How does your calculation take into account the number of additional deaths due to the millions that are unemployed and without insurance and die as a result of that? There was a poster that had a stat of 40,000/year dying because of unemployment. That was when unemployment was 4%. What happens when that quadruples (which it is predicted to do by June)? Do we quadruple the death rate due to unemployment? I really don’t know. Is that accounted for?
How does your calculation take into account what happens after restrictions are lifted? Is the assumption that in 3 or 6 months or whatever, the virus stops spreading? There was an ongoing thread on how this ends. It is an open question. If this doesn’t end until we have a vaccine, shouldn’t this be run out for another 12 months? Maybe 18? What happens to the unemployment rate if we are on lock down for a year to a year and a half? If we are going to socially isolate to keep the death rate as low as possible, we’ll be doing this a long, long time. Not just 6 months.
Your calculations seem to suggest that at a minimum, with an “8” scale of social distancing, we’ll see over 700,000 deaths in the workforce alone. Does that seem accurate to you given the numbers we are seeing now?
Maybe I’m misinterpreting what your Fermi analysis is demonstrating. I’d say you are off by at least an order of magnitude.
a) The usual goal of a Fermi estimate is to give you a sense within an order of magnitude. E.g. between x10 and ÷10.
b) It’s too high. At the moment, we look to be on target for about 250k deaths, across the entire population. The numbers I used were pulled from real numbers that I’ve seen (though I forgot to link to all of them). Assumption #8 comes from Marc Lipsitch but I’m guessing that he’s being pessimistic even in his best-case scenarios. (He is doing the whole planet, not just the US, so that might be fair since it includes India.) I would need to play around with some numbers to compare to how reality has behaved, and then write up a simulator to get a proper result.
250k in the entire population. Which is twice the workforce. Which is going to necessarily include a lot more 55+ folks than you included. Obviously this means AT most 125k will be from the workforce. Your low end lists 700k deaths in the workforce.
I’ve heard the white house talk about between 100k and 240k. Some folks are saying the estimates at IHME are grossly under estimated, but they are at a median of 81k right now.
Where is the 250k coming from?
The analysis can’t be single sided in assuming the number of deaths is single sided. Both options will cause deaths the other won’t. In the case of this virus, you’d be trading killing mom for grandma if you use 8 instead of 2, at least to some percentage. Maybe that tradeoff is ridiculously small, as has been pointed out previously. I certainly couldn’t do anything but do wild guesses. But it has been cited 40,000 people died last year due to a lack of health insurance. That doesn’t seem insignificant compared to talking about 250k dying due to the virus.
I used the U of W estimate of how many beds are needed going forward. If you assume that the total number of people who have died went to hospital about 11 days ago or earlier, and track forward, calculating the area under the curve, that’s about the number you end up at.
Their estimate is based on the US and American policy, and updated regularly. Lipsitch was doing the whole planet and was basing his results on how things looked a couple of weeks ago.
Also, just because you tell people that they can go out and hang out at a bar despite the pandemic doesn’t mean that they actually will. I’m not staying home and social distancing because the governor tells me I have to, I’m staying home and social distancing because I don’t want my chronically ill wife to get Covid-19 and die.
I can’t find the U of W estimates. Can you link to them?
I’m wondering how different some of the various models are at this point. Lots of people point to the IHME models, but as I said, there is a serious question about those.
Models of infection which I have seen have generally had a strong leftward skew rather than being a fairly standard bell curve:
That makes me a bit skeptical of the people who are doing this stuff but I haven't had a chance to dive into the question to try and determine why there's such a notable difference and which way to lean.
The basis for my number was to assume that people stay in hospital for about 11 days before dying.
In the IMHE graph, the growth in deaths and the growth in beds both happen at the same time. There’s no offset. With my view of things, the death curve should lag the bed curve by some amount. And, if that view is correct, and we’re seeing a death curve of the size we are this early, then we need to magnify their prediction by quite a ways.
So how about we do something about the health insurance situation, instead of saying ‘well we’ll just have to decide to kill a lot of people so other people can keep their only-if-you-have-a-job insurance’?
And, yet again: if everybody gets this more or less at once, why do you think all of those people will keep their jobs anyway?
Sage Rat proposed that “letting grandma die” wasn’t such a great economic idea afterall because the numbers killed were so much more if we did nothing. I called him on his numbers. I didn’t proposed just letting this this rip, so we aren’t talking about a situation where everyone gets this more or less at once. His analysis was faulty by at least 3x, and maybe more. It is off by an order of magnitude if you take experts’ advise instead of believing he knows more than the experts.
I’m dearly hoping we get universal health care out of this and some serious other improvements. I’m also hoping they find a vaccine tomorrow or this thing mutates into something far less deadly. I can hope and wish for lots of things. Reality doesn’t care. Reality says millions are now unemployed while so many sit around hoping the people in charge do the right thing. Millions are now going to struggle, and many die, because what we hope for isn’t happening.
You don’t want to kill grandma? Fine. But at least have the courage to understand you are now killing mom or dad or junior.
Right now we are all going to be sitting around for the next 2 months so we don’t overrun hospitals and everyone who needs a ventilator can get one. Guess what? Way more than half (best number I’ve seen is 67%) of those that go on ventilators die there.
If we weren’t killing people in the process of saving people and if this were just about money, sure, kill the economy and we’ll figure it out later. But it isn’t that easy. To discount the misery and anguish we are going to be causing millions for the sake of saving hundreds of thousands is selfish.
