« Open Mike II: It Only Makes Sense | Main | Punctum Def »

Tuesday, 19 March 2024


Feed You can follow this conversation by subscribing to the comment feed for this post.

So you're saying that it's safe for me to leave the house? No, I'll just pretend it is still dangerous and keep my introverted lifestyle going. No laugh emoji because it's more true than joke.

Comparing numbers like this makes not very much sense though it is often done.

If I live in a village with 100 inhabitants and 50 die in a year, I notice this (and I have a 50% chance of dying in that year other things being equal). If I live, I start to look for a new disease, a serial killer, or probably Putin.

If I now move from my gab to a city of 10 million inhabitants and 50 of them die in a year, well again I notice, but this time I look for medical miracles, aliens or very poor record-keeping, because I would expect about 10,000 people to die each year from this population.

So, for example, using data from The Economist (here: https://www.economist.com/graphic-detail/coronavirus-excess-deaths-estimates, no subscription required for this), official US death toll from covid to 19th March is 1,179,342. Real value is probably 1,300,000 to 1,500,000, about 20% higher. US civil war deaths I will take as 750,000.

But but. US death rate per 100,000 from covid is official 346.3 probable actual 380 to 450. Civil war deaths per 100,000? well assume a population of 35,000,000 (census in 1860 was 31,443,321, in 1870 was 38,558,371) and it was about 2,140.

So the US civil war was about five times more deadly than covid in the US: you were five times more likely to have know someone who died in the civil war if you lived in 1865 than of covid today.

This has been a public service announcement by the gypsy mathematical service.

Yes , but does long Covid also come in waves, or is it long enough on average that the active cases monotonically increase, so that at some point all of humankind will be crippled (extreme scenario, admittedly). Moreover, will people who encountered the virus early in their lives do better than those who were older at the time the virus emerged (wet market? lab? we’ll never know) in 2019? If the young exposures do better, Covid may exacerbate the ills of societies where fewer and fewer active young people support a growing cohort of disabled old. Covid is less than five years old, so despite the unprecedented (no hyperbole) speed and success with which vaccines and Paxlovid were developed, projecting that it is going to be progressively milder is wishful thinking. It might become another irritant, or it might remain a major cause of morbidity. The pathogen certainly is morally indifferent, so it will keep killing and crippling preferentially those who ignore precautions.

The USA population in 2018 was about 103 million people, so about half of 1 percent of the American population died of Spanish Flu.

And now, COVID in 2024, as of last week per CID here, has killed 1.184 MILLION people out of a total population of 332 million. Those deaths are the entire population of cities like San Jose, San Diego, or Dallas. All of these are in the top 10 cities ranked by population. Imagine that, one of the biggest cities in America, with everybody dying. Over 4 years that is 24,600/month.

Or another way to think about it, COVID killed a third of 1 percent of the population, and that's WITH our knowledge base about disease, and the treatments available. And on the flip side, that's WITH the anti-vaxx, anti-mask, anti-public heath measure idiots.

Of course, deaths in the Civil War were much greater as a %age of the population....

As for the lower numbers, I'm not sure the virus is becoming that much less fatal. More likely, I think, is that thanks to vaccination and infection/recovery, we're now much better equipped to fight it. Well, those of us who survived are better equipped - it's also the case that the first wave or two in 2020 killed a high proportion of those who were most vulnerable. In fact at the moment we might be at the sweet spot - those of us who survived haven't got much older and therefore become significantly more vulnerable due to advancing age, and there aren't many babies and young children who, having neither had the disease nor been vaccinated, have little immunity. But give it another ten years and the picture will have changed.

I immediately had the same question as Gary regarding the relative impact of the numbers by knowing the percentage of the population impacted. I believe the US method was to attribute "cause" of death to covid if the person tested positive for covid. In that elderly population that was suffering the greatest mortality, their "actual" cause of death was probably most often not covid. Like most things, it seems these statistics and numbers can be bent, twisted and manipulated to say almost anything you want?

It would be interesting to know, though I suspect the numbers will always be kept hidden from us, the percentage of people who died after being vaccinated and the percentage who died unvaccinated.

Dear John,

They could be... but careful cross-checks have shown that they weren't! There were both overcounts and undercounts, but those have been accounted for by back-studies of the data, and we have pretty solid numbers.

A good and easily-explained example of such a cross-check is analyzing "excess deaths." From quarter to quarter, one can make a pretty good estimate of the number of people who will die in the US from all known causes. It varies with the season, and it can vary with circumstances like an especially good or bad flu season. The accounting takes that into account (he said redundantly).

What you can look at, then, is how the reported deaths due to Covid match up with the excess death numbers. When that is done for the United States, it looks like we've underreported Covid deaths by 10–15%, which is really very good! Some countries are notorious for massively undercounting, like by more than 50%!

(It's worth mentioning that the US numbers for the first year of the pandemic very badly undercounted Covid cases. If you looked at the "official" Covid case numbers versus conservative death numbers, you'd conclude that Covid was 5–10% fatal! This is where some of the wild, apocalyptic reports that appeared in the popular press in the first year came from. Medical people KNEW this was wrong, because only a fraction of the people with Covid were hospitalized and way fewer than 5 to 10% of them died. We were pretty sure that the actual death rate, pre-vaccination, in the 0.5–1% range. More data with better statistics and back analysis confirmed that was a good first guess.)

I think that satisfactorily answers your question about whether the number of deaths have been significantly exaggerated.

There are other ways to cross-check the reported numbers and they all say that it's good (within 10% or so) but explaining those would require me to go way into the weeds of biomedical statistics. longissimus, non legi!


Dear Stephen,

Why would you think that?! That data is out there, it's very well-known, has been from the beginning—nobody's been keeping it secret, not even close! Relative death rates were a key part of testing the efficacy of vaccines. After that doctors and researchers tracked whether the people who got sick and died were vaccinated or not, because we needed more field data after fast-tracking the vaccines.

Vaccination massively reduces the death rate, even with less effective vaccines, easily by a factor of five.

It also decreases the incidence of long Covid by at least a factor of two, although that number is very soft because it's still hard to get good statistics on that. We're getting better.

(please excuse any word-salad. Apple Dictate's fault)
pax / Ctein
-- Ctein's Online Gallery http://ctein.com
-- Digital Restorations http://photo-repair.com

There are still people who call Covid a hoax. There are still people who think vaccines are dangerous.

(Tongue-in cheek warning) When an anti-vaxxer steps on a nail, do they get a tetanus booster?

re: "known to history as Massasoit"

That word evolved to be the name of the region where King Philip lived: Massachusetts.

re: "the virtual extermination of the Narragansett"

This word lives on as a town in Rhode Island, a nearby bay and a beer that was once more popular than Budweiser and has recently been brought back as a boutique brew.

[Actually there was a tribe called the Massachusett in Southeastern present-day Massachusetts South of Boston and North of Plymouth, and the tribal name is the origin of the State name. And today there are about 5,000 Wampanoag descendants, many of them on Martha's Vineyard. According to the Narragansett Tribal website, there are currently about 2,400 Narragansett descendants on the tribal rolls, most of them residing in Rhode Island. To be a tribal member, candidates must be able to trace their ancestry to the 1880-84 Roll, taken when the Narragansett tribe was "detribalized" by the Rhode Island State government. Tribal rolls are currently closed. --Mike]

Per Robert Roaldi's TiC question, "Fountainhead" author Ayn Rand collected Social Security late in her life. And a good friend once declined a lucrative job offer from the Christian Scientists because... they had no health care program. So there we are.

Thank you Ctein - I think that is a very fair assessment and answer to my question. I thought from very early on that the more valuable number was the excess death total which seems to remove much of the political and societal noise. Another thing I always wanted to consider with Covid was estimated lost "life years." In no way do I want to discriminate against the old, but as a society we were very fortunate that this disease was not very lethal to the young. If the average age of death from Covid was very close to the average life expectancy of 77.3 years, very little "life" years were lost. But when a 5 year old dies of some disease, they have lost 72.3 years of "life." I don't know if I have seen a statistic for this because it may seem somewhat cold, but it is real.

Dear John,

I'm glad my brief answer was helpful. I didn't want to get too deeply into the weeds, because I know way too much about this. I needed to get so good at the epidemiology in the first year that I wound up being able to *quantitatively* calculate the risk of infection from a certain activity during 2020.

I didn't find your question about age to be "cold;" it's a very important socioeconomic question about what's the impact of this on society as a whole. That's a whole lot different from the deplorables who were saying, "Well, it's just the elderly. Who cares if they die off!" The nicest word for them is despicable.

I don't know about recently, but that question was investigated in the first months of the pandemic by the econometricists at the University of Chicago. I should mention, not in passing, that they have a top-notch economic school, and I have huge respect for their competence, even though I deplore their politics, which run heavily libertarian (they worship at the altar of Milton Friedman). I mention that in case people might think they were biasing the results, consciously or unconsciously.

Early on, they ran an econometric model of what the impact would be of "opening up" — removing all restrictions on activities and letting the epidemic run its course, as the alt-right types were pushing for. They included age, economic status, and sex in their partitioning of the data and (knowing them) probably a whole bunch of other factors that I'm unaware of. Those people are GOOD!

Their conclusion was that if we removed all controls, in a matter of months 1.5 million people would die directly of Covid before the epidemic burnt itself out (yeah, I'll get back to that). But even more than that would die of other causes as a result of the healthcare system being overwhelmed and crashing.

The economic cost of this? Approximately one GDP! That wouldn't all hit in the first year, of course, because it was analyzing future impacts. It would be spread out over a decade, but it would be heavily front-end loaded. So, maybe 25% GDP lost in the first year? Somewhat less than the second year, and declining further after that.

Short form: a Great Depression.

I bet this went a long way towards silencing the "open it up" folks. Even the wouldbe-dictator-in-chief couldn't ignore that.

They did a good job with the data they had, but from what we've learned later their assumptions were overly optimistic. For one, they ignored the impact of Long Covid, which was affecting 20–40% of the infectees. The data was too sparse to be quantifiable at that point. For another, it assumed that people could only get infected once, and then they would be immune. Boy, did that turn out to be wrong!

With good politics and good policy, we could've kept it down to maybe two hundred thousand deaths (not much less without fabulous good luck), but had we gone the other way we'd now be looking at about 3% of the adult population having died of Covid.

So, y'know, as we're all thinking we had it really bad and things could've been so much better, there are some alternate timelines in which the folks are saying, "Oh my God, they got off easy!" A sobering thought.

(please excuse any word-salad. Apple Dictate's fault)
pax / Ctein
-- Ctein's Online Gallery http://ctein.com
-- Digital Restorations http://photo-repair.com

The comments to this entry are closed.



Blog powered by Typepad
Member since 06/2007