Coronavirus Daily Death Rates by Country

Very high in Italy and Spain, followed by France, Switzerland, Belgium, and the Netherlands, followed by Denmark, Portugal, Ireland, Austria, the U.S., and Germany.

|The Volokh Conspiracy |

Every day in the developed world, about 20 to 30 people out of every million die. (The typical yearly death rate is about 7.5 to 10.5 per thousand, depending on how many younger people there are compared to the elderly; 7,500 to 10,500 per million, divided by 365, equals 20 to 30 per million per day.)

Judging by the WorldoMeters data from the last three days, here is the mortality owing to coronavirus, which pretty much adds to that daily death rate (averaging deaths over the last 3 days):

  • Italy and Spain, with 14 to 17 deaths per day per million.
  • France, Belgium, the Netherlands, and Switzerland, with deaths in the 4 to 5.5 per million range.
  • The UK at a little under 3 per million.
  • The US, Germany, Austria, Sweden, Denmark, Ireland, and Portugal, at a little under 1 per million (Germany, 0.9) to a little under 2 per million (Denmark, 1.85); the US is at 1.2, though the daily numbers in the US have risen sharply in the last two days. (Keep in mind, of course, that some of these comparisons may be less significant if countries have different protocols for listing cause of death.)
  • South Korea, China, and Canada are very low, below 0.25 per million (though there may be reason to be even more skeptical about Chinese data than about data generally).
  • Japan and Taiwan have been barely hit at all; likewise for most of the rest of the world outside some Western European countries that I haven't listed. (Russia is ostensibly in this category, though no Russian ever trusts numbers from Russia.)

Of course, the numbers can be much higher in particular regions. News accounts report that 84 people died per day in New York City Thursday and Friday. Over a population of 8.6 million, that's about 10 deaths per day per million, not far off the national Italy and Spain numbers.

Likewise, most coronavirus deaths in Italy are still in Lombardy: 542 in one day, apparently Friday, out of a population of 10 million, for a daily death rate of 54 per million. That means that in Lombardy the daily death rate is basically triple the usual number. (Of course, especially since the dying are mostly the elderly or chronically ill, a much higher than usual death rate now will probably mean a considerably lower than usual death rate after this epidemic passes, because there will be fewer elderly, more-likely-to-die people left, assuming those who survive the illness won't be permanently weakened by it. But in the meantime, it's still people dying sooner than they otherwise would.)

The question, of course, is where are we going? (And, while we're at it, what's with this handbasket?) Will the daily surplus death rate substantially increase? Will it substantially decrease? How quickly? That we do not know.

NEXT: DHS Adds Workers for Gunmakers, Gun Retailers, and Shooting Ranges to "Essential Critical Infrastructure" List

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. The US death rate varies by month, it’s higher in winter than summer, largely because of the flu. See CDC data for 2017: https://www.cdc.gov/mmwr/volumes/68/wr/mm6826a5.htm

  2. Make it a point to notice. Based on Johns Hopkins data, the worldwide death rate attributed to coronavirus has been creeping steadily upward. Last week it was in the upper 3% range. As I write it is at ~ 4.6%. I have no idea what accounts for that. Is there some way to use statistical methods to make it seem encouraging?

    1. ” Is there some way to use statistical methods to make it seem encouraging?”

      Yes — actually USE statistical methods, which is not yet being done.

      First, all of these mortality percentages are totally irrelevant because we have no idea what our actual infection rate is. To know what it is, we’d have to test *everyone* (or a large enough random sample to accurately represent everyone), and we’d have to do that daily. It’s an unjustified use of scarce resources, but you have to remember that we are not doing it.

      But what you can do is look at a similar situation where we don’t test everyone (for the same reasons) — the flu. Flu mortality of those who are tested is 3% — while the actual mortality is 0.1%. The reason for the discrepancy between the two numbers is that (a) the vast majority of the people with the flu are never tested and (b) only those who are the sickest are.

      Second, all the death statistics mean is that the Wuhan virus was found at autopsy — NOT that it killed the person. It is entirely possible (and common) for one to have multiple infections, and we have a particularly virulent version of the flu going around right now as well.

      Think of OUI and single car fatal crashes. If a drunk driver dies, it’s blamed on the driver being drunk — with no mention of the bald tires, the black ice, the treacherous curve, or the inexperience of a 16-year-old who just got his license. So what really caused the car crash?

      I’m not saying the Wuhan Flu isn’t serious — it is. But one has to remember what the mortality statistics are actually telling us — and what they are NOT telling us…

      1. Dr. Ed, you do not seem to have addressed the increasing trend in the fatality rate at all. It is a statistic, however (or however erroneously) derived. Assuming for the sake of argument that it is erroneous, what real-world errors are responsible for its steady rise?

        I ask because I am having trouble imagining any explanation except that maybe the virus kills some folks quickly, and others more slowly. Thus, the rise reflects that second wave of mortality in earlier-affected regions, as they are added to the first wave occurring in later-later affected regions.

        But I have no confidence that is a correct interpretation. The only justification for mentioning it is that it seems superficially plausible, and not self-contradictory—which is pretty weak justification where evidence is needed. What mechanism can you propose to account for an observable trend?

        1. “maybe the virus kills some folks quickly, and others more slowly”

          Well, some of the people put onto ventilators will die anyway (i.e. more slowly) and hence create a blip in the statistics.

          1. @Dr Ed: FYI, EVERYbody dies eventually. Some just do it later than others.

        2. “what real-world errors are responsible for its steady rise?”

          It’s difficult to parse what you are actually saying, but if you are observing that the simple death rate (number of deaths divided by number of cases) is rising, that’s what you’d expect in an epidemic, just from the math (and, eventually saturating the hospitals will also have an effect).

          Consider a hypothetical epidemic that has a very predictable outcome: exactly 50% of people die, and they do so exactly 4 weeks after infection, and the number of infected people exactly doubles each week. We’ll start week one with two infected people, one of whom will die in 4 weeks. Given the formatting difficulties, following along with a pencil and paper might make it easier.

