The Course of the Pandemic: What’s the story with Excess Deaths?

This post is by Phil Price, not Andrew.

A commenter who goes by “Anoneuoid” has pointed out that ‘excess deaths’ in the U.S. have been about as high in the past year as they were in the year before that. If vaccines work, shouldn’t excess deaths decrease?

Well, maybe not. Anoneuoid seems to think vaccines offer protection against COVID but increase the risk of deaths from other causes. Or something. I don’t much care about Anon’s belief system, but I do think it’s interesting to take a look at excess deaths. So let’s do that.

I went to https://stats.oecd.org and searched for ‘excess’ in the search field, which led me to a downloadable table of ‘excess deaths by week’ for OECD countries. “Excess deaths” means the number of deaths above a baseline (which I believe is the average over the previous ten years or something, perhaps adjusted for population; I don’t know the exact definition used for these data). “Excess deaths” over the past couple of years have been dominated by COVID deaths but that’s not the only effect: at least in the first year of the pandemic people were avoiding doctors and hospitals and thus missing out on being diagnosed or treated for cancer and heart disease and so on, suicides and car accident numbers have changed, and so on.

Below is a plot of excess deaths, by week since the beginning of 2020, in nine OECD countries that I selected somewhat haphazardly. You can download the data yourself and make more plots if you like.

“Excess Deaths” by week, as a percent of baseline deaths, in nine OECD countries, including the US.

If you had asked me a year or so ago, “what do you think will happen with US COVID deaths now that we have vaccines” I probably would have guessed something like what has happened in Italy or the UK or Belgium or France: there would be some ups and downs, but at substantially decreased magnitude. Instead, the US really stands out as being the only country that had high excess mortality prior to the vaccines and also has high mortality now.

But then, I also expected that just about everyone in the US would get vaccinated, which isn’t even close to being the case (about 20% of US residents haven’t gotten any COVID vaccination, and about 30% are not ‘fully vaccinated’…a term that is a bit misleading, perhaps, as the effects of the vaccines wears off for those of us who got our booster several months ago).

Also, there are competing factors — competing in the sense that some tend to make excess deaths increase while some make it decrease. Vaccines provide substantial protection, and doctors have gotten better at treating COVID, so those tend to lead to lower COVID deaths. But most people seem to have resumed normal life without many COVID precautions, presumably leading to higher infection rates than there would otherwise be. And of course there are still traffic accidents and suicides and drug overdoses and so on that could either increase or decrease compared to baseline.

I find the figure above really interesting. Here are a few things that stand out to me, in no particular order:

  • Denmark had no excess mortality through early 2021! That’s remarkable, they saved a lot of lives compared to the other countries.
  • Canada looks like the US in temporal pattern, which kinda makes sense, but with mortality at about half the US level.
  • I knew Italy got hit very hard early on, northern Italy especially, but hadn’t realized Belgium had it so bad. Jeez they had a terrible first year.
  • The time series in the U.S. is much smoother than in the other countries. Belgium, France, Sweden, the UK, Italy…they all had a big initial spike and then dropped all the way back to 0 excess deaths for a few weeks before the next spike. The U.S. went up and never came all the way back down, even briefly, until a few weeks ago. The U.S. has a much larger population and much larger geographic area than any single European country; perhaps the data on some small part of the U.S., like just New England or just Florida, it would look more like one of these other countries.
  • If the U.S. had matched the average excess mortality of the rest of the OECD countries, hundreds of thousands of Americans would be alive who are now dead.

I guess I’ll leave it to commenters to provide insights on all of this. Go to it!

This post is by Phil.

86 thoughts on “The Course of the Pandemic: What’s the story with Excess Deaths?

  1. There is an idea out there that mRNA vaccines cause lots of excess deaths because they cause the body to create a lot of the “spikes”, which are in themselves dangerous. If a much higher fraction of the vaccines in the US were mRNA types compared with the other countries, these data could support that idea.

    I don’t know if that’s the case, but it just goes to show how complex it is to draw good conclusions in this area.

      • There’s also an idea out there that all-cause mortality was higher in the mRNA vaccine group than it was in the placebo control group during Pfizer’s clinical trials.

        Do you know why that particular idea is out there?

        Because it’s true.

        It’s in their data.

        I got their vaccine. I’m still skeptical there’s any cause for concern.

        But I’m not so certain as to suggest that these concerns are in the realm of flat-earth theory. Not yet.

    • Seems like this would be easy to detect at the population level if it were true. I don’t mean to go full “economist sees a bill on the sidewalk” here but… I find it very hard to believe there would be no convincing epi evidence of this after this much time if it were really happening.

    • That’s a weird one What’s the evidence?

