But *when* will you favor a shift in coronavirus strategy? (no Straussians in a pandemic)

I agree with the numerous sentiments, for instance as expressed here by Ezra Klein, that we are not facing a dollars vs. lives trade-off, rather the better solutions will improve both variables.  Also read this Tom Inglesby thread.  Furthermore there is a concrete path forward toward general improvement, for instance read Zeke Emanuel (NYT, I don’t agree with every detail but the overall direction yes).  And don’t forget these costs cited by Noah.

But we are economists, not mood affiliators, and so we must address the classic question of “at what margin?”  At what margin would you favor an actual shift in strategy because the virus already had reached so many people?  And yes, such a margin does exist.  At that margin we would continue some of our defensive responses, but the overall approach would have to change away from the above links.

Let’s say everyone had been exposed to the coronavirus except yours truly.  Should we shut all (non-take out) restaurants just to limit my personal risk?  Clearly not.  And likely I would end up getting exposed sooner or later in any case.  Then you should “let it rip,” and let Tyler decide when he wishes to go outside or not (but of course offer him health care).

So what is the margin of bad outcomes where, after that point, a major change in strategy should set in?  Has to set in?  That is the question we all need to answer.  And what should that strategy change be exactly?

We like to say “speed is of the essence,” but a less frequent spoken corollary of that is “at some point it is too late to stage the defense we had been hoping for.”

What if we made no further progress against Covid-19 after two more weeks?  Three more weeks?  How about a bit of progress on testing across the next month and a modest increase in mask capacity?  How much longer is the cut-off?  Given how rapidly the virus spreads, it can’t be that long from now.  It cannot honestly be “four months from now.”

(For the record, I am still optimistic, but not at p = 0.8, so this eventuality is by no means purely hypothetical.  And it is perfectly correct to note that Trump’s own incompetence is to some extent making the whole dilemma come true, and that itself is deeply unsettling.  Agree!  We should have “gone Singapore” months ago.  But the dilemma is now here nonetheless, noting that we are hardly the only country in this bucket.  You can’t just condemn Trump and stop thinking about it.)

Or what if New York and seven other regions are hopeless but the rest of the country is not?

I am fine if you agree with me, Ezra, Tom Inglesby, Zeke Emanuel, and many others, including most of the Democratic Party public health establishment.  We all favor “speed is of the essence.”

But the next part of the message never quite gets delivered.  And no one wants to talk about what the next strategic stage — if we fail — should look like.

It is imperative that you consider where your line lies — if only mentally — when you would jump ship and indeed…confess a significant degree of defeat and then formulate and push for a new strategy.

Addendum: Straussian Tyler is not entirely comfortable with this post, as he, like his brother Tyrone, prefers to tell the Noble Lie and maintain the illusion that the preexisting struggle must continue across all margins and at all times.  But perhaps, these days, there are no Straussians in foxholes.  So pick your “no return” point, write it down, and then get back to me.  The honesty of our policy response requires this, yes?  I’m not even making you say it out loud.

And don’t you find it strange that no one has been willing to raise this point before?  Could it be that we are not being told the entire truth?  Or are people not telling the entire truth to themselves?  Isn’t that the same mistake we’ve been making all along?


So basically the ship is sinking, and we're trying to disentangle the few lifeboats that were properly maintained, the captain has abandoned ship, but some of it's crew are still trying to slow it sinking while possibly sacrificing themselves. And some of the first-class passengers want to pause on the deck and do a classroom ethics exercise about who gets to climb into those lifeboats, and whether it's really worth the bother to lend a hand to those sailors down below trying to plug the leak while the water is chest high.

Is this really what the Marginal Revolution is about?

I look forward to all the wise cost-benefit analyses that will be forthcoming, virtually all of which will look idiotic in about two weeks.

@KevinK - "Is this really what the Marginal Revolution is about?" - that's life. What do you think "triage" is? Would you rather have a doctor or economist decide the fate of who lives and who dies? I rather trust my life to TC, having seen how many doctors operate. A doctor, even Dr. Fauci, is in many ways no different than a used car salesman. Seriously.

Cost benefit analysis here: Cass Sunstein-type cost-benefit analysis from two different people shows a lockdown (quarantine) does pay.

@Student from themoneyillusion.com 3.25.2020 -" Suppose we shut the entire us economy for 6 months (no one is saying, and it’s not even possible, that but just suppose). That’s about $10.5 trillion. Now suppose we go herd immunity. Say 33% of Americans get it. Of that, say 3% die (remember hospitals will be swamped so it could be double that percentage). That would be about 3 million people. So, if a person is is worth more that 3.5 million dollars (10.5 trillion / 3 million), [a person is worth $10M--RL] it would make sense to shutdown the economy for 6 months."