          Week / #new infected / cumulative infections / #new deaths / cum deaths /% fatal
          1 / 2 / 2 / 0 / 0 / 0%
          2 / 4 / 6 / 0 / 0 / 0%
          3 / 8 / 14 / 0 / 0 / 0%
          4 / 16 / 30 / 0 / 0 / 0%
          5 / 32 / 62 / 1 / 1 / 1.6% (one of the orig 2 patients dies)
          6 / 64 / 126 / 2 / 3 / 2.3% (two of the week 2 pat. dies, 3/126=2.3%)
          7 / 128 / 254 / 4 / 7 / 2.7%

          and so on for some time, until the curve starts to flatten. So the observer fatality rate is climbing from 1.6 to 2.3 to 2.7, even though the actual fatality rate is constant.

          1. Okay, I follow your math. Thanks for the explanation.

            So tell me what the fatality rate is after the epidemic is history. Isn’t it 50%? So why not draw an inference that a steadily rising week-to-week cumulative death rate is indicative of an actual fatality rate higher than the week-to-week figure indicates, at least until that stabilizes.

            How does the pattern you suggest get interpreted to mean cumulative week-to-week death percentages are inaccurate because they are higher than the actual fatality rate? Your example shows the opposite, that the week-to-weeks are lower than the actual death rate. That is the anomaly I am trying to unravel—the notion that a steadily rising cumulative death percentage can somehow be indicative of, or consistent with, a much lower death rate for the pandemic.

            1. It’s obvious that if you only test the most severe cases, 100% of your cases will be severe.

            2. “Isn’t it 50%?”

              Yep.

              “How does the pattern you suggest get interpreted to mean cumulative week-to-week death percentages are inaccurate because they are higher than the actual fatality rate?”

              ElvisIsReal nails it. We know the numerator – the number of deaths – but we don’t really how many people have been infected but were asymptomatic or had a mild enough case they were never tested. There is some limited data from e.g. cruise ships, but unfortunately we just don’t have hard data at this point.

              1. Actually some of the best data is out of Iceland currently, where they’ve tested 3.4% of the entire country. (12615 tests out of a population of 364,000).

                https://futurism.com/neoscope/half-coronavirus-carriers-no-symptoms

                They’re showing a non-symptomatic rate of almost 50% within the study.

                1. Not mentioned is how they selected that 3.4%, and if it was a statistically valid method. Notwithstanding that, this is really interesting…

              2. Absaroka, seems to me you and ElvisIsReal are content to argue that there are plausible reasons, supported by some evidence, to suppose that mortality from coronavirus is lower than it now seems. I get that. I already understood that reasoning. But neither of you seems to be offering any way to understand how an upward trend in the worldwide mortality rate could be consistent with your other argument.

                1. The upward trend on the mortality rate is exactly what you would expect with an increase in testing. As those jurisdictions that prioritize testing of people with symptoms (like the United States) increase, that will increase the Case Fatality Ratio because more symptomatic people are getting tested. It is possible that it is going up in some hot spots because of triage as well, but that should not be a significant cause of the increase. As testing goes to a broader group of people (including the asymptomatic ones) the CFR will drop.

                  The CFR is not the same as the Infection Fatality Ratio. The scientists are putting the IFR generally below 1%. Oxford researchers think it’s between .1% and .26% (although that has not yet been peer reviewed). Though that will change significantly by country, due to demographics, comorbidities, etc. The Imperial Academy estimate for the K was .9%, based on an earlier study that had it at ~.5% in China.

                  1. The upward trend on the mortality rate is exactly what you would expect with an increase in testing

                    High testing finds mostly the ones who are infected but don’t show symptoms, or very mild.

                    This increases the confirmed cases, but not deaths, which reduces the mortality rate.

                    It can also uncover infections who later die. The death count is understated by an unknown degree, from deaths by unconfirmed infections.

      2. Think of OUI

        Oh, yes. I remember that magazine. A favorite of my teen friends and mine way back when.

      3. If you aren’t going to use the correct names of the virus then your opinions on this subject are worthless.

        1. Woo woo, the language police have arrived.

        2. Are you really suggesting that people use SARS-CoV-2 whenever discussing the novel coronavirus? That’s a mouthful and it’s not as if it’s set in stone to be the name forever, nor is it known yet whether it will end up being multiple viruses responsible for COVID-19, as gets more likely with every day (which is not the name of the virus but the disease).

          1. nor is it known yet whether it will end up being multiple viruses responsible for COVID-19,

            Everyone knows what you’re talking about when you say COVID or COVID-19, don’t pretend otherwise.

            It’s DEFINITELY not a flu, so Dr. Ed is making a definitely wrong choice for…reasons.

      4. “Second, all the death statistics mean is that the Wuhan virus was found at autopsy — NOT that it killed the person. It is entirely possible (and common) for one to have multiple infections, and we have a particularly virulent version of the flu going around right now as well. ”

        That is not true of the Italian statistics. If covid was detected at any point, a death is counted as a covid death. That obviously over counts.
        The method that you cite is close to the German method in which covid is the proximate cause of death.
        Nonetheless the German “rate” has been getting larger every day

    2. Sudden increase in exposure (as in the now global spread and in all states) kills a lot of already dying and those at high risk before we know all of the less at-risk people who’ve been infected. We won’t ever know most of the people who were infected but asymptomatic. The rate will probably stabilize over time, once we get more into a steady state.

      There’s also the problem with widespread testing not being available for those who aren’t in much danger but still may have it. The obviously sick and endangered tend also to be the ones who die, so it’s a bit of testing bias.

      There’s also the possibility that it is more deadly now. I don’t know how the theory is doing now (it depends highly on Chinese reports, which have gotten increasingly suspect), but the initial Wuhan infection seemed to be more deadly than subsequent infections in other regions of China; it was expected that a mutation that decreased lethality allowed it to “sneak” out. It’s also possible that some strains of the virus are more lethal than they were before. It’s had a ridiculous number of chances for mutations to increase lethality.