      Seems highly unlikely for several reasons:

      1. the mRNA doesn’t result in the formation of lots of spike protein (“spikes”) in the body. The spike protein produced in cells that have taken up the mRNA isn’t secreted from cells – it lacks a signal sequence (required for secretion). Instead it’s displayed on the cell surface and is recognised as a foreign body that stimulates an immune response. Some proportion is also displayed as fragmented peptide antigens on the cell surface. The mRNA is actually modified to produce a slightly altered spike protein sequence that is locked in a pre-fusion conformation which is nicely immunogenic.

      2. There’s a pretty good correlation between protection from Covid-related deaths and vaccine take up – for example, a recent study shows this quite clearly by analysing US county-wide vaccination rates and Covid mortality ( https://www.bmj.com/content/377/bmj-2021-069317 ). Not surprisingly where vaccination take up is high risk of Covid-related mortality is lower and vice versa. Since most of the vaccinated individuals received Moderna or Pfizer vaccines which are mRNA spike protein vaccines, that rather counters the idea that you raised. I’m pretty sure the Jannsen vaccine which isn’t mRNA also ultimately makes spike protein which is presented at the cell surface in same was as as that from mRNA vaccines.

      3. The correlation between vaccination levels and mortality is also found for excess deaths (i.e. not specifically Covid-related deaths). So in countries where vaccination levels are high, excess death levels are much lower post vaccination (e.g. from around late Spring-early summer 2021 when most people in developed countries could have been jabbed if they’d chosen to do so). You can assess this by combining data from these two sites (one for vaccination rates and the other for excess deaths (whih I think may be the same as Phil’s one in the top post:

      ( https://ourworldindata.org/excess-mortality-covid for excess deaths and https://vaccinetracker.ecdc.europa.eu/public/extensions/COVID-19/vaccine-tracker.html#uptake-tab for vacination rates)

      Again the basic fact that in places where a very high proportion of the population took up the vaccine the excess death levels dropped, combined with the fact that pretty much all vaccines (not sure about the Russian/Indian ones etc but we could easily find out), are made from genes (RNA or DNA) that produce cell-surface presented Spike protein indicates an interpretation that is diametrically opposite to the one you propose.

      Would be interesting to know what the evidence for your idea is

      • Wow seems like everyone’s a little touchy on the issue of mRNA vaccines. I have no idea of what a potential mechanism would be (probably not creating a bunch of spikes), but recent follow up analyses of the Pfizer/Moderna trials suggest either a null effect on all cause mortality in the case of study number 1 or a risk of adverse events exceeding protection from hospitalizations for study number 2. Obviously all of this is far from certain especially without access to IPD data and no follow up studies I know of. But doesn’t seem like a completely outlandish idea to question the benefit of mRNA vaccines especially for populations that are at very low risk of bad COVID outcomes (like children).

        Study 1. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4072489
        Study 2. https://www.sciencedirect.com/science/article/pii/S0264410X22010283

      • “The spike protein produced in cells that have taken up the mRNA…[is] displayed on the cell surface and is recognised as a foreign body that stimulates an immune response”

        You’ve haven’t specified the nature of an immune response. The main immune response is the production of lymphocytes that attack and destroy infected cells.

        Those lymphocytes are the beneficial aspect of mRNA vaccines. The destructive part is that the lymphocytes may kill and destroy those mRNA receiving cells with the spike protein on their surfaces. If those cells are near the injection site, they are most like fat and muscle cells. If those cells are liver, heart, nerves or other unique and vital cells, seems like there could be some life threatening unintended effects?

    • > There is an idea out there …

      Sounds a lot like “Some people are saying….” that was such a familiar rhetoric from a certain unnamed person:

      I think much of the “idea out there” comes from stuff like this:

      https://alexberenson.substack.com/p/another-week-with-deaths-far-above

      Plausible deniability heaped upon selective use of data w/o any real analysis. It would be a mistake, IMO, to underestimate the reach of Berenson and people who are similarly saying that the vaccines were ineffective at best but most likely extremely harmful. Just take a look at the comments.

      Brett Weinstein, Joe Rogan, there’s a long list and they have a lot, lot, lot of followers.

      • Yeah, I agree, this sort of thing is damaging. But it’s hard to know what advice to give. There’s rarely complete scientific consensus about anything, so unless you are an expert yourself you are going to have to choose who to believe. You can go with the weight of authority — which I think is the right default, actually — but that will sometimes lead you astray.

        On the other hand, if the choice is between trusting the most prestigious authority and choosing at random from among the gadflies, you’re surely better off going with the prestigious authority. That’s what I really don’t get about the people who are skeptical of whatever the leading experts are saying: sure, OK, you don’t trust the experts, that’s fine, I agree they get stuff wrong sometimes…but instead you’re just picking some random dude who is spewing a bunch of BS on some podcast or something? It’s very frustrating, and damaging.

