Hacker News new | past | comments | ask | show | jobs | submit login
[flagged] 1 in 3 Americans know someone who died from Covid-19 (axios.com)
36 points by cwwc 13 days ago | hide | past | favorite | 71 comments

I know a few people who've gotten it and recovered. I know one person who got it and died. He was a healthy 48 year old mountain climber. A statistical anomaly, though that is of no comfort to his family. They went through hell, he was in the hospital for a couple months before passing.

That’s unfortunate, the 45-54 age range has less than 1/6th the risk of death of a 75-84 year old, but that’s a long way from zero.

This is a pretty useless metric for anything other than shock value.

Yes, it's just about finding a way to come up with the largest and thus most shocking number. It's sensationalist but not useful.

The US population is 330 million and 500k people have now died of Covid. So for 1/3 of all Americans to know someone who died of Covid each victim must be known by 220 people on average. Now, a quick look at Google suggests that this is within the range of the number of people known by the average American (and maybe on the low side) [1] so saying that 1/3 of Americans 'know' someone who died of Covid really does not convey any new information.

[1] https://www.nytimes.com/2013/02/19/science/the-average-ameri...

Edit: Ahah I see you did the same calculation in another comment!

It is useful when considering the social effects of the pandemic.

What social effects does this inform us about?

That there'll be a lot of people with varying levels of grief, loss, and trauma. This may have some long-term impacts on psychological health that we should be prepared to mitigate.

Grief, loss, and trauma are hardly specific to covid-related deaths. If that's our concern, we should really be considering overall mortality rates (looks to be about 10-15% higher than previous years) and how many Americans know someone who has recently died, regardless of cause.

I'd argue that mass casualty events have some specific characteristics that differ from the normal background noise of deaths that happen on a routine basis.

9/11, for example, had a pretty disparate impact versus the raw number of deaths it generated.

Sure, but the impact of a mass casualty event like 9/11 is not limited to people who knew the victims. Even though COVID-19 kills more Americans daily than the 9/11 attacks did, 9/11 has a comparable social impact because to most Americans it was an unanticipated, unprecedented act of human violence and we didn't know what to expect moving forward.

1 in 3 people know someone who died on any day of any year. Number of deaths of Covid is not that high.

Non sequitur -- a new single cause that is equivalent to the combined sum of all causes of death in a typical year is a "high" number by any definition.

I'd think it's pretty universally agreeable that the COVID pandemic has has some social effects. Step one to understanding the size and scope of social impact is to start measuring it. This is one of many measurements that may lead us to some conclusions.

It's a useful metric to toss back at the Just the Flu crowd.

Not really. If you want to convince people that COVID-19 is significantly different the seasonal flu, all you have to do is compare COVID-19's ~500k death toll in 2020 to the flu's ~50k average death toll for recent years. That said, I haven't heard of anyone using the "just the flu" argument since the first couple months of the pandemic.

>That said, I haven't heard of anyone using the "just the flu" argument since the first couple months of the pandemic.

I'm jealous of the circles you run in.

It's good if you want to understand how Americans as a group perceive the world. It affects which measures they will support and how they will respond to different arguments.

How does this inform us about how Americans perceive the world? The best I can figure out from this data is how many people the average American "knows".

329,484,123 (US pop) / 3 / 500,443 (COVID-19 deaths) = 219.46

Apparently the average American knows about 220 people. Impressive.

EDIT: That's assuming each of the 1/3 Americans only knows exactly one person who has died of COVID-19. If we consider some of them know more than that, then the result should be even greater than 220.

I was wondering the same thing, so thanks for putting this together. My one suggestion would be to include only the adult population, since I'm guessing their data / surveys only include adults.

196M (approximate adult US pop) / 3 / 500K (COVID-19 deaths) ~= 130

But the numbers, i.e. average American knows somewhere between 130 to 219 people, seem a bit high.

When someone you know dies, it shakes you up and changes your perspective. This process has happened to 1/3.

I'm sure more than 1/3 Americans know someone who has died in the last year. I'm sure that was also true last year, and the year before that, and the year before that, etc. If the psychological impact of death is your concern, there's little value in focusing on covid-specific deaths.

Would probably correlate with 1 in 3 Americans know someone who is obese and has resulting comorbidities.

More than 1 in 3 Americans are obese themselves.

Not remotely, no. That number would be more like 99 in 100.

More than 40% of Americans are obese.

Comparing a major health crisis to another major health crisis and is comparable. Doesn't sound great.

I know one, a co-worker. He was not young, and like the rest of the organization had been working at home since mid-March.

A women down the block, whom I did not know, died of it last year. Nor was she young, nor in particularly good general health.

Hmmm, the naive combinatoric guess would be:

1 - ((329500000 choose n_connections) / (330000000 choose n_connections))


That comes out to about 14% of American should expect to know someone who died of covid if everyone knows 100 people and its completely random who dies. With some clustering of social networks this seems like a pretty unsurprising number. It mostly depends on how people defined "knowing" someone. I would have guessed the percentage would be higher personally, but then, I don't know anyone who died from COVID.