    3. It’s telling that the only articles on the internet highlight how it’s going from bad to worse. NY saw a drop day over day in deaths from 578 to 257 and you can’t find anyone remarking on that drop. Instead, you’ll see site after site saying the sky is falling because deaths are over 1000 now.

      Obviously this is a scam.

      1. Uh, deaths are over 2,500 and doubled in just two days,
        No conspiracy. Italy dropped after hitting 983, then skyrocketed back.

        A one day drop says nothing, when the disease takes up to 14 days to show symptoms. On the day the deaths drop, thousands of Americans are walking around, each infecting 2.5 others (on average) for as long as two weeks.

        New York’s deaths also fell on the 26th, then went back up.
        If you’re not informed on the issue, even the fundamentals, you’re certain to remain a victim of manipulation by the political elites,

        In other words, it’s you fomenting hysteria, and also a victim of hysteria. Hysteria is far more contagious than coronavirus!

      1. Balancing human life against an economic collapse is not easy to contemplate, but an economic collapse will cost a lot of lives also. Sometimes there is no good choice.

        1. Not near as many lives. Imagine these deaths keep doubling until there’s a vaccine, over a year by most estimates,

          At what level of deaths do people refuse to leave home? Most of the stockpiling is to minimize going out, and shopping.

          Trump’s delays and denials — a Democrat hoax — for over a month, have permitted at least half of all our total deaths.

          1. “Imagine these deaths keep doubling until there’s a vaccine, over a year by most estimates”

            That is not going to happen — for the same reasons a Ponzi scheme will not work.

            1. You need a better reason.
              And learn what a Ponzi scheme is!

              1. Have you actually done the math to see what 1000 doubled every two days for one year gives you?

                1. That would be stupid, since I neither said nor implied that.
                  Are you seriously denying that Trump’s failures, and over a month of needless delays (DEMOCRAT HOAX) has allowed hundreds of needless deaths … eventually thousands? I’ve done THAT math.

                  1. It’s stupid to ask if you did the math because you didn’t say nor imply that you did the math? That’s… not a statement that instills my confidence in your suppositions.

                    You did imply that we should do the math, though: ““Imagine these deaths keep doubling until there’s a vaccine, over a year by most estimates”

                    1. It would be stupid because I neither said nor implied such a thing. And EXPLICITLY said the opposite.
                      Stupid for ME because I know enough that they fluctuate, and why.

                      “Imagine these deaths keep doubling until there’s a vaccine, over a year by most estimates”

                      Where do you see “every two days,” Skippy?

                      You also missed THIS.

                      The PACE of increased deaths depends almost solely on human contact, which is why every country recommends social distancing … enforced by police, in a growing number of countries (Most recently, England).

                      Umm, PACE means how often it doubles.

                      Anything else?

                    2. If one plots the number of US deaths over the past month, you will see that is proportional to the number of cases. That number is rising exponentially. At a rate slightly higher than the rates in Italy and Iran. If you remove NYC from the US statistic you’ll find a doubling every 3 days.

          2. Let’s talk about Trump for a bit.

            Trump’s job as president here is to balance the need for further precautions with the need to reassure the country and avoid panic. Which I think he’s done as well as can be expected. Some of his actions would good (closing travel with China, with Europe). Some could’ve been better (faster testing). He’s got to rely on the experts to an extent, and the experts (especially with regards to testing) have been failing…utterly.

            Trump also has to balance the politics of the situation. To be honest, he should’ve quarantined New York at least 2 weeks ago. But that was politically not feasible. You’ve seen what the response to that suggestion was, even today, from the liberal side of the aisle.

            So, that’s where we are.

            1. Why do you say there was no national crisis until 16 days ago?
              He should have recognized a crisis at least a month before he did. Instead he was delaying, denying, lying and just plain nuts,
              Did YOU believe it was a Democrat hoax?
              Did YOU believe it would just magically go away.
              When we had only 10 infected people, did YOU believe it would be at near zero in a few days?

              Why do YOU agree that
              1) Trump should do NOTHING to increase ventilators, WHILE PEOPLE FUCKING DIE?
              2) Trump should provide aid to only Governors who are nice to him (while he’s screwing them over)?

              If Presidential leadership matters, how many of all these delays, shortage and deaths were allowed by his total lack of leadership, until a mere 16 days ago?

              1. ” Trump should provide aid to only Governors who are nice to him (while he’s screwing them over)?

                Prove it!

                That’s a damn serious charge — it actually would be grounds for Impeachment were he really doing it, but he isn’t. Of course the supplies for NYC are in NJ — I doubt that there are any large parking lots in NYC where one could drop a lot of 40′ trailers. (The Feds like to use Home Depot parking lots for this as they are near major highways and designed for large vehicles.)

                Would you have preferred Trump to have shipped the stuff to Buffalo??? Yea, it’d have been in New York State, but how many hundred miles away?!?

                1. He bragged about it, on Fox.
                  He has EXPLICITLY REFUSED to supply states with more life-saving ventilators, using an executive authority not available to governors. Which I mentioned in boldface

                  Whatever supplies you’re talking about has not as shred of relevance to anything I said.

                  1. Oh shite, Hiln said something in boldface (not that he overuses it) so how could you miss that; it must be true.

                    1. NO, I PUT IT IN BOLD … AND HE DIDN’T SEE IT.
                      SO I’LL TRY ALL CAPS, TOO.

              2. You’d lose. Again.

            2. “Which I think he’s done as well as can be expected” . . .

              as expected by diffusely bigoted, stale-thinking, science-disdaining, downscale, elderly white males . . . in other words, conservatives and Republicans.

              1. Of course Rev.(???) DeBlasio did very well when he had the power and authority to order a police-enforced lockdown in NYC 3 weeks ago. Instead he put NYC money ahead of public health. No wonder Cuomo is disgusted with him.