        • Phil –

          > That’s what I really don’t get about the people who are skeptical of whatever the leading experts are saying: sure, OK, you don’t trust the experts, that’s fine, I agree they get stuff wrong sometimes…but instead you’re just picking some random dude who is spewing a bunch of BS on some podcast or something? It’s very frustrating, and damaging.

          I observed this up close in the climate change domain for years. The dynamic is pretty standard and seems to generalize across domains.

          There are “experts” who fall out on different sides of an issue like vaccines. For most issues, there’s not really enough of a controversy for the opposing “experts” to make much of a difference. For the most part, most people just go with the flow. Sure, there are people who believe the “experts” who say the world is flat, that aliens really do abduct people (have you watched Tucker lately?), etc. But all-in-all a significant plurality go with the apparent consensus without much turbulence.

          But some issues get polarized. Beliefs become a part of identity. Then people filter “expert” opinion so as to confirm identity orientatation – even if the “expert” opinion is on minority side of the consensus – in an alignment much more like 50/50, reflecting the overall 50/50 split in the country. It’s largely an American phenomenon, although it does exist with some issues in other countries to some extent, even if it doesn’t usually fall out at 50/50 (say, views on climate change in the UK or Australia).

          So they’re not just picking some random dude. They’re picking and trusting an “expert” who aligns with their ideological orientation.

        • Joshua,
          Yes, confirmation bias and socio-political identification are surely part of what is going on. But that’s not all there is to it. Some of my friends have bought into various forms of claptrap, and although they do choose claptrap that is, at least, not in conflict with their biases, they also don’t seem to choose whatever is most closely aligned with their biases. I think a lot of people judge credibility based on style of delivery, with different people leaning towards different styles and thus susceptible to different types of hucksterism. Some people respond to calm, rational-seeming arguments, while others respond to confident expressions of outrage. Or maybe there’s just a big random component that is hard to pin on any particular thing. My acquaintances who became 9/11 ‘truthers’ didn’t seem particular susceptible on the basis of political beliefs, and aren’t the ones I would have chosen if asked which of my friends were most susceptible.

          I assume social scientists have tried to look into what sorts of people fall for which sorts of bullshit. I wonder if any of that work is any good.

        • How skeptical one should be depends on what the skepticism is about, no? Skepticism should never be based on what authority is or isn’t saying. It should be based on the quality and soundness of the available information and the rationality of the interpretations or claims derived from it.

          Just the same, I’d say at the moment the record of “Prestigious Authority” is in decline and the forecast is for more of the same. If it takes prestige to make something stick, that something is already in trouble.

        • Phil –

          I agree with your comment there; not all of this is explained by ideological biases in the sense of political ideology. Certainly, there is a relatively small number of people who hold beliefs that are not in alignment with the majority of people who share their political orientation. If beliefs on these types of issues were merely a function of (political) ideological orientation, that wouldn’t happen.

          I think that its like there are two overlapping circles as in a Venn diagram, where one circle would represent group identity orientation, and another would represent orientation (or bias) at a more personal level. At the personal level, most people want to be “right” and to discover “truth” and to be even-handed and un-biased. A such, we all weight things like style, as an indicator of veracity or thoroughness – and different people attach different weights to those attributes. The weight of those personal-level aspects, relative to the weight of things like group ideological orientation, will vary by individual. Probably pretty much everyone is driven to some extent by things like the “repetitional risk” of holding beliefs disfavored by one’s broader ideological group. But that’s going to vary. And we all belong to different groups. For some people, the group of being a “skeptic” or “contrarian” carries more weight than other group orientations.

          I know you don’t particularly like listening to podcasts, and I’ve only listened a bit to this one and thus can’t really vouch for the quality – but here is an academic who discusses much of this, in particular how the most influential “group” biasing effect can vary. At the link there is a breakdown by topics along the timeline, which could make listening more efficient:

          https://rationalreminder.ca/podcast/214

    • There are lots of ideas out there, but this one appears to have no evidence behind it, and has been debunked multiple times. See eg https://www.reuters.com/article/factcheck-vaccine-safe-idUSL2N2NX1J6 (Nevertheless, of course it will circulate forever, like the infamous autism-vaccine thingy).

      Vaccines are one of the most important achievements of medical science, and mRNA vaccines are simply amazing. I hope they get them working against a lot of other viruses and cancer too, that would be a game changer. I expect that in the long run the majority of people will just accept mRNA vaccines routinely, as they accept more traditional vaccine technologies (no, I don’t live in the US).

      • Miraculous at killing tens of thousands of people, yes, I agree. You do realize even the inventor of mRNA technology said it’s unfit for human use, right? Or that the former Chief of Science for Pfizer of 16 years said it’s a mass depopulation event and he’d never take their vaccines.

