misreading.

    My guess is most of the discrepancy between states is just differences in how accurate states are in assigning the second and boosters correctly.

    1. azumbrunn

      This is a good point. Is there a connection to the resistance against "vaccine passports"? To make the data unreliable makes it impossible for anyone to prove that they are vaccinated.

      Puts US international travelers at a disadvantage against travelers from Europe or Asia. But Real Americans don't care about international travel.

    1. Jerry O'Brien

      There does seem to be something about the Midwest, and it's interesting because the Midwest is pretty evenly divided between Republican and Democratic influence.

  8. gvahut

    The take-home is that if you're between 50 and 79, you have the most to gain from vaccination, and likely the most to lose (in functional years) if you're not. The numbers in younger people don't mean much mortality-wise when the overall mortality rate is so low. And none of this addresses long Covid considerations, which we're almost completely in the dark when it comes to data. And if you're not vaccinated and boosted by now, you're likely SOL because Omicron is at our doorsteps. By the time they could whip up a new vaccine the surge will be over.

    1. Spadesofgrey

      Omicron is not on our doorstep. It's already here and irrelevant. Look at the SNE states. Cases like last year, hospitalization way down. Omicron is nothing but asymptomatic cases.

      1. megarajusticemachine

        Our very own Tucker Carlson, ladies and gentleman! Even if he doesn't have the charisma, he still has the same grasp of the facts.

  9. Vog46

    For all you past infection imunity folks out herer:
    https://www.bloomberg.com/news/articles/2021-12-17/omicron-gets-around-previous-covid-infection-u-k-study-warns

    A previous Covid-19 recovery provides little shield against infection with the omicron variant, a research team from Imperial College London showed in a large study that underlines the importance of booster shots.

    Having had Covid probably only offers 19% protection against omicron, the study showed on Friday. That was roughly in line with two doses of vaccine, which the team estimated were as much as 20% effective against omicron. Adding a booster dose helped dramatically, blocking an estimated 55% to 80% of symptomatic cases.

    The Imperial College London team analyzed all the PCR test-confirmed Covid cases in England between Nov. 29 and Dec. 11, making it one of the most expansive examinations yet at omicron’s potential to evade the body’s defenses. The results were in line with the picture emerging of the variant’s capacity to elude protection from previous infection or inoculation and spread faster than earlier iterations of the virus.

    There was no evidence of omicron cases being less severe than delta, based on the proportion of people testing positive who had symptoms or went to the hospital, the team said

    “It’s very early days,” said Neil Ferguson, a professor at Imperial, who helped lead the study. Rising hospitalizations in places like Manchester and London will provide clearer evidence, he said. “We should have data within a week to make a more definitive judgment on relative severity.”

    In South Africa, which announced the discovery of the variant on Nov. 25, authorities said on Friday the rate of hospitalizations seems to be lower than during the country’s earlier wave of delta infections.

    Europe is bracing for an omicron-driven fifth wave even as intensive-care units in many areas remain filled with patients sick with the delta variant. Some governments are already imposing new measures to slow the omicron wave and buy time for booster campaigns to gear up.

    Another open question is how long the shield provided by a booster shot will hold against the new variant. In a worst-case scenario, with efficacy dwindling as quickly as it does after the first two doses, boosters might prevent only four-fifths of hospitalizations by two months after they’re given, a second Imperial study showed on Friday. That makes it important to think about a long-term strategy, ideally with shots that could work against multiple variants, said Azra Ghani, a professor at Imperial who led that research.

    The proportion of omicron among all Covid cases was probably doubling every two days up to Dec. 11, the U.K. team said, estimating that every person infected with the variant passed it on to more than three other people.
    *************************************************************************

    This is what we feared all along. A mutation that could evade IMMUNITY protections. The manufacturers deliberately made the decision NOT to tweak their formulation for DELTA (The formulation was made for Alpha and Beta variants). Now we have a different variant.
    You need booster shots IF you have been vaccinated, or remain partially vaccinated then get fully vaccinated. The co-existance of Omicron and DELTA is a dangerous mix.
    Especially for those un-vaxxed and never infected, previously in ected, and partially vaxxed and those that have waited for their boosters. Fully vaxxed AND boostered is far better than anything else. Wait till the next variant comes out. !!!!