          3. Disease do not follow exponential curves. They can’t. To make the extreme point, if diseases increased exponentially, at some point the next day’s doubling would require killing more people than exist on the planet.

            Diseases follow an S-shaped curve, sometimes called a Logistics curve but in the context of epidemics, more precisely called a Gompertz curve. A Gompertz curve starts out looking sort-of like an exponential curve but flattens and then begins to approach an upper asymptote.

            It’s worth noting that “social distancing” and “flattening the curve” are all about stretching out the time for the infection to pass through the environment. While that delay has some value in the context of scarce medical facilities, it does nothing to change the position of that upper asymptote. Eventually the same number of people are going to get sick regardless.

            This means that most of the disease deaths are unavoidable. All you’re doing is changing the timing. And while that has some value, it does not necessarily justify shutting down the economy and causing the many, many new deaths that will result from the shutdown.

            That’s a cold-hearted analysis. But if we’re going to address this problem with reason and not emotion, we have to start working from actual facts and science.

            1. To make the extreme point, if diseases increased exponentially, at some point the next day’s doubling would require killing more people than exist on the planet.

              It doesn’t double every day.

              PAST acceleration, til today, means Trump’s delays have allowed 2/3 of the deaths.

              The whole point of flattening and reversing g the curve is that the fewest number of people die, while a vaccine is being developed and tested. Why do you dismiss that goal?

              While that delay has some value in the context of scarce medical facilities, it does nothing to change the position of that upper asymptote. Eventually the same number of people are going to get sick regardless.

              Learn what vaccines are!

              1. You obviously still don’t get it. And obviously still haven’t done the math. Your exact quote above was ” Imagine these deaths keep doubling until there’s a vaccine, over a year by most estimates…” My rebuttal is that I don’t have to imagine that because that’s a wildly unrealistic hypothetical. It will never “keep doubling”. That not how disease progression works.

                I dismiss vaccine production because it can not be ready in enough time to do any good. No matter how much you try to “flatten the curve”, this virus will have swept through the entire population long before the year or so that it takes for vaccines to be developed and mass-produced. Waiting for vaccine is not the goal of flattening the curve at all. The goal is to prevent the overload of ventilators and other scarce medical facilities that we already have.

                1. Your exact quote above was ” Imagine these deaths keep doubling until there’s a vaccine, over a year by most estimates

                  THANKS FOR PROVING ME RIGHT!

                  YOU:

                  To make the extreme point, if diseases increased exponentially, at some point the next day’s doubling would require killing more people than exist on the planet.

                  It doesn’t double every day.

                  And you QUOTE me NEVER saying daily doubling!!!!

                  that’s a wildly unrealistic hypothetical. It will never “keep doubling”

                  NOT EVERY DAY .. AS YOU FALSELY CLAIMED.
                  “the next day’s doubling”

                  (Did he forget his own point? Or change it when he was called out? Perhaps Both?)

                  1. 1. You haven’t proven anything.
                    2. You said “keep doubling”. Whether daily, every 2 days, weekly or even monthly doesn’t matter. It can’t keep doubling – and it won’t. That’s not how epidemics work.

                    By the way, bolding and all-capping makes your rants look increasing strident and irrational. Give it up. You clearly don’t understand the math.

                    1. 2. You said “keep doubling”. Whether daily, every 2 days, weekly or even monthly doesn’t matter. It can’t keep doubling – and it won’t.

                      ONE MORE TIME FOR THE CRAZED PSYCHO …. HE quoted me saying THIS!

                      Your exact quote above was ” Imagine these deaths keep doubling until there’s a vaccine, over a year by most estimates
                      https://reason.com/2020/03/29/coronavirus-daily-death-rates-by-country/#comment-8187735

                      until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine, … until there’s a vaccine, …

                      By the way, bolding and all-capping makes your rants look increasing strident and irrational.

                      THAT’S RIDICULE
                      *SNEER*.

                      You clearly don’t understand the math.

                      Until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine, … until there’s a vaccine, …until there’s a vaccine,

                      ***DOES YOUR HEALTH INSURANCE COVER THE EGO TRANSPLANT YOU NEED?

                      Disclaimer: Not all Trumptards are THIS bad.

            2. The problem I have with “flattening the curve” is that it overwhelms medical resources via exhaustion. People can work 18 hour days — for a while, not indefinitely. A lot of other routine medical stuff (e.g. cancer surgeries) can be put off — for a while, not indefinitely.

              We already are accepting a higher morbidity and mortality rate for “other stuff” in redirecting resources to the Wuhan Flu. At what point is the marginal return on the latter exceeded by the former?

              1. Of course the cost-benefit analysis would be critical, but off course since there are so many uncertainties in that analysis at this point its basically useless. Furthermore, the question is who bears the cost for whose benefit, you know the moral hazards problem. Geez if the was just a plug and chug solution to the problem we could make that decision expeditiously.

              2. The problem I have with “flattening the curve” is that it overwhelms medical resources via exhaustion.

                I does the PRECISE opposite. The curve is almost vertical. (that means A LOT of more new cases per day). Flattening the curve means, literally, the same number of new cases every day.

                People can work 18 hour days — for a while, not indefinitely. A lot of other routine medical stuff (e.g. cancer surgeries) can be put off — for a while, not indefinitely.

                Which is WHY we MUST flatten the curve.

                The rest is untrue, especially if one knows what a sharp upward curve means.

                1. ” People can work 18 hour days — for a while, not indefinitely. A lot of other routine medical stuff (e.g. cancer surgeries) can be put off — for a while, not indefinitely.

                  Which is WHY we MUST flatten the curve.”

                  You missed or don’t understand the point. Dr. Ed is pointing out that spreading out the sickness, especially if it doesn’t significantly affect the total number of cases/deaths, can actually be worse because while people are actually pretty good at pulling on hidden reserves of strength and powering through extreme stress for short periods of time, that are not good at doing the same thing for weeks or months.