RayLopez estimate: Superforecasters site saying 350k or less deaths in 1 year in USA. US GDP will suffer 50% drop in lockdown. Israeli scientist says you need six weeks lockdown minimum for Covid-19 to drop to low numbers, let’s make it 2 months, and assume 0% GDP during these two months (not realistic but worse case). US GPD is $20.5T/12 = $1.7T/month Do the math: 2 months*$1700 B / 0.350M people dead = $9.8M per dead person, right at borderline for the value of a human life, so shutdown is economically justified (also health care for survivors not factored in, and 'precautionary principle' since there's a rare chance Covid-19 survivors don't have immunity to SARS-CoV-2)

Federal govt in US ‘overvalues’ human life: $9.6 million (Department of Transportation, Aug. 2016)[Wikipedia] (some estimates say it overvalues human life however). Given Covid-19 has long term health costs, it makes sense to do a 2 month shutdown now. However, shutting down the economy for more than two months might have long term problems, so 2 months max is my recommendation. Besides if you can’t kill Covid-19 with a two month quarantine, likely more months won’t matter.

Bonus trivia: When is the last time an economist became US president? Tyler Cowen for president! Would be a big improvement over both parties.

@ Addendum (Tyrone) - another point to make is that 'de facto', the USA is on a path of 'herd immunity' (do nothing) since there's not a nationwide quarantine. Florida Man will infect you. De facto, once hospitals fill, doctors will perform triage and mostly older people will die, like they did when Hurricane Katrina struck. Society is not comfortable 'seizing the bull by the horns' and solving the trolley problem. BTW, my analysis above is the best ("Student's" is too biased towards shutdown) and none of the links show such a pithy analysis. It shows that, barely, a 2 month shutdown is called for now, just like they had in Wuhan. But sadly, I doubt the USA will go down that route, and people will needlessly be exposed to the virus, with mostly older folk dying, and some younger folk having long-term lung problems and male infertility. That's "American Exceptionalism" at work.

I don’t think we are even close to being ready to let it rip. We can’t agree on the fatality rate (ranges from thousands to millions), we don’t have random sample testing for people to make informed decisions about risk, we don’t have PPEs to keep the healthcare system in tact, I am not sure the national guard is ready to go in case law and order starts to get dicey.

If we are going to, “let it rip” we need to get our house in order. We need the above and we need to do this in a coordinated manner.

Honestly, the thought of this terrifies me and the rising odds in my mind that this is inevitable is giving me anxiety.

@Student - right. You'll note we did not take into consideration long-term health benefits, nor that during quarantine the US economy is not zero (probably only down 50% at best), which would make quarantine an even better bet than before.
And a challenge to "let it burn" herd immunity people: if you're so confident Covid-19 will not harm society, and you're young and in good health, volunteer to let it infect you, then you an be either a data point in some database or on the front lines fighting the disease and immune from it. Any takers? I did not think so.

Excellent blog you have here.. It's hard to find high-quality writing like yours these days. I truly appreciate people like you! Take care!!|

It’s not 9.8 M per life. I calculated that the average number of years lost is 8.5 per death or perhaps 850k.
I used demographic and life expectancy tables and accounted for the fact , the fatalities in the 50-80 years brackets have worst life expectancy than the average of their bracket due to co-morbidities.
So your estimated may be off by a factor of ~ 10
btw : Candidates canceled. What has the world come to?

Yup. I doubt the $9.8M anyway. I need to read more about where these come from, but I doubt revealed preference or lifetime utility calculations can possibly get that high.

@Catinthehat - show your work. Seems you are assuming only old people die, which is not true, and you fail to account for the health problems Covid-19 survivors have.

@Chuck37 - one liners with "yup" don't cut it, even and especially as a troll. Raise your game. And I said the $9.8M a life figure is a high, but in today's developed country society it's probably a safe assumption (risk adverse)

Ray. I don't owe you anything. It's very simple to do. I am sure you can do it. Don't be lazy about it. you can age weight the deaths and the demographics and the life expectancy. I use $100k per value of a year. ( you can change that number).
You're the one that acts like every fatality had 80 years left to live.
Show me your work that a 70 year old dying or 80 year old dying should be counted as 9.8M. At least 80% of the deaths are over 65. A lot of these have co-morbidities such that their life expectancy is already reduced.

Who in here wouldn’t pay 10 million to save their mother if they had 10 million? There are probably people that would pay it (again if they had it) to save their dogs. Human beings are not goods. They are our families.

from Wikipedia: United States
The following estimates have been applied to the value of life. The estimates are either for one year of additional life or for the statistical value of a single life.