  10. Vog46

    I have been amazed at the speed in which Omicron is spreading.
    If you wonder why I constantly harp on our info here in the United States the study I referenced above is the EXACT reason.
    Our political divide has affected just about every facet of our lives, including and especially referring to science and or medicine.
    In the UK with their National Institute of Health care system they can determine with a fairly good degree of accuracy how many people have or haven't been vaccinated because everyone is covered by them. You get a shot? Its recorded. Get sent to the hospital? It's recorded.
    Here in the U.S. there are 50 distinct government agencies that report to the CDC. Some governors have interceded in the reporting. So, when TFG said it would go away AND That the more you test the more cases you find - what happened? That normal 1% to 3% of our population that was anti vax, then got missed in with everyone who wanted to be anti- DEM or, more succintly put anyone who was anti TFG was now placed in doubt.
    So testing was NOT wide spread.
    At THIS POINT we as a nation have no idea who has NOT had COVID because we didn't test enough. We are left with estimates with searching around for a way to calculate just who's left that UNprotected. This was THEIR version of the "long game" - sew doubt into the science or at the very least into the numbers.

    Unfortunately the United States is a minority in this situation but the internet is a wonderful thing. We can now check to see just how dangerous Omicron is to the naive, the post infection crowd and the vaccinated crowd by just looking at overseas reports. That are not as sullied as ours due to our infatuation with TFG and his bizarre policies. Keep in mind I don't blame him for COVID - but his response has ping ponged all over the place only to have him end up in the hospital and eventually admitting he got vaccinated (in his own words)
    At this point his supporters are stuck with either admitting they might have been misled OR they try to find any shred of truth that would make TFG look right.

    So, if you start a sentence with "I think" I don't respond to you. If you claim without citation stats are wrong - you make yourself look like and ass. And if you couple that with anecdotal evidence of one or two people that somehow did this and you are using THAT as proof of your point? Nobody can help you.

    Look overseas for data. Look at the English, the French, the Germans and other countries that put politics aside when it comes to the health and well being of their countrymen. They report numbers because they have them. We don't.

    Our healthcare system has let us down during the pandemic - just in record keeping alone.

    1. golack

      Public health is underfunded. Obama tried to up our game, indeed we maintained or helped maintained, some of the largest emerging diseases networks in the world--partnering with many companies. Trump? His budgets basically tried to kill of the CDC.

      You're right about the US... it is a 50 state mess.

  11. lisagerlich

    The thing about omnicron is that all evidence from South Africa says it spreads faster than delta, but it is less lethal. The chart below from ourworldindata shows that South Africa has a low vaccination rate. Like the UK, Ireland has a very high vaccination rate, with the US coming in a distant third. As far as I have deduced, hospitals in South Africa are not being overwhelmed. Yet, Ireland is so frightened of OMNICROM that they are imposing an 8 PM "hospitality" curfew. Britain is troubled and talking about a two-week "circuit breaker" lockdown. I live in the US and say no more panic, no more lockdowns, and stop with the vaccine and mask mandates. Wear a mask or get vaccinated if you so choose. It has been almost two years since the famous "two weeks to slow the spread." Historically, most plagues run their course in about two years. Public safety measures can do more harm than good. It is all about balance. Many on this blog think Australia is where we should be. I couldn't disagree more.

    1. Vog46

      Lisa -
      The best way to compare Omicron to other variants is to go back to pre vaccine days. Why? Because we are mitigating the seriousness with some form of immunity.
      To do THAT we would need to know who had NOT been vaccinated and who has not had a previous bout of COVID. If THAT person were to get Omicron - how dangerous is it to them?
      South Africa has very sparse stats on this - as do we

  12. Vog46

    Kevin made the point for Joe Bidens large employer vaccine mandate by pointing out how effective the vaccines were for the working aged group from ages 30 through 64

    Companies WANT their employees to be there working. Insurance companies do NOT want to pay out death benefits. Companies do not want to have to hire and train employees as replacements.

    The ONLY question is this - how badly would excessive sick outs be on the company's bottom line should they have a break out of Omicron? Think about this for a second. WE just had a large scale tornado out break that affected millions of Americans - some of whom worked at Amazon and that candle factory. Both companies DENIED the employees requests to leave - under the penalty of termination.
    Now, what are the chances that those specific locations would get hit as hard as they did? It is very remote. But they HAD WARNINGS. Sirens went off 15 minutes prior to the strike. They had been under a tornado warning for at least an hour.
    Now, the employees are suing both companies. If the companies did not have a shelter-in-place plan and moved employees to the safety of a hallway or other less dangerous location - all I can say is I hope they win their lawsuits. If they did all they could to protect their employees then, well, they did what they could.
    The key here is that the companies determined the risk TO their employees and their determination was bad - with death resulting. If YOUR company is so worried about losing 45 minutes of production only to have the tornado turn away leaving your company unscathed then you can take your - Just in Time, Lean Six Sigma, Very Lean Six Sigma attitude and stick it where the sun don't shine because you are cutting it way too close.