                  1. You missed or don’t understand the point

                    His premise is wrong.
                    He said flattening the curve INCREASES exhaustion.
                    From fewer patients!!

                    And YOU got his point backwards.
                    Spreading out the cases REDUCES the workload.
                    You’ve said that reducing the workload makes it more stressful. This suggests you don’t know what it means either.

                    Which is most stressful: 100 patients in a week? Or in a month>

                2. “Flattening the curve means, literally, the same number of new cases every day.”
                  Wrong. It means that the growth is sub-exponential, not that it is linear. It is not even linear in Korea, Taiwan, and Singapore.
                  Once one can lower the peak demand below existing capacity, the case rate will fluctuate non-linearly, hopefully at a low enough rate that there will be capacity of elective and non-critical medicine.
                  There is very good precedent to look at, namely the comparison of Philadelphia and St. Louis in the 1918 Spanish flu epidemic.

            3. While that delay has some value in the context of scarce medical facilities, it does nothing to change the position of that upper asymptote. Eventually the same number of people are going to get sick regardless.

              The value of the delay, “in the context of scarce medical facilities,” is that medical care has the potential to save lives—a lot of lives. So it always saves lives if you give a medical delivery system running at full capacity more time to do its job. Your comment seems self-contradictory.

              The assertion that, “the same number of people are going to get sick regardless,” leaves out the possibility of a vaccine.

              Please stop trying to justify Trump’s blundering, incompetence, and reflexive malice toward Americans who do not support him.

              1. You’re wrong on two important points, Stephen. First, there is no evidence that advanced medical care is actually very effective at helping people recover from this particular virus. There’s also little evidence that it’s ineffective. We just don’t have enough data yet. We can, however, say that it does not necessarily save “a lot of lives” and can be skeptical of claims that the isolation procedures save more lives that they cost through other channels.

                And, yes, I leave out the possibility of a vaccine because there is no reasonable chance that a vaccine can be developed, mass-produced and distributed in time to do any good at all. China, which has no vaccine, saw cases top out in about 2-3 months. South Korea is seeing cases max out at about the same time. Isolation procedures might stretch that time a bit but there’s no way they will be able to stretch to the year or more that vaccine development requires.

                1. I think Stephen is right. Even if “advanced medical care” doesn’t help, hospitals have to triage when they run out of beds, which means that some patients unnecessarily die. And when the number of patients shoots up the hospitals run out of ventilators, also causing unnecessary deaths. And hospital personnel run out of protective equipment, which means that more Doctors and Nurses die or get sick, which in turns means that patients get poorer care, causing more of them to die. That happened a LOT in northern Italy, and it would have happened a whole lot more if they hadn’t locked down the country. It’s happening in New York too, but fortunately the shutdown is limiting the damage.

              2. lathrop, a vaccine available for mass use is 12-18 months away, in the most optimistic of scenarios. Unless we want ‘right to try’ in this area which I would not favor.

                Stay safe, stay healthy.

                1. Testing begins, I think today, on a vaccine that uses antibodies extracted from those who have recovered. It works quite well for some viruses, not at all for others.

                  This could be a game changer on timing.

          4. but an economic collapse will cost a lot of lives also

            Not near as many lives

            I’m not sure this mathematically pans out. If technology is set back six months or a year over where it would otherwise grow to, that drag continues killing more year-over-year.

            1. What does that have to do with an economic “collapse?”

              If broadcast technology is set back six months or a year, from where it would have been, how would that kill people?
              Like 500 per day, on our way to 1,000?

              1. Of a cure for cancer is set back?

                1. What do radiation and chemotherapy have to do with economic collapse, “Doctor?”.

                  I’m fairly certain they’re both over a year old, since they cured mine over three years ago.

                  No offence, but might you be misspelling MR. Ed?

    1. It took 26 days to reach 1,000 deaths. (Feb 29 – Mar 26)
      The next thousand deaths in only TWO days (Mar 27 – 28)
      And that’s with the President’s current 15-day (mild) lockdown

      Growing from 1 to 1000 in 26 days is about a 30% daily growth rate, or 69% over two days. Allowing for error in these counts, it shouldn’t be two surprising to find it doubling in two days.

      1. You fail at what I mentioned. It is not a flat curve
        Each infected person infects 2.5 more. To keep it simple use 2.0

        1 adds 2 =3
        3 adds 6 =9
        9 ads 18 = 27.

        See?
        The growth keeps accelerating. Even wore because Trump dela

        1. See?
          The growth keeps accelerating. Even worse because Trump delayed for over a month. Did YOU believe it was a Democrat hoax.

          Plus, Trump was forced to reverse his irresponsible plan to reopen America by Easter. He’s admitted that 15 days is “not long enough” for all parts of the country, and is considering STRICTER measures for the most troubled areas.

          Does it not concern you, that he has reversed all of his initial positions, and sided with the professional, eventually?

        2. “You fail at what I mentioned. It is not a flat curve
          Each infected person infects 2.5 more. To keep it simple use 2.0”

          You do realize that there is a finite number of people in the world — don’t you? Or has Trump Derangement Syndrome so addled your brain that you don’t realize this?

          I’m not even going to try to explain the concept of susceptibility to infection, or that Corona-type viri (at least those not engineered in a laboratory) don’t do well in months that don’t have “Rs” in them.

          As an aside, is Trump Derangement Syndrome the medical equivalent of the Neutron Bomb?

          1. It’s rhetorically laying the groundwork to attack, regardless of the number of deaths. You can’t judge from these claims because they would be the same either way.

            I will quickly point out this is the result of Fox News balls-out behavior like this against Democrats. It’s sad CNN and so on just gave up and duplicated their behavior in the other direction.

          2. You do realize that there is a finite number of people in the world — don’t you?

            You realize we have these things called VACCINES — don’t you? For quite some time now!

            Do you get flu shots? That’s a VACCINE!

            The whole point of this is to reduce the number of deaths to the minimum possible, while new vaccine is developed and tested.

            Or has Trump Derangement Syndrome so addled your brain that you don’t realize this?