I do think we can open up parts of the economy to keep it functioning, albeit at a lower level. The hospitalization rate for those under 65 in NYC is 0.4%. You really think having 0.4% of the people in the hospital is going to shut down the economy? Even more, only 70% of the people will get this, so that number is even lower.
That’s a fair point. I was struck by how well economies that were patient and enforced strong social distancing measures over a longer period of time recovered better than those that seemed to either not take it as seriously or went back to work early. I do think that part of it is something we should look at.
Mom and dad and junior would also be at risk if almost everyone gets sick at once – partly because yes, mom and dad and junior sometimes die of covid-19; and partly because it’s not going to help them to have health insurance if they can’t get medical treatment anyway because the medical system got entirely overwhelmed with covid-19 patients, meaning both that they can’t get hospital beds or surgeons or diagnostic tests and also that the health care people who could otherwise have treated them are down sick themselves for lack of protective gear.
yet again, yet again, yet again – if nearly everybody gets sick at once, mom and dad and junior are very likely to lose their jobs in any case. Dead people won’t patronize their businesses, sick people won’t patronize their businesses, people who are desperately trying to nurse their loved ones through disease at home for lack of hospital beds won’t patronize their businesses, people who are still healthy but are just plain terrified of going outdoors won’t patronize their businesses, people who found out that six of the clerks at their business were diagnosed yesterday with covid-19 won’t patronize their businesses. You’ll just be killing grandma and the economy at the same time.
Your point 2 is a good point and it was one I had considered. I’ve never proposed letting everyone get sick at once. Hopefully that is clear at this point.
Sage Rat tried to quantify all this and I think he is off in his calculations. I’ve stated my case in other places on the board and I’m happy to outline it again, but I’d mostly be repeating myself. I’m truly happy to discuss it with you if you’d like to hear my thoughts on your two points. Your points are valid, to be sure, but I do have counter arguments but I don’t want to hijack further.
The magnitude of things was the important element of the numbers, not the absolute values.
In either case, the OP posted a better analysis than mine so getting lost on someone’s back of the envelope sanity check isn’t particularly useful to the topic.
The article in the OP explicitly states the 1918 flu is rather difficult to compare to now, so it certainly isn’t an ironclad determination that the economy would be better if we kept things shut down longer. Your analysis, while actually quite impressive, I think needs to be tweaked to get better estimates on the trade offs before we can use it to get even an order of magnitude evaluation of impacts.
Things are just too different now than they were in 1918 to make any comparison to what happened then valuable for what we are doing now. Shutting down the economy back then meant mostly* saving the lives of the most productive parts of society - people under 50. That isn’t true with this disease. Further, we have the complication of the impact of losing jobs has on people today that it did not have back then - losing health insurance.
*Please, please, please do not restate this it not just an old person disease and those dying aren’t just those over 80. The VAST majority of those severely impacted by this are over the age of 60 or 65.
But they didn’t save a lot of those lives. A lot of those people died. Productive people. From what I hear, society should have gone back to the stone age for decades to come.
And yet, somehow that didn’t happen. In fact, wages rose following the pandemic. And why not. A couple of your coworkers die, you don’t, it’s going to be better for you as a worker.
Economies are circular. The one thing I will say is that there is more slack now. Not because people are lazy or stupid, but essential services consume less of every dollar. All of those productivity improvements of the last 100 years have to go somewhere. Lately they’ve all gone to capital. Capital then seems to have an increasingly tough time finding good investment opportunities, as witnessed by our ever falling interest rates.
So maybe fewer jobs are “needed” as opposed to 1918, but it’s not like there’s anything else for these people to do anyway, and they are needed as consumers. I am not worried about the national debt, because frankly even if they paid it all back the money would just sit in some vault unused. I am totally an advocate of doing UI, money drops, whatever it takes until we are through this and don’t have to focus on this again. The only thing that can make things really bad is to get “serious” about the situation, try to call all of the unpayable markers in, and insist on “real money” or the gold standard or some crap like that.
Help me out here, so I’m clear.
Article in the OP: Shutting down the economy in places saved lives and those places that shut down sooner and longer recovered quicker.
Jay Z:It didn’t save a lot of lives. Many died. Therefore, the economy recovered quicker because there were fewer laborers.
We need to get the moral of shutting down in 1918 straight. Either shutting down saved lives and that was good for recovery, or it didn’t really save many lives and that was good for recovery because it became more of a workers’ paradise.
As far as saving lives, then as now different things were done in different places. Philly had a parade, St. Louis cancelled theirs, more people died in Philly. Then as now, the “stay open” choice is a false choice anyway, once the bodies pile up everything closes in every society and what did you gain by staying open those few extra days?
So in 1918, when the flu was sweeping through, businesses did indeed close, factories closed for a time. Now there are several economic questions.
Why weren’t the economies crippled by the businesses and factories not producing for 3 weeks, let alone 3 months or so?
Why weren’t the societies crippled by the factories and businesses not pumping out product, using processes that were less efficient than we have today?
Why weren’t the post flu economies crippled by having so many of the productive workers die, not being around to pump out more stuff? When they’re further dealing with the problems in 1 and 2, in a society that had far less room for slack?
I just think that after 100 years of progress, 100 years of productivity improvements, 100 years of further automation in factory, farm, and business, that we’ll be crippled if we reduce our production AND consumption for a time to deal with this… it’s absurd. 100 years, and we have to keep working like slaves and consuming like we don’t know where our next meal will come from.
Take care of the people that need help, get people well, get rid of the virus or it goes away, then these businesses can reopen.