$50,000 per year of quality life (the "dialysis standard",[30] which had been a de facto international standard most private and government-run health insurance plans worldwide use to determine whether to cover a new medical procedure)[31]
$129,000 per year of quality life (an update to the "dialysis standard")[32][31]
$9.1 million (Environmental Protection Agency, 2010)[33]
$7.9 million (Food and Drug Administration, 2010)[33]
$9.2 million (Department of Transportation, 2014)[34]
$9.6 million (Department of Transportation, Aug. 2016)[35]

Granted , I didn't consider, potential health problems of survivors. I didn't see anything explicit about it on MedRxiv.
If you have a link, please share

Hand waving by Catinthehat noted. See also the below, where instead of doing 0% for GDP, as I did (which favors 'doing nothing'), in fact the Fed paper says GDP does not suffer much in a partial quarantine like was done with the Spanish flu. Which makes the case for quarantine overwhelming. You're dismissed ...but you owe my an apology. -RL

New Fed study finds efforts to slow pandemic don’t depress the economy
Published: March 27, 2020 at 8:52 a.m. ET By Steve Goldstein

The economy isn’t helped by rushing back to work in the face of a pandemic, according to a new study by Federal Reserve researchers.

At a time when the lieutenant governor of Texas has said senior citizens would be willing to sacrifice their lives for a better economy and when President Trump has fretted that the cure may be worse than the disease, a new research paper looked at the impact of the 1918 Spanish flu on the U.S. economy.

The most likely outcome is that we let up and "let it rip" on a slower and more controlled basis so that there is no panic.

So over the next few weeks the US should:
1. Ramp up PPE production (CA nurses refuse to work without it)
2. Ramp up testing capability and develop antibody tests
3. Widespread mask distribution and handwashing education
4. Wait for the first wave of infected to die or discharge so you can clear up hospital space.
5. Develop a nationwide triage system (for liability reasons) to prevent hospitals from being swamped. e.g. Over 55, you get oxygen only and are automatically DNR. Over 70, you don't even get oxygen, just a morphine drip.

Then you can relax restrictions and let people decide for themselves whether to go out and mingle. Some will get sick and be hospitalized or die but at least they will be masters of their own fate.

Yes, we will combine the worst of both world's. Partial shutdown for 2+ months most places, with longer local shutdowns based on governor competence. This will not be long enough to really stop COVID-19. Then, Trump's TV watching will ensure we do "let it rip," squandering a lot of those short-term grains from the shutdowns.


I think we can be relatively confident the government will consistently choose to incur maximum cost while simultaneously failing to turn costly but at times necessary policy into actual benefits.

Which is basically our modus operandi, see War on Terror.

All the right “experts” claim that we need to “flatten the curve,” which is just what you’re saying. Note that this is not the same as “defeat the disease.”

The defeat the disease went out the window when those in charge did nothing between Christmas and the Ides of March. Stock piling PPEs, getting random sample testing up and running, preaching Hand washing and social distances from the rooftops (which probably would have been ignored either in any case). Bend the curve was always about showing the outbreak to avoid overwhelming the health care system. What could have been if everyone was masking in February? Those darned counter factuals.

Actually, they won't be masters of their own fate if someone decides to go out and mingle as an asymptomatic carrier who then infects someone and so on. The whole point of the shutdown is to cut the number of interactions. This is like arguing that firing a gun into the air is just taking a risk that the bullet will come down and hit oneself and completely ignoring the possibility of the bullet coming down and killing someone else.

1. A drop in GDP also leads to an increase in deaths, ceteris paribus.
2. A substantial percentage (single digit?) of those who die would have died anyway.

Also, right now we have 1 serious Covid case per 4 hospitals. Probably stressing the system in a few places, but not exactly a crisis yet.

It would be nice to have a tweakable model of disease spread, # needing hospital care, recovery, etc. I think I saw some graphic like that a week or so ago, but not something where one could change the parameters.

We've seen a dramatically rising GDP and a rising death rate among white males. Perhaps a falling GDP would reverse this.

The spread was slowing (at least given the growth factors) for a couple of days but I am starting to think it was an artifact of testing lags (or even worse, less testing, not more. I am almost beginning to wonder if that is deliberate at this point)...

It’s kicked back up yesterday and today. The GFs were >1.2. Today it was 1.28. It’s been, what 8 days since the shutdowns started, and we are still way exponential. We are on pace in the USA to hit over 1 million cases by April 4. Given the median time to showing symptoms is 5 days, I would have expected to see a slow down. It appears as tho there is still to much local transmission and to much travel between the clusters. It’s behaving exactly like a fractal.