        Even if you completely ignore the data, you don’t care that the vaccines have killed more people than every other vaccine ever invented put together, you still are faced with looking at a cost-benefit analysis. There is no benefit whatsoever, it doesn’t protect you against infection, and it’s not true that it makes you more likely to survive. In fact a lot of the data shows the opposite, that you’re more likely to be hospitalized or die after being vaccinated. A Canadian Provence showed exactly that. Even if you stuck your head in the sand and totally ignored all of the available concerning data, ignored who’s making a profit on what, and ignored the fact these drug companies asked for 75 years of sealed records, you’d still be faced with the reality that Covid never needed a vaccine. Period. It never even came close to any sort of threatening level of death for any healthy person. In fact, virtually nobody who was healthy died at all of Covid.

        If they had pushed the vaccines on those people 65+, I would have probably ignored the entire thing. I was excited they were developing the vaccines! I celebrated when the data first came out. I thought this is great, this should help the vulnerable. But then a funny thing happened… they started pushing unnecessary vaccines on young, healthy people, which isn’t medically justified whatsoever. Any potential benefit is so small it couldn’t possibly cover up the risk of barely tested vaccines (it turns out the trial data wasn’t good, anyway, and they covered that up too).

        • Andrew, just because you don’t like or disagree with a concusive statement, doesn’t mean it disinformation. We can hopefully establish a few undeniable facts.
          1) Vaccine development typically takes an excess of 10 years.
          – and with this multiple years of trials to ascertain the safety of the vaccine.
          2) the C-19 vaccine was “developed, tested and distributed” in less than a year. No long term studies were avaiable by the simple fact that a year is not long term.
          3) The CDC lied about a number of issues surrounding the disease and the vaccine.
          – Ivermectin is dangerous and ineffective against C-19 (lie)
          – You will never get Covid if you take the shot (lie, hence the comparison of excess deaths between vaccinated vs unvaccinated).
          – Everybody needs a Covid shot, even infants! (lie, for those under 18 influenza is more dangerous than Covid)
          4) Pfizer and Moderna are protected against any legal ramifications involving. You are unable to sue them for vaccine harm. Is this standard practice?
          5) Historically, viruses mutate and become less dangerous with each strain. All except for Covid (according to the CDC).
          6) A significant number of doctors (Dr Malone, Dr Gold, etc) have had their careers and lives adversely impacted professionally and financially because they disputed the government degreed narrative. I tend to believe people who are not acting out of their own self interest.

          Am I a doctor? No, but I am a critically thinking human who has heart disease, diabetes and is over 60. I did not get the jab. I did get Covid, took Ivermectic and get better within 6 days. (My wife only took 3 days). Anecdotal? Yes, but there you go. Suprisinly enough, I have not been sick since. While all of my friends who did get the jab and booster have now have C-19 multiple times. But the science!!! you say.
          But the money!! I say.
          “But the CDC wouldn’t lie!!”, Really, they’re a government entity and the government NEVER lies, right?

    • Bingo!

      Phil says, “Vaccines provide substantial protection” with ZERO proof. In fact, the deadly Covid vaccines and deadly, nonsensical Covid protocols like #Remdesivir while suppressing life-saving #Ivermectin have killed or injured many.

    • Seems the US state data is available at the CDC via the link John shared. I would be interested to see how US States who had differing policies compare to the EU, or at least CA vs TX, or NY vs FL.

      There’s a git repo at https://github.com/COVID19StatePolicy/SocialDistancing of policy changes and dates for things like emergency declarations, school closings, restaurant closing as well as relaxations of these policies for US states.

      • Dace: Any meaningful order would be fine. Simplest would just be decreasing order of population.

        Anon: I agree that “Don’t do X” isn’t as useful as “Do Y instead of X,” but it takes less effort and still has some value. Once you recognize that alphabetical order is generally a mistake (as it makes graphs harder to read and wastes a dimension of display), you can then think about what variable you’d like to use for the ordering. And, just speaking more generally, “Don’t do X” advice can be useful because of its generality. Every graph has its own story, but “Don’t display alphabetically” is usually (though of course not always) good advice.

        • I guess alphabetical ordering is a useful indicator that the presenter hasn’t tried to influence the viewer with a sneaky ordering (.e. the viewer doesn’t have spend time thinking “hmmm…what does the ordering mean”?)

          A little like setting multiple choice questions. Nowadays we’re supposed to list the possible answers (1 correct, four wrong) in alphabetical order and the students know this – and so they can clear from their minds the possibilty that the order of the possible answers might have some meaning that they might be able to divine!

    • Yeah, I thought about pointing out that I had used “Afghanistan First” ordering, just to avoid having you criticize me for doing so.

      In the text I call out Denmark, France, Belgium, isn’t it faster to find those if I use alphabetical ordering?

      I thought about other orders, like ordering by mean excess death rate or something, but none of my specific ideas seemed better than any other, plus I would have had to explain whatever I chose, so I thought fuckit.

  2. By looking the graph of US and Canada, I am moderately surprised that (1) the beginning of 2020 is exactly 0 and (2) the curve happens to hit zero. The mortality rate has natural fluctuation from year to year: the curve here almost appear too stable by chance.

    • Not sure about the OECD data but CDC say their excess death stats are truncated at zero when the deaths for a period are below the expected number.

  3. Fwiw, just sayin’…i think it’s pretty tricky to try to reverse engineer from population level epidemiological data – at least if you don’t have a foundation in biological or other evidence to ground the epi data in the real world. Just so many confounders – especially if you’re looking across nations or other widely different context factors. For example, how do you tease out the lack of seasonal flu deaths the last couple of years, when trying to understand excess mortality and what that might say about COVID, vaccines, etc.? How would you tease out the impact of the timing and difference in virulence as to when different variants hit, or the impact of vaccines, or better therapeutics?

  4. I’m intrigued by the “US states are more analogous to EU countries in pop size, public health response” hypothesis. Without asking for data that isn’t there (state-level), I wonder if a whole-EU data series would be illuminating? At a glance, it seems like different EU countries have second peaks (late 2020-early 2021) at different times, which I would expect to look somewhat like the US curve when superimposed.

  5. BTW if you want to make your own plot (and do a few other things) you can start with this. Download the file and save it as HEALTH_MORTALITY_2020_to_2022.csv’, then run the following R code:


    library(dplyr)
    library(ggplot2)

    mort = read.csv('HEALTH_MORTALITY_2020_to_2022.csv')

    mort = mort %>% mutate(week_from_start = (YEAR - 2020)*52.5 + WEEK,
    month_group = round(1+(week_from_start - 46)/53 ),
    year_frac = YEAR + WEEK/53.5)

    selected_countries = factor(c('USA', 'BEL', 'CAN', 'ITA', 'DEU', 'DNK', 'FRA', 'SWE', 'GBR'))
    selected_europe = factor(c('BEL', 'CAN', 'ITA', 'DEU', 'DNK', 'FRA', 'SWE', 'GBR'))

    ggplot(mort %>% filter(COUNTRY %in% selected_countries, AGE == 'TOTAL', GENDER == 'TOTAL', VARIABLE == 'EXCESSPC'),
    aes(x = YEAR + WEEK/53.5, y = Value, col = Country)) +
    geom_line() +
    geom_hline(yintercept = 0) +
    xlab('Year') +
    ylab('Excess mortality (%)') +
    facet_wrap(~Country)

    mean_selected_europe = mort %>% filter(COUNTRY %in% selected_europe, AGE == 'TOTAL', GENDER == 'TOTAL', VARIABLE == 'EXCESSNB' ) %>%
    group_by(year_frac, week_from_start) %>% summarize(tot = sum(Value))

    ggplot(mort %>% filter(COUNTRY %in% selected_countries, AGE == 'TOTAL', GENDER == 'TOTAL', VARIABLE == 'EXCESSPC'),
    aes(x = year_frac, y = Value, col = Country)) +
    geom_line() +
    geom_hline(yintercept = 0) +
    xlab('Year') +
    ylab('Excess mortality (%)') +
    facet_wrap(~Country) +
    annotate('line', x = mean_selected_europe$year_frac, y = mean_selected_europe$tot/600 ) # Superimpose the scaled mean over the selected european countries
    # Scaling should be based on the baseline deaths but here I (PNP) have just scaled it to match the initial US peak, for comparing shapes.

    # Total deaths in (a) first six months, (b) following year, (c) the year after that.
    smry = mort %>%
    filter(COUNTRY %in% c('USA', 'BEL', 'CAN', 'ITA', 'DEU', 'DNK', 'FRA', 'SWE', 'GBR'),
    AGE == 'TOTAL', GENDER == 'TOTAL', VARIABLE == 'EXCESSNB') %>%
    group_by(COUNTRY, Country, month_group) %>% summarize(tot = sum(Value))

  6. The more people who have had covid, the more people (like me) who have long covid. Lack of energy to get much exercise; concentration on the few foods which still taste good may result in malnutrition; fogginess causing a lot of dumb mistakes; and not as much will to live might be secondary factors contributing to secondary waves of excess death. (Or maybe I’m just getting old.) Plus there’s Trump. How many other countries have had peak Trump along with covid? (Maybe North Korea.)

  7. Most of the comments here are on the mRNA kills people hyperbole and the alphabetical order of the graphs.

    Anyone ever considered looking at the actual demographics of the countries listed? The US has dramatically different demographics than Canada or than any country in Europe higher rates of immigration as well. The impact on different groups in the US population surely explains some of the higher mortality in the US.

    • Yeah, it’s funny that nearly no one is discussing the substance of the data. I find this pretty interesting but maybe most people don’t.