    But to Kevins point - if you are cutting it this close then having your employees here makes the company WANT to have the vaccine mandate. It's no cost to them and protects THEIR bottom line(s)

  13. Vog46

    Glad to see the courts upholding Bidens vaccine mandate for now
    I think this is one case where the federal government will have some sort of victory on.
    In spite of the rhetoric companies WANT their employees safe and healthy - especially in this time of tight labor markets. They can ill afford many people being out on sick leave all at once. At that point you might as well close down temporarily. but that will throw another monkey wrench in the supply chains.

    Lets face it - the government pays for the shots! It's a no cost to me argument the company's will make. And in spite of some of the far out politicians I can tell you that most company's WELCOME OSHA rules and regulations. Its a way for THEM to take harsh actions against employees who do stupid stuff, believe it or not.
    Now, as for the states? Well, they can run their own OSHA program, it can be tougher than federal OSHA's programs but not weaker - but make no mistake - the politicians in charge can tweak their programs to emphasize things that wins votes and not makes enemies. I served under 2 commissioners of labor in NC. One DEM and one republican - both had their differing priorities. It does happen.
    Delaying the OSHA vaccine mandate until Jan is probably a good thing, but it should happen. It gives employers another tool in the box to keep their company's running.
    IF this were being paid for by health insurers or workman's comp insurers? Every darned working man and woman would be vaccinated or a card carrying member of the "I can't get vaccinated because if this _________ reason" club. Their main mission is to minimize payouts for any reason.
    In a sad way it's funny that we would shut down schools in a heartbeat over an unsubstantiated Tik-Tok threat of violence at schools - but for something that is proven to cause illness we talk about the kids loss of freedumb - never mind what it does to the teachers and staff at that school.

  14. Vog46

    Break through cases with Omicron will become the norm I'm afraid
    The problem is not with the vaccinated. It's with the un vaccinated. They probably WILL get Omicron and just the sheer numbers of them will lead to increased hospitalizations.
    The fact that Omicron affects the bronchia more so than the lungs I'm afraid that it will be much more contagious because of that. The question is how many different strains of Omicron will we get? We got to 50 with Delta -each showing a significant change to the original delta. Yet Omicron just blew those changes away.
    This is not over.

  15. Vog46

    *************SEVERELY OFF TOPIC***************************

    https://www.rawstory.com/donald-trump-2656067519/
    {snip}
    Former president Donald Trump will soon be indicted for criminal racketeering under New York state law, according to Pulitzer Prize-winning investigative journalist David Cay Johnston.

    Johnston indicated Saturday afternoon that the charges will stem from Manhattan District Attorney Cy Vance's ongoing investigation into whether Trump's company misled lenders or tax authorities about the value of its properties.

    "I anticipate they're going to bring a racketeering charge against Trump," Johnston said. "Certainly Trump's team, when he's indicted, and I'm certain he will be indicted, is going to try to lay the blame on everybody else, and so what the prosecutors want to show that is if (Trump Organization Chief Financial Officer) Allen Weisselberg phonied up documents, it was at the direction of Donald Trump."

    MSNBC host Yasmin Vossoughian then pressed Johnston to confirm that he is certain about an impending indictment.

    "Oh, yeah," Johnston responded. "They would not have done all of this and know how much they know ... if they weren't going to to do this. Yeah, they will indict him. Exactly when? I don't know. I don't expect it will be on a straight tax charge. I think there will be a tax charge, but the key charge will be racketeering."
    {snip}

    I'll believe it when I see it............

  16. ruralhobo

    Misleading title. I'm vaccinated but it does not increase my odds of staying alive from Covid by 2,500% but more like 0.4%: from something like 99.5% to 99.9%.

    1. rational thought

      OK. Semantically you are correct and kevin should have said it reduces the odds of dying instead with those percentages.

      I did notice that too as I pick up on little semantic distinctions given that I have to deal with legal interpretations in my profession.

      But really somewhat petty a criticism and I passed on commenting on that as i doubt anyone took it semantically literally. Plus I am aware that I have been annoying to people when I point out petty semantic mistakes so I try to refrain .

      But personally enjoy that discussion.

      1. ruralhobo

        Let's say it's unbelievable that people wouldn't want their survival chances to increase by 2,500%, but quite understandable that they don't get excited when it's only 0.4%. So I get that it's just semantic, but precisely when you see it's only semantic, as most people indeed do, you may think "it's not all that bad, so why is everyone trying to scare me?"

        Putting myself in the shoes of the vaccine hesitant, which I am not myself, I might also ask: "so what happened to that disappearance of Covid when enough people got vaccinated? It's endemic even in states and countries where the target vaccination levels were reached."

        In short overpromising and over-alarming only increase distrust, and vocabulary does make a difference therein.

Comments are closed.