Coworkers of a family friend and my uncle died from it, even though their jobs could be done remotely (my uncle's a sysadmin ffs)

Our oldest daughter's friend didn't get it despite it putting her 3 siblings and 2 parents on their ass for over a month.

My wife and I suspect it played a role in her father's death. This occurred soon after nursing home shutdowns in the state and he had been dead of "heart failure" far longer than they had said in the initial phone call. They were fairly evasive when asked follow up questions about what had occurred and my wife and mother in law had gotten influenza (and hospitalized from it in my mil's case) from the same nursing home. He was in there recovering from a physical injury but had been dealing with health issues for a few months already, so who knows. The important thing was that my wife had her first viewing of her dead father completely alone, followed up by a "funeral" of immediate family members. Oh yea, and this all happened soon after his birthday that she wasn't able to celebrate with him due to the lockdown.

I feel like focusing on total deaths, death rate, etc. sometimes misses the point and allows those who rebel against any measures taken a rhetorical out. I don't like being sick, I don't like my family being sick. We've had our entire family wiped out from things like the flu and gastroenteritis. It sucked and it was an absolutely miserable experience. So even if I was fully guaranteed that no one would die (which is unlikely given what happened previously with my mother in law who is currently living with us) I would do what I could to avoid coming down with it.

Yes, many miss the impact of the illness not just the death. This is why a lot of white collar remote-able firms have kept us home.

It is in your companies self-interest to have everyone home&working than having large numbers of staff out sick for days to weeks in batches as the virus cuts through open office floor plans and poorly ventilated conference rooms..

It wouldn’t be uncommon for a hundred people to be counting the very same COVID victim.

No one's asserting 1/3 of Americans have died, correct.

Also from the source polling:

Our survey finds that the public remains fairly evenly split on the accuracy of the reported death toll, whether it is higher than reported, lower, or about right.


Survey method is on page -2:

This Axios/Ipsos Poll was conducted February 19 to 22, 2021 by Ipsos using our KnowledgePanel®. This poll is based on a nationally representative probability sample of 1,029 general population adults age 18 or older.


Those in lower socioeconomic classes are the ones I've known who've gotten it, mostly because they were forced to work in offices or some public facing role where they had to interact with a lot of people. There is a terror of small business owners who act like despotic lords of a fiefdom, that I find to be some of the worst actors in general.

Polls are designed to sway public opinion, not measure it.

I find that remarkable, because I barely know anybody who even got it, and nobody who was hospitalized. Presumably that says more about my social circle: people followed the rules. Which isn't perfect protection, either, but apparently better than nothing.

[EDIT: I am drawing "social circle" very broadly, to include a very, very large set of online-only friends.]

I had said at the beginning that there was a large crowd who would not take it seriously until they personally knew somebody who had died. They were also the ones most likely to not follow those rules. So I wonder to what degree their minds have changed.

I think it is just a question of odds. If there was a bit more of 1,5 deaths per 1k people and and the average us citizen knows about 600 people, then about 1 in 3 will know someone who died.

I do not think it has to be someone exactly in your close circle, I think 'know' here means people you meet and maybe you have no contact with.

Source: https://www.nytimes.com/2013/02/19/science/the-average-ameri...

I know at least 3 who died directly from it. I know far more who you could reasonably say died indirectly from it.

What? You know "far more" than 3 people who died indirectly from COVID? Like, from suicide or avoiding going to the hospital when they were sick with something else? You must "know" a huge number of people.

You must really be invested in the idea that covid doesn't kill to have this reaction. My experience is anecdotal and doesn't have to be representative of yours to be true.

"I know people who died indirectly from covid" is a confusing sentence.

I feel very fortunate to not know anyone who has died. My brother got it and got over it pretty quickly. So far he's the only one.

One of my friends traveled to 11 states, ate out regularly, etc. and never got it. Crazy.

(please don't take these statements as any sort of COVID denialism, just sharing some personal experience).

I had two dead colleagues (both older but neither elderly, out of a ~50 person company) to COVID by the end of April.

I think a lot of this depends on when COVID hit your area in force. The areas that got hit early in the pandemic knew quite a lot less about treating it effectively.

Or your friend got it and was asymptomatic.

Yea that's what I was thinking, but people tend to have not so great reactions from the first dose if you have the antibodies.

One of my friends traveled to 11 states, ate out regularly, etc. and never got it. Crazy.

Some people have had it and been largely symptomless.

I don't know anyone who has died from it either. This is probably mostly commentary on my lack of social connections more than anything else.

He has gotten both doses of the vaccine too. I think in all the testing he's had he never had any evidence of the antibodies. Not that he couldn't have had it.

> One of my friends traveled to 11 states, ate out regularly, etc. and never got it. Crazy.