      You say “the US has dramatically different demographics” but I’m not sure what you mean by that. I don’t disagree with it, depending on what you mean, but I would have thought the most important demographic factor would be the age distribution, and that’s not that different from some European countries. We do have higher rates of obesity and diabetes and perhaps other co-morbidities, I think, but a few European countries aren’t far behind.

      One thing that’s very clear is that the US data are driven mostly by COVID. I say this based on data not in evidence but easily available: take a look at the plot of COVID hospitalizations vs time. No doubt the pandemic has also changed things with regard to suicides and overdoses and car accidents and murders, perhaps to a degree that is of great practical significance (I don’t know) but the temporal pattern is dominated by COVID.

      Another thing that requires looking elsewhere for data: booster shots. Somehow I had forgotten that people are counted as “fully vaccinated” even if they haven’t gotten a booster. And only about 30% of Americans have gotten boosters! Presumably that is heavily skewed towards the most vulnerable people, but still…jeez, wtf is stopping people from getting that extra shot?

      The large between-country variation in excess mortality is interesting. I didn’t expect it. At least as far as mortality goes, the pandemic ended in Sweden over a year ago, and only _started_ in Denmark about a year and a half ago!

      Belgium, France, and Italy all look about the same, which maybe makes sense given that they are contiguous.

      To me the big question is: what are the successful countries doing that the unsuccessful ones aren’t? A piece of the puzzle could be vaccinations — I think the U.S. is lower than any European country in vaccinations — but I don’t think that could be the whole story. I’m kinda hoping someone else with an interest in this will find a source of vaccination data and maybe demographic data and do some sort of analysis or at least make some more graphs.

    • Phil: this is what I mean:

      https://www.kff.org/coronavirus-covid-19/issue-brief/covid-19-cases-and-deaths-by-race-ethnicity-current-data-and-changes-over-time/

      Check out Figure 3, where the (multiply adjusted) rate of mortality among Hispanics was almost 2.5x higher than for whites in F20/W21. Hispanics are >18% of the US population, but this data implies Hispanic mortality was very roughly numerically equal to white mortality in F20/W21. In other words, Hispanic mortality was so large that, if the mortality rates had been equal to that of whites, it would imply roughly another 100m people in the US. I’m not sure exactly how that translates into “excess deaths” but it’s gotta be a pretty big punch.

      In the fall/winter of F21/W22, Hispanic mortality was much lower than the previous winter, only slightly above that of whites. But, while Black mortality in F20/W21 was only about 45% greater than whites, it fell only slightly in F21/W22 and, as white mortality fell more, the difference between blacks and whites actually increased.

      I ascribe these differences to several demographic factors that are unique to the US. First, the has much higher overall immigration rates than Europe/Canada. Second, a larger percentage of immigrants to the US come from 3rd world or undeveloped countries and arrive speaking little or no English. Third, we have by far the highest Hispanic immigration of any non-Hispanic country. I can’t say for sure but my bet is these immigrants are overwhelmingly poor and uneducated.

      The immigration factor is important because, in addition to not speaking English and having little education, many immigrants live in large family groups were COVID is easily transmitted. This is normally how they build a strong financial base and become prosperous in US society, but during the pandemic it was a liability.

      Not all the answers and a little garbled but I think this covers a good part of the difference between Europe/Canada and the US.

      This is not to say we should not accept these immigrants. But it’s clearly a factor in COVID mortality in the US.

      • Hispanics are >18% of the US population, but this data implies Hispanic mortality was very roughly numerically equal to white mortality in F20/W21.

        Hmmm… we can look at the actual numbers in the CDC datasets which relate to your figure 3 (click on the data link below the figure). It doesn’t make sense to focus on one particular period of the epidemic unless one is specifically interested in factors underlying time-dependent differences as a function of race. In terms of race contributions to Covid mortality the full Covid period is much more meaningful (especially if one wishes to make decisions about whether to “accept these immigrants”).

        Total Covid deaths in period 1.1.2000 to 27.8.2002 = 1,042,006
        Deaths in “white” population = 677,198
        Deaths in Hispanic population = 164,423

        So rather than “Hispanic mortality …very roughly equal to white mortality”, if assessed through the full epidemic Hispanic Covid-related mortality is 24% of white mortality. Since the Hispanic population is around 19% of total US population, averaged over the full pandemic the Covid-related mortality as a proportion isn’t so different though it is a bit higher (for reasons that are discussed in the CDC pages).

        I vote that we accept these immigrants!

        • “I vote that we accept these immigrants!”

          I make no vote on immigrants or immigration, but apparently you think I am voting on that, so, to protect your views on immigrants, you’re forced into the position of minimizing the impact of the Hispanic population on mortality F20/W21 by spreading it out across the entire pandemic, rather than recognizing that it was an outsized contribution to the F20/W21 wave of the pandemic and a significant cause of higher US mortality.