            I know what a vaccine is … so it’s kinda wacky to assume the transmissions keep doubling until the entire planet has been infected!

            I’m not even going to try to explain the concept of susceptibility to infection,

            You already FAILED on the easy part … vaccines.

        3. What did I fail?

          1. What I said. Your “averages” only work if it’s a flat curve,
            Your own words show the fallacy.

            Growing from 1 to 1000 in 26 days is about a 30% daily growth rate, or 69% over two days. Allowing for error in these counts, it shouldn’t be two surprising to find it doubling in two days.

            “If it grows at 30% a day for nearly a month, it should not be surprising if it suddenly jumps to 69%”

            And your math is wrong
            1000 at 69% growth for 2 two years = 2856, not 2000.
            Just adding .69 to .69 = 1.38 GROWTH where 1 is doubling,

            And 1 at 30% for 26 days = 705, not 1000.

        4. As the curve flattens, Ro declines. In Korea and China it is less than 1.0.
          Your suppositions are not based on any epidemiology with suppression measure being enforced.

  3. To “…where are we going?” I’d suggest “To hell, in this handbasket”. That’s because this planet can support no more than 1.5 million at a decent Western standard of living (according to e.g. Worldwatch.org, based on available land and water supplies).
    Global population already is ~7.6 billion or five times that. Thus, to “Will the daily surplus death rate substantially increase?” a reasonable, rational observer might well answer “Will it substantially decrease?” with “One hopes so”, and “How quickly?” with “Ideally, two centuries ago”.
    That’s why some hard science fiction treats Global Thermonuclear War as a desirable alternative future. That’s wrong, of course, if only because detonation of 100 reasonably dirty Big Ivan bombs would kill all life on Earth, but a more immediate objection is that this sort of calculus impelled the Nazi’s search for Lebensraum, with all the well-known consequences. Hoess of Auschwitz described a Jewish mother and her five-year old boy confronting him with this moral dilemma; he persisted in what he perceived as his duty, though coming to repent in the days before his execution. I can only recommend: Duck, and Cover.

    1. Whatever the antonym of aptronym is, you’re that.

    2. “this planet can support no more than 1.5 million at a decent Western standard of living”

      The problem with that fallacy is that we would not even be able to feed 329 Million Americans were we still using 1920’s agriculture technology. Land resources are based on water resources and have you ever heard of something called “desalinization”? Yes, it’s a fledgling technology that isn’t quite here yet — but look at where your smartphone was 40 years ago…

      We didn’t really have refrigeration a century ago — much of the country didn’t even have electricity….

      1. Exactly.

        Why material resource shortage impact predictions are always so wildly wrong.

        In a free country, people come up with amelioration and substitute solutions faster than the issues become serious, and costs continue to drop and plenty continues to increase.

        There are ups and downs, but the long term trend is down, always down.

      2. The point is a good one, explored in other hard Science Fiction works such as Flint’s “1632” and especially the associated tech manual. But all such arguments fail on two crucial points.
        First, we have long since passed the point of diminishing returns on the application of food technology to a finite resources such as arable land. For example, Desalinisation as a technology I am somewhat familiar with, but it is barely capable of meeting the drinking water needs of a population, let alone the much greater (by several orders of magnitude) needs of agriculture.
        Second, the laws of thermodynamics get in the way. To apply some of these technologies (hydroponics, desalinisation…) requires applying vast amounts of energy having serious thermodynamic inefficiency – with predictable consequences for global warming.

    3. We’re already supporting more than 1.5 million people at a decent Western standard of living, and we’ve been doing this for a century in just one country. Include the UK and we push that back even further.

      1. Indeed, and the figure should be 1.5 billion, not 1.5 million. I corrected this and other small errors, but the Reason website did not seem to accept the submission. I must have clicked the submit button several times without apparent result, till on checking some time later it became clear that each “Submit” command was interpreted as a separate comment! I do apologize.

  4. The numbers of infected (and deaths) we are seeing right now are mostly people who were exposed prior to the 15 days to slow the spread guidelines were put into place. We can expect things to worsen for the next few weeks before we see improvement.

    What I do not see discussed is the impact and cost of 5% of patients who need extreme measures (like a ventilator) for a prolonged period of time, and then go on to recover. This will be a tremendous burden on the healthcare system and the economy.

    1. You flunk math, Bigly.
      The 15 days began March 13. The longest incubation is 14 days.

      Yes, on the ventilators, and Trump still refuses to take a leadership role on this for the states, which are filled with people. Instead, he whines they aren’t nice enough to him, as he screws them over.

      1. Liberty, I thought the start date was Sunday, March 15th or Monday, March 16th. I think we are both close. Even so, I don’t think that changes my points.

        – We can reasonably expect things to worsen over the next few weeks before they start getting better

        – There is a tremendous burden and cost by the 5% of Wuhan coronavirus afflicted patients who require extreme measures and then go on to recover. This is a gigantic cost. That makes it quite different than the flu, by itself. Not even considering the mortality rate.

        There is a time and place for Orange Man Bad. This really is not the time. We are all – regardless of political viewpoint – engaged in an enormous national effort to overcome the Wuhan coronavirus. That really needs to be our focus. There is plenty of time afterward for the political bullshit and carping from the sideline.

        Right now, it is an American problem. There is no Team D or Team R crisis; there is a American crisis. We all need to pull together and deal with this. Then help the rest of the world.

        Stay safe, stay healthy.

        1. When he announced the 15-day program is when he finally assumed leadership. Kinda.

          If I accept your dates, then he delayed assuming leadership even longer, right?

          Right now, it is an American problem.

          Do you deny that requires an American leader?
          Did we not elect a President to do that? YOU make it Orange Man Bad, to deny all his many errors here, and the resulting needless deaths. (I mean the collective you)

          There is no Team D or Team R

          How is that even remotely relevant to what I said? And it’s MY POINT!
          It was TRUMP who made it partisan with his bat-shit crazy claim that it was a Democrat hoax. Did YOU believe Democrats sneaked over to Wuhan, and infected people? Then infected that guy who brought it from there to Seattle?