I think the mix of asymptomatic or mild cases has been really bad for a coordinated response. It’s not hard to find people who were lucky, and there’s a constant chorus of people saying that’s all anyone healthy needs to worry about. Then you talk to a doctor, nurse, EMT, etc. and hear what it’s like on their side with unlimited overtime and patients ranging from the expected older / comorbidity cases to young people previously in good health.

That;s it! So 100% of the country in lockdown because only 30% even know of a person who has died? And how many of those 30% are the "friend of my cousin's friend's niece..." type knowing.

1 in 3 Americans know someone who died of old age in the last year

Correlation is not causation

What would the implied causality be here?

Fortunately you can't catch old age from touching a doorknob.

Probably won’t catch COVID from a doorknob either:


And I knew someone would post this snarky response. Yes, COVID is definitely different from literally every other microbe we know of that spreads on common surfaces. Also "probably not" still means hundreds of thousands of infections at the scale of this pandemic.

I pointed it out because referencing a questionable mode of transmission undermined the punch of your original "snarky response".


Now do other causes of death

That'll be heavily confounded by the fact that the other causes of death have been around longer than a year.

There are none

This is an extremely trivial assertion to debunk.

Just look at total deaths - all causes. The "they're just reclassifying normal deaths" conspiracy theory falls flat when you scroll to the "Weekly number of deaths (from all causes)" chart.


(If you can't find it, here's the screenshot: https://imgur.com/a/6bk0PtF)

This is an extremely trivial assertion to debunk.


"Not a single case of influenza has been detected by public health officials in England for the past seven weeks"

In US it is very regional, socio-economic and political. I knew 0 people who had it from March->October. I then knew 30+ people who had it across 4 different outbreaks between Nov 1 and Xmas. Broke down to - - Working class / front-line workers who had to physically report to work

- People with kids in college who came home

- People who were more right wing (its all a hoax)

- Extroverts who refused to stay home / went to bars as much as allowed

And the commingling of above.

Fall/Winter wave was crazy. Of the 30, 1 went to hospital, none died. Most of them were on their ass in bed for a week, and some have not had a return of normal taste/smell after 2 months.

I do have an indirect "know of" relative-of-relative who died, as a frontline nurse in area that took few precautions, during the summer wave sadly.

People still don't understand how quarantine and testing lag work, so I know of multiple scenarios like this:

* Person A feels off but tests negative in first 3 days so socializes with Person B. Then tests positive 2-4 days later.

* Person B starts to feel off after 2-3 days, gets contacted by Person A that hey I got it

* Person B goes and gets tested and pops a negative in first 3 days ...

From here if Person B quarantines and keeps re-testing until positive or clean for 14 days (unlike Person A) it's over.

However lot's of Person Bs kept going out and spreading.. One in my family turned a single kid coming home from college into 4 entire households sick in a week, and the guy who ended up in hospital was a Person C in this scenario.

> Most of them were on their ass in bed for a week

It's really interesting to me how accounts of the virus differ so greatly from person to person. In my personal experience, I know of about a dozen people who have tested positive - all of whom showed few or no symptoms (usually exhaustion and a cough) and felt fine within a couple days and having no long term effects.

I'm sure part of it is influenced by the demographics of who you know (everyone I know was under 40). Confirmation bias might play a part (severe cases stand out more and may seem more representative than they really are), but I'm probably only suspecting that because my personal experience has been so different from yours. Could it have something to do with different regional strains with varying severity of symptoms?

It's a poll, not a meaningful analysis. Anecdotal experience is all that's needed to know it's BS.

Polls are meaningful analysis. They always publish the details somewhere, in this case the analysis is here: https://www.ipsos.com/sites/default/files/ct/news/documents/...

1,000-person panel, weekly poll over the last 52 weeks.

That's a little small, but not that small. I wouldn't trust statements about demographic minorities out of their tiny fraction, but the 1-in-3 number across all 1,000 probably scales up with a single-digit MoE.

The anecdotal experience is just as garbage though. Something like this will have very clustered behavior. I.e one grandma dying will color up quite a large number of people (probably 50+). The number is very fathomable.

I think it depends a lot on what type of job you and your family members have and where you live.

And yet there are more than 500k deaths in a population of 330M. I mean, the odds are there.

So what are the odds that 110M know one of .5M USA persons? "Dunbar's number is a suggested cognitive limit to the number of people with whom one can maintain stable social relationships. . . . It has been proposed to lie between 100 and 250, with a commonly used value of 150."[0]

So .5 * 150 = 75; .5 * 250 = 125, so 110 is in the upper range of Dunbar's number which may reflect asymmetric "personal knowing," whereby someone may be personally known to others even if the others might not be personally known to someone. To me this makes intuitive sense since the disease mortality is skewed toward older people who I think would be known by more people.

0. https://en.wikipedia.org/wiki/Dunbar%27s_number

Applications are open for YC Summer 2021

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact