Spain Coronavirus Death Rate Over the Last 2 Days Comparable to Italy's, 40% Addition to Usual Total Death Rate

|The Volokh Conspiracy |

According to Worldometers, Spain had 391 coronavirus deaths yesterday and 539 today (recall that it's tomorrow over there now). [UPDATE, March 24: Today's death count was 680.] Italy's has averaged 668 over the last 4 days, 625 over the last 2. Since Spain's population is about 3/4 that of Italy, Spain's death rate has been pretty close to that Italy's.

Spain's death rate last year was 9.115/1000, which amounted to about 425,000 deaths per year (out of a population of 46.66 million), or 1165 deaths per day. This means that Spain's death rate over the last two days apparently increased by about 40% as a result of coronavirus.

Of course, keep in mind all the usual cautions about possible diagnosis error, reporting error, and the like; plus, while the coronavirus death rate may go up or may go down, the one thing it's not likely to do in the next several days is stay the same.


NEXT: Is privacy in pandemics like atheism in foxholes?

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 or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. I don’t think you can assume that the Coronavirus deaths and other deaths are additive. For example, it has been said that flu deaths are a major cause of death. How many people that died of Covid-19 would have died of the flu give or take a few days? Same goes with heart disease, cancer… A better approach would be to look at total actual deaths now and compare them to expected deaths extrapolated from previous years.

    1. What are the odds that the people who died of coronavirus in the last two days would have died of heart disease, cancer, or the flu in those two days, as opposed to months or years later? Sounds like these really are additional deaths — though I appreciate that, given the apparent demographic of those who have the highest fatality rate, the Years Lost for this epidemic may well be lower than for some others.

      1. It doesn’t have to be one or two days apart. Somebody that was destined to die from cancer today might have died of Covid-19 a couple of weeks ago, so there is one less cancer death today.

        The point is that detailed analysis may show (and probably will show to some degree) that non-Covid deaths went down during this period in Spain and in the months after the pandemic is over.

        1. We all die sometime, who gets bent out of shape about shifting the timeline around a little?

          1. Understanding the effectiveness of policy at preventing, postponing, or prolonging death is utterly critical to marshaling finite resources in the middle of a crisis. How do you propose we begin triage unless we understand shifting timelines? Are we going to give David Lat’s ventilator to a 97 year old man who was just diagnosed with stage 4 colon cancer? Note that the all-cause mortality rate (per year) of an average 97 year old man is >30%.

            In a world with finite resources, prioritizing resources to extend the the life of a person who has months to live, by shifting resources from people who have years to live, is pointlessly cruel and immoral.

            1. I dunno man, that’s sounding like death panels.

              Also, pretty sure triage is already something well-practiced among medical professionals already; not something we need a national guidance on.

              Certainly there’s some countervailing effect, but other than noting it, I don’t think we should essentialize it.

              1. Triage can mean the decision a doctor has to make in real time at a hospital. It can also mean the broader policy of the extent to which we want to kill the virus versus kill ourselves. One way to destroy the virus is to send out humvees with kill squads to shoot anyone who leaves their homes. Presumably that would go beyond what even you believe is prudent to stop the spread. On the other end is do nothing. There are a million places in between.

          2. The entire world gets bent out shape about shifting the timeline if it shifts enough.

            If you ignore that, then nothing that kills people matters.

            1. Yeah, I was sarcasting. This whistling past the graveyard lets restart the economy nonsense is pretty dumb.

              1. Yes, but at a societal level we do have to make cost benefit decisions. Restarting the economy right now is stupid not because some people will die but because the CBA almost certainly comes out against it. In twelve months, that calculus would almost certainly be reversed. Where the tipping point is, I have no idea. Neither does anyone else, but they have to try to figure it out.

                TL;DR: “If it saves one life, then it’s worth it” is a terrible argument.

      2. I think Freddy’s point is valid, taken over a time period.

        Suppose on March 1 we have 1000 people with some advanced cancer, who will likely die during the month. One hundred of them die from Covid-19 instead, as do another hundred who were healthy on March 1.

        So, for the month the death rate is 10% more than expected, not 20%. OTOH, since the Covid-19 deaths occur sooner than expected the days of life lost is much more than 10 or 20%.

        1. If we actually knew total deaths per day, we could compare that to last year’s daily average and, assuming that Covid-19 is the only real shock, that would tell us what the increase due to the virus is. But I think it has to be done over a longer period of time than two days, because of day-to-day variation in “normal” death rates.

    2. How many died of BOTH this AND the flu? And which actually killed them?
      I find the initial high negative testing results (of people whom a MD thought had the virus) to be interesting — they all had something else.

      It’s like if a tree falls on a drunk driver, the fatality is listed as OUI while the accident was actually caused by a few tons of snapped-off Pine tree falling out of the sky. (True story)

      Italy & Spain are listing fatalities where CV19 was found, which is not the same thing as saying it killed them, particularly when cause of death is usually either fluid in the lungs and/or heart attack caused by that. It’s like if someone is both shot & stabbed — death is caused by loss of blood, but which actually caused it? Or did both?

  2. Demographics and culture matter. In italy (and also Spain I think) there are a lot of multi generational households. Young people are relatively asymptomatic and give it to old people. Most deaths are older (80 yrs +).

    The same demographics are playing out in NYC- the mortality is much higher for older people than the flu.

    1. In addition to age, add smoking-related lung issues and cardo issues.

      It’s like driving in the snow on bald tires — that definitely doesn’t help….

  3. As WorldOMeters notes, calculating the death rate while an event is ongoing presents some challenges, but the easiest acceptable method is to divide (the total number of known deaths) by the sum of (the total number of known deaths plus the total number of known recoveries). Italy’s rate of 44.98% (at this moment) is slightly higher than Spain’s 40.78%, but that is really close. None of it compares to the death rate due to HIV, though… and I’m amazed that nobody really notes that 1 in 25 people (40,000 ppm) on the African continent are coping with HIV infection.

    1. You can’t do that because you don’t know the total number of infections.

    2. the easiest acceptable method is to divide (the total number of known deaths) by the sum of (the total number of known deaths plus the total number of known recoveries).

      That depends on whether those tested are a random sample of those infected, doesn’t it?

  4. Let’s say that 50% of all pregnancy tests are positive — that does not mean that half the women in the US are pregnant. No, it’s half the women who took the test — something very different.

    The death rate (US) for those tested for the flu is 3% while the actual death rate for it is 0.1% because most people weren’t tested. Only the sickest were.

    Unless you test EVERYONE (or a statistically-valid sample) for CV19, you can not determine a mortality percentage from it.

  5. Professor Volokh,

    You are comparing a daily rate for a couple of days to a yearly average.

    How much does the daily rate vary from day to day. It may vary a lot. A particular day or couple of days might easily be 40% above the yearly average by chance.

    Moreover, there may be a seasonal pattern. Perhaps the daily rate goes up in late winter/early spring.

    As a law professor, and one interested in technology and law in the modern world, you should have enough quantitative literacy so you don’t take this sort of information uncritically, at face value, and automatically assume it means something major.

  6. Its fascinating (and disturbing) watching Reason commenters bent over backwards to explain what’s going on as “normal” and they would of “just died of something else”.

    1. Maximum capitalism plus tribalism leaves precious room for reality, it seems.

      1. Reality is that some human beings will die miserable deaths because of COVID mitigation measures, and some humans who won’t die miserable COVID deaths due to mitigation measures, would have otherwise suffered near-term miserable deaths, anyway. It is annoying that there are adults pretending to find this wholly ordinary social utility calculation “distrubing”. We should be focused on minimizing human misery.

        The tribe at issue here is humanity. “Maximum capitalism” (however defined) is a tool for helping that tribe. It’s either a useful, or not useful, tool. But it shouldn’t be derided if it is a good mechanism for dealing with finite resources.

    2. Well you really do have to consider that according to the official statistics the percentage of people in the world with covid-19 is .00000053%. In the U S its 0.0000016%. And of course the death rates are a fraction of that, in the US 2 people per million population. In Spain its 59 per million or 0.000059%.

      I’m certainly not saying its mindless panic or a hoax, it is panic over the potential.

      But keep in mind that nowhere has the potential actually been realized to a truly devestating pandemic, unless of course you think my numbers are wrong, and your welcome to correct them.

    3. Many DID die of something else.

      1. Along this vein, i just saw a news article the other day which was a perfect example of a bait-and-switch on covid. Title was something like ’19 year old with covid19 dies in Britain’. Article mentions he tested positive, he had other health conditions, and then, second to last sentence (and last substantive sentence), that he died from something other than covid19. If it wasn’t covid19 that killed him, why did the headline emphasize covid19, and why did you wait till the end to mention actual cause of death?

        Want to bet he got counted as a covid19 death in the official statistics?

  7. I shouldn’t do math at 3am. Actually my math was fine, my conversion to percent was off. 0.0053% of the world has gotten the Wuhan virus. 0.016% of the U.S. has been infected, and .0059% of Spaniards have expired from it.

    But m point I still intact. The panic is in the potential, not the current reality.

Please to post comments