        2. By February 25, there was a cluster of 53 cases in the US reported to WHO. The first 2 deaths were reported on March 3 based on 62 cases. That number had doubled by March 5. The next doubling took 2.5 days; the next, 2.5 days. The doubling time increased to 3 days by March 15
          (WHO statistics)

      2. “The longest incubation is 14 days.”

        Are we sure about that?

        I remember when we thought that Haitians had a special vulnerability to AIDS — and a few years later when we learned that wasn’t true.

        1. 14 days is the longest known incubation, 5 days the shortest.

          1. We don’t even yet know what percentage of the population has it.
            We’re about where we were with AIDS in 1982 — and we’re already testing a vaccine — that’s damn good and Trump should be credited for that.

            We THINK 14 days — but were you the person who deliberately infected the person? Of course not — and all we can do is make an educated guess as to when the person got infected.

            An educated guess….

            1. Diversion.
              Irrelevant

              and all we can do is make an educated guess as to when the person got infected.

              Then you need to educate yourself.

              They trace an infection back to contact with an infected person. This has been WIDELY reported all along.
              Ted Cruz and others in contact with the carrier at CPAC.
              Rand Paul.
              Donald Trump.

              Of course, we hear more about the ones who test negative, but that shows us how the connections are tracked.

            2. We know that the average time to symptoms is 4.5 days. I have never seen a report of the variance of that number. But is is not credible that the distribution function abruptly terminates at 5 sigma. Most MDs are saying 3 to 14 days

  5. So, I ran these numbers a while back. One common comparison is the Spanish Flu of 1918, and a comparison of the mortality rates in 1918 in the US compared with 1917 and 1919.

    In 1917 the mortality rate was 14.3 per 1,000. In 1918, 18.1 per 1000. In 1919, 12.9 per thousand. Roughly speaking the Spanish Flu increased the death rate by 4 people per thousand (or ~30% increase in the death rate).

    https://www.cdc.gov/nchs/data/vsushistorical/mortstatsh_1922.pdf

    How is that comparing to the current death rate increase under the Wuhan Virus? The annual death rate in the US is currently ~8.8 per thousand. We’ve seen a 2229 people die of the Wuhan virus, out of a total population of over 300 million. That’s an increase in the death rate of 0.007 deaths per thousand. Assuming every Wuhan Virus death wouldn’t have happened otherwise.

    Now, these are national annual death and death rates, not broken down by month and day and region. But at the end of the day, the total death rate in the US may go DOWN in 2020. To give an example, the typical US motor vehicle fatality rate is about 11 per 100,000 people (33,000 per year). This lockdown is likely effectively dropping traffic on the roads (and the resulting fatalities) by 50%. On the other hand, other causes of death (suicide) may increase.

    1. Ummmmm… your numbers include all the flu deaths for 1917 and 1918. But, unfortunately, I don’t think the current covid numbers include all the 2020 covid deaths.

      1. True, I don’t include all 2020 Wuhan virus deaths (for obvious reasons, like 2020 not being completed). Some extrapolation is needed. How one performs that extrapolation, is of course critical.

        But what I wanted to do here was put the death rate in perspective, in comparison to both past historical data, and in current accident death data.

        1. It’s profoundly innumerate to compare three months to two years, when when know the average daily deaths keep increasing, rapidly

          For this, the average deaths were 4.7 for the first 16 days.
          The next two days averaged over 500 deaths each.

          1. Again, we are limited to the data we have. And it’s important to use accurate data, which yours doesn’t appear to be. Assuming you’re using the deaths in the US prior to March 26th, it’s an average of 80 per day. (IE, 1295 deaths divided by 16).

            But we have 3 months of data, not a year’s worth of data. And as I noted, some reasonable extrapolation should be done. Emphasis on reasonable. We have other countries we can look at, for example South Korea or Italy, in terms of curves and expected extrapolations. It’s not an exponential increase in deaths per day, it levels off. If you wanted to increase the total number of US deaths by a factor of 10 for the year (IE, 20,000 deaths) that may be reasonable, based on existing data. Depends.

            But, it’s important to keep that number in context to existing deaths via other causes.

            1. accurate data, which yours doesn’t appear to be. Assuming you’re using the deaths in the US prior to March 26th, it’s an average of 80 per day

              Non-responsive. Read it again.

              But we have 3 months of data, not a year’s worth of data

              Which is MY point. And you IGNORED that these diseases escalate as more infections cause more infections.

              Ooops, I had a typo. Shoulda been (using round numbers)

              1000 / 26 days = 38.4 per day (not 80)
              1000 / 2 days = 500 per day.

              THERE we can see the escalation.
              If we do the entire period, as you did with swine flu
              2000 / 28 days = 71 per day — totally useless for comparing with a much longer period of time.

              Swine flu would have also escalated, but at a different pace.

              1. “Oops I had a typo”…

                Multiple typos. 16 to 26…. 4.7 per day to “38.4” per day…. And still using the wrong number of deaths to get there (IE, not 1000 but 1295). Your math is all over the place.

                And you’re failing to account for reasonable extrapolations.

                Until you can make a reasonable argument with real math, there’s no point to discussing things further. Please inform yourself better.

                1. Multiple typos. 16 to 26…. 4.7 per day to “38.4” per day

                  16 was the typo, which means the RESULT changes!

                  Your math is all over the place.

                  Here’s what YOU said

                  deaths in the US prior to March 26th, it’s an average of 80 per day. (IE, 1295 deaths divided by 16).

                  The first death was February 29th.March 26 was the 26th day, not the 16th day,

                  Repeat:
                  The first 26 days averaged 38.4 per day.
                  The next two days averaged 500.
                  That escalation is why you cannot compare three months with two years.

                  True, I don’t include all 2020 Wuhan virus deaths (for obvious reasons,

                  The vast majority of those deaths haven’t happened.