          FYI, likely a big factor in the drop off of Hispanic mortality in the F21/W22 wave was a set of massive outreach /education programs in affected states. Credit Hispanics – unlike whites – for reacting to that outreach in a positive way.

          It **does** make sense to focus on part of the pandemic to explain that part of the pandemic! If you live in an area with a large Hispanic population, as I do, you would know that a likely reason for the later decline in Hispanic mortality is a gargantuan effort on the part of the government to educate the Hispanic population.

          If you don’t want an accurate explanation of what happened during the pandemic (or, what happens socially in general) in the US, you can persist in the belief that “race” is just a skin color people accidentally got splashed with at birth, rather than recognize it for what it is: a skin color shared by one or more cultures each of which has to varying degrees distinct behaviors that may both contribute to and detract from its success in different environments.

        • I’m a Brit and not subject to the tedious partisanship that you seem to be displaying – so no need to attempt to define my “belief”s through your perspective.

          Your post rather explicitly equated some (misinterpreted) aspect of race-specific response to Covid with some ideological point about whether one “should accept these immigrants”. I’m pointing out that that seems rather dubious and is in any case ill-informed and illogical.

          This thread is about excess deaths. Rather than your insinuation that a defined period of age-adjusted deaths implies a near equivalence in hispanic and white mortality, the actual numbers indicate that the proportion of hispanic Covid deaths is not so different from the proportion of US hispanics.

          In reality the contribution to excess deaths in the US is dominated by the white population. We don’t need to impute some moral significance to this (although it would be nice if a large proportion of the US population didn’t engage with vaccine misrepresentation). It’s a consequence of the relative age distributions of the populations. The data you linked to, if you care to look at the numbers, shows this. So of the 677,198 US “white” Covid deaths, 548,842 are in the 65+ age group (81%). Of the 164,423 hispanic Covid deaths 97,627 (59%) are in the 65+ group. Because of this age-related disparity there are more white Covid deaths relative to the proportion of whites in the US population than there are hispanic Covid deaths relative to the proportion of hispanics in the US population.

          Let’s kick out the whites and send them back to Europe!

        • It seems that there’s more focus lately on the limited utility of “Hispanic” as a distinct voting block (since it lumps together a population that’s widely disparate and comprising sub-groups with fairly distinct voting behaviors). Although black, non-Hispanic white, and Hispanic seem to be still widely used when examining health outcomes, I wonder to what extent that type of analysis likewise has limited utility.

          That said… I recently read that while initially Hispanics were getting vaccinated at a rate lower than their % of the population, that has turned around and in total they’re now vaccinated at a rate higher than their share of the population.

        • Chipmunk
          You seem to think you are focused on the facts, but you conveniently apply your own reasoning to the data to the exclusion of other views. The KFF data you point to, contains the following speculation about why Hispanic death rates are so high:

          “The higher rates of infection among people of color likely reflect increased exposure risk due to working, living, and transportation situations, including being more likely to work in jobs that cannot be done remotely, to live in larger households, and to rely on public transportation.”

          Of course, that is speculation and further analysis can be done that might support that speculation or not. But these reasons do sound a lot like a group that has skin “splashed with color.” By that I mean that had other demographic groups had these characteristics (working, living, and transportation) they might have had the same mortality as Hispanics. But you seem convinced that it is something intrinsic to that population. We just can’t tell without more data and analysis.

      • To add to the confusion here, a point that came out during this blog’s discussion of the “deaths of despair” brouhaha was that the Black and Hispanic populations’ death rates in the relevant age groups was/is far far higher than that of the whites that were maybe (the whole deaths of despair thing was probably an artifact of incorrect age correction of the statistics) seeing a sligh increase in death rates.

        This difference in base rates must make calculating, reporting, and understanding excess deaths in different parts of the population a bear.

        Speaking of statistics, here in Japan, which has until recently had a long period of almost no immigration, about 11% of new tuberculosis cases are amoung recent immigrants, most of whom are in “work training” programs (and nowhere near 10% of the population). Context: Last year, Japan became the last industrialized country to make it into the “low TB” classification, although the reduced rate of TB is probably due to the medical system being swamped with Covid and not testing for TB. “Cough? PCR negative? Not a problem. Go home” Oops.)

      • The US has a much LOWER immigration rate than Canada. Wikipedia gives 21% of the population of Canada as foreign-born, compared to 14% of the population of the USA (the EU would be complicated because it varies from country to country and because of what kind of immigration this trope is about). Xenophobes in the USA often try to blame poor public health statistics on visible minorities or immigrants, but I don’t think that works for COVID deaths.

  8. I must say what I thought in 2020 and 2021: ‘next years there will be far less mortallity because all the frail people will die from covid.in 2020 en 2021. I got it wrong.