                  Mine was a typo.
                  That’s four failed “corrections” on the page.
                  Will there be more below?

          2. What date did you start counting?
            Ten days after the first cluster (53 cases) the death rate was 3%.
            Since then it has varied down and uo between 1% and 1.8%
            (WHO statistics)

    2. Yeah Armchair, you’re asking good questions we are all asking, in one form or another: Just how bad is this Wuhan coronavirus? The dataset we now have is in excess of 1MM observations, and within the next 48-72 hours, we will be able to make zip code based population assessments (if we cannot already). I figure there are roughly 40K zip codes, divided by the 1MM+ observations. Pretty sure we have at least a couple hundred observations per zip code by now (or will shortly)*.

      *shortly: The recently passed CARES law requires commercial labs to report data to the Feds. That data will come on line in the next 48 hours. Prepare to be surprised.

    3. In 1917 the mortality rate was 14.3 per 1,000. In 1918, 18.1 per 1000. In 1919, 12.9 per thousand. Roughly speaking the Spanish Flu increased the death rate by 4 people per thousand (or ~30% increase in the death rate).

      1918 was also a war year — by the Summer of 1918, we were sending 10,000 soldiers per day to France. Many died there, many died later from war wounds (notably lung damage from poison gas).

      Hence while the Flu killed more Americans than the war, doesn’t the 1918 mortality rate include both? And I imagine that having Chlorine-damaged lungs didn’t help one fight off the Flu…

      1. Crowded military camps certainly helped spread the infection – and then there were the mass troop movements. After the war, troop ships coming home brought it back to the USA.

        1. Oh, it’s worse than that. Troops came home to Boston in the midst of the Boston Police Strike so they were all thrown in jail which didn’t help, and then the military housed them in massive rooms so one person infected the whole bunch.

      2. 1918 was a war year. But American deaths due to WWI were relatively modest (53,402 deaths to to combat). In fact, the US military lost more soldier to disease (63,114)…largely due to the Spanish Flu.

        On the other hand, the US population in 1918 was just over 100 million. At a death rate of 4 per thousand in excess of the normal death rate of the time, that’s an extra 400,000 deaths. (Conventional research puts it at 675,000 deaths due to the Spanish flu in America).

        Point being, US deaths in combat due to WWI were an order of magnitude less than US deaths due to the epidemic.

  6. I hate to complain, because I haven’t done any such analysis myself, but your methodology seems underwhelming. To actually answer your question you would need: What’s the current weekly or monthly mortality rate, compared to the historical weekly or monthly mortality rate, and it would be even better if it were temperature adjusted.

    I did see someone do something similar for CDC data up to the 10th week of the year, which of course was before the recent explosion, they found total mortality and pneumonia significantly lower than normal, but they don’t link the source of their data.

    It’s from Sara Hoyt’s blog who is a successful science fiction writer, so it’s not just a rando, but I’m not sure what to think:
    https://accordingtohoyt.com/2020/03/27/covid-19-and-us-mortality-by-i-ratel/

  7. Projections of deaths are incomplete in one important respect. If we reach a situation where medical facilities are overwhelmed there will be other deaths, from heart attacks, accidents, etc. that could have been avoided had ER capacity been available.

    In measuring the impact of the coronavirus on mortality, these incremental deaths should also be included, even if the victims did not die of the virus.

    1. Yeah, you’re right but that is incredibly hard to estimate accurately (too many confounding variables) = In measuring the impact of the coronavirus on mortality, these incremental deaths should also be included, even if the victims did not die of the virus.

    2. That’s part of what’s really lacking in the numbers. There’s been anecdotal reporting of hospital capacity issues and supply shortages, but in the end, isn’t the whole point of tanking our economy to forestall “health care system collapse”? And if so, shouldn’t the reporting be on:

      # of COVID-19 ICU patients
      # of open hospital beds
      # of open ICU beds

  8. Here is the most recent update to a page updated daily. I have a previous comment with two links which is “awaiting moderation”, so let’s see if a single link gets through immediately.

    1. Two links will result in that.

      Also, I’d argue that this is a far more trustworthy source of information:

      https://coronavirus.jhu.edu/map.html

      1. Trustworthy? In what context? All that page I linked to does is convert official statistics into pretty graphs. How is that untrustworthy?

        In addition, your site, pretty as it is, doesn’t seem to show anything except current totals — no trends to help gauge progress.

        Methinks you favor pretty “official” facts over any kind of individual post-processing.

  9. Other sources tell me that Japan has not really been that successful, that they are hiding their death toll because face-saving.

    1. Links, please.

    2. How can you say that?
      The case rate in Japan has been a steady exponential with a doubling every 4.5 days since Feb. 11.
      The mortality reported is now 3.5% and has been rising steadily since 13 days after the first cluster of cases.
      (WHO statistics)

  10. Was reason.com undergoing maintenance last night? At various times reason/volokh became completely unreachable, then would respond for a few seconds, then a blank page again. Strange.

  11. Of course, especially since the dying are mostly the elderly or chronically ill, a much higher than usual death rate now will probably mean a considerably lower than usual death rate after this epidemic passes…

    By that yardstick, abortion is good because it lowers the infant mortality rate (of those who actually get born).

    1. It’s not “good” – it’s just something that has to be taken account of in the math. A different way of saying it is that we should not be measuring success based on ‘lives saved’ but on ‘life-years saved’.

  12. here is the mortality owing to coronavirus, which pretty much adds to that daily death rate

    How do you know it adds to the daily death rate, as opposed to replacing other causes of death?

  13. Change the rule, get different measurements. The ‘best’ data is a change in historic death rates between various countries. Sweden is going to be interesting because of herd immunity. CDC COVID19 Mortality Reporting Guidelines: “COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death,”: https://www.cdc.gov/nchs/data/nvss/coronavirus/Alert-2-New-ICD-code-introduced-for-COVID-19-deaths.pdf

Please to post comments

Comments are closed.