  9. I think it would be very interesting to look at cumulative excess mortality as well.

    It might just be that the US is having a more spread out infection compared to other countries.

  10. > I don’t know the exact definition used for these data

    > that I selected somewhat haphazardly

    The above statements, and the other random musings, are not the quality of statements I expect from this website.

    • There’s nothing wrong with selecting a grab bag of countries to compare to.

      As for “excess deaths” or “excess mortality”, there are minor variations in how this gets calculated — do you calculate the baseline over 10 years or 15 years or what, do you do mean or trimmed mean or median, do you do any smoothing over weeks, etc. For exploratory purposes as shown here, and indeed for many other purposes, it is not important to know the exact definition that was used.

      You are, of course, free to dislike the post!

    • Boosters are completely unnecessary result of miscommunication of why and how vaccines work, that became a policy.
      Vaccines do not prevent transmission. They ‘shrink’ the peak of contagiousness and that’s it. Their primary benefit is in cellular immunity to protect one from dying.
      The figure of 30% of vaccinated that are also boosted hasn’t changed for months and it’s clear it doesn’t have any beneficial effect, beyond that of vaccination. Any dose, initial or a ‘booster’ raises the level of antibodies for a few weeks, but that’s for extracellular immunity only. Once the cell is infected, ABs are useless and all that protects you is cellular memory from the initial shot.
      There is much larger within group variability in viral shedding, length of the ‘peak’, severity etc. among individuals, than the between group difference b/w boosted and not boosted.
      Again, miscommunication of how these things actually work.

      • RE: “There is much larger within group variability in viral shedding, length of the ‘peak’, severity etc. among individuals, than the between group difference b/w boosted and not boosted.”

        I know it’s only been 2-1/2 years and that real science (not pseudo-scientific posturing) is difficult and takes a while. But I’m struck by the paucity of insight gained so far into that “variability in viral shedding…” and related questions.

        For instance, some people seem to become infected after fairly limited exposure to just one infected individual. Yet other people can live in the same house and even share the same bed with a person suffering protracted COVID illness but never catch it from their partner.

        What are the factors underlying what seems to be a multiple orders of magnitude range of susceptibility among individuals? Why does COVID seem very contagious in one situation and not at all contagious in a superficially identical one?

        I keep hoping a literature will emerge that sheds some light on these questions but all I’ve seen so far are unconvincing, broad-brush ecological or secondary-data models combined with anecdotal clinical experiences.

      • Navigator –

        > They ‘shrink’ the peak of contagiousness and that’s it [. ..] it’s clear it doesn’t have any beneficial effect, beyond that of vaccination.

        Seems a rather definitive statement. Given the high degree of certainty we’ve seen throughout the pandemic, I’m a bit skeptical about that degee of certainty. Can you provide some links for (some of?) the evidence that makes you so certain?

  11. FWIW, the numbers on excess deaths in Japan are in: negative for 2020, slightly positive for 2021, and probably very bad this year.

    Life expectancy in Japan continued its upward trend in 2020, due to fewer deaths amongst the elderly due to less influenza (everyone, and I mean everyone, started wearing masks in mid-March, and we still are)), but due to what was, IMHO, somewhat problematic handling by the government, life expectancy was down in 2021. (According to the Japanese press a couple of days ago.) Still, they made an attempt at getting all us old folks vaccinated and reduced the number of people travelling (both into/out of and within Japan), and made bars and restaurants close early. Not quite adequate, but it was working somewhat.

    But this year, the Japanese government has essentially completely given up on doing anything about Covid, and (infections and) deaths during the current wave are higher than they’ve ever been (over 300 deaths a day in a country 2/5 the size of the US). The press reports 60% of deaths are in the over 80 age group and 20% are in the 70 to 80 age group, so it’s only going to minorly reduce life expectancy, but it will do so. Well, what is the government doing, you might ask? I mean, this is major medical emergency, hospitals are swamped and more people are dying than ever before. It must be doing something, right? Wrong. It turns out that reporting all Covid cases has become a paperwork burden for the health care system, so they decided to make said reporting optional*. No states of even semi-emergency, no travel restrictions within Japan, and increased numbers allowed to enter/leave Japan.

    * There was even a seconday farce here. The original idea was that reduced reporting was to be optional on a per-prefecture basis, but Osaka and Tokyo said “No thanks, we want to know what’s actually happening”, so the government is making the reduced reporting thing non-optional.

      • I didn’t know that (i.e. that Unherd is on “skeptic” side). I thought it was a very strightforward and enlightening discussion with an excellent choice of interviewee.

        Very occasionally I had the impression that the interviewer (the posh Brit) was trying to push things a little towards the perspective of mal-effects of vaccines but on balance it seemed like he was just trying to explore evidence for this particular POV.

        I very quickly stopped looking at the viewer comments

Leave a Reply

Your email address will not be published. Required fields are marked *