Unless testing increases like 1000% in the next 8-16 days, we will be in the millions of cases by Easter.

It’s looking like we are gonna go herd immunity by default. It’s looking like we just can’t get it done. We just can’t do it. I fear the cat is already out of the bag.

You see the presser today...

"When people start talking about 20 percent of a population getting infected, it is very scary but we don't have data that matches that based on the experience,"


What the hell data is the good doctor looking at? Can anyone explain that? I would love to know. I pray to see it.

As of today, we are undeniably on pace to hit millions of cases in the USA by Easter. we will be at 150,000 in 2 days. Just look at the damn trends people. I would like to know what makes anyone think the models are wrong? I don’t see a shred of evidence of a slowdown (at least one rapid enough one to make any significant difference in the near term).

Can someone please calm my nerves here. I am praying that we bend the curve, I just don’t see a lick of evidence that that is happening.

It takes longer than 8 days to see results. Mean time to symptoms is about 5 days, but in most places (especially places with a high density of cases), the guidance when symptoms show up is "stay home and self-quarantine"; tests are mostly reserved for about-to-be-hospitalized patients (plus high-risk groups). Reagents are slow, so test results can take multiple days to come in. Bottom line: I think the majority of "new confirmed cases" today were infected substantially more than 8 days ago.

'Is this really what the Marginal Revolution is about? '


America is at total war against COVID-19. Total wars are won with sustained economic focus on winning. The American Economic Machine has to focus on two objectives:

1. Isolate the Illness
2. Fortify the Health System

The world is in a tight OODA Loop. Now is not the time to focus on anything that is not relevant to the current loop or the trends that will orient us in the loops ahead.

> America is at total war against COVID-19.

You must be young. This is nowhere near what total war looks like. This isn't even a street brawl.

I am not young. The important parallel to Total War is that the fight is prolonged and ultimately is determined by a nation-state's economic capacity rather than its military strategy or capacity. Rather than being young or old, let me guess that you misunderstand the dire nature of the scenario that just started in New York City. I hope that my estimate is wrong, but it looks like America is on track for a scenario that is much worse than Wuhan, and much worse than Italy.

Over under that corono deaths in 2020 will be more or less than influenza deaths in Italy?

Right now Flu well ahead.

What were influenza deaths?

I'm not able to find recent numbers for Italy flu deaths this year. The best I can find is 240 from January. That's significantly less than the 8000 attributed to COVID-19. Do you have some more recent numbers? If not, what odds can I get on the over?

We don't know how long this battle will last but we know what the end game looks like. Once an effective intervention is widely deployed, the growth will continue for four weeks until it's peak, then another four weeks for the steep decline of the skewed Bell Curve to finish.

Cowen says:

What if we made no further progress against Covid-19...

We have a Suppression Strategy Bell Curve with a proven eight week end game and a Herd Immunity Bell Curve that peaks when less than 50% of the total population is Susceptible. Health System Capacity tops out when a tiny fraction of the population is infected. This looks like the
curves you see
in the classic SIR (Susceptible, Infected, Recovered) Model you see on Wikipedia.

The Herd Immunity Strategy is uncharted waters. The question is how society functions when the hospitalization rate of COVID-19 is above say 10% and the Health System is weeks/months beyond capacity.

"We have a Suppression Strategy Bell Curve with a proven eight week end game ..." We have one example that worked. And it was done in an extremely quick, all encompassing, and disciplined manner from what I understand. China locked down with a handful of deaths and ~500 known cases and more localized outbreak. In the US, at the city/regional level, the lockdowns might be happening in a context similar to what happened in China, but as a whole, the country is in a much weaker starting point. With multiple hot spots, haphazard, inconsistent implementations of lockdown, lack of state enforcement, lack of testing, transportation in and out of hotspots remains complete open...I am pessimistic about the prospects here. Still I am definitely willing to see what happens over the next two months, and would be surprised if the spread did not slow, but I am currently skeptical of US being able to squelch it like china did. But again, it still seems worth trying.

KevinK, I like you. You don't think a libertarian. Let these surrender monkeys run some numbers they pulled out of their collective asses. It keeps the little hamster wheel in their head going and lets them sound so self-important on the internet because in real life nobody would go to war with these losers.

You know why libertarians never won anything in the history of the US? Well you are seeing it now. They would rather give up and rationalize surrender as "herd immunity" or "but what about the economy!" All horseshit. You fight to win. Notice how comfortable they are with losing. These aren't people you want on your team. You want to send them to other team or out of your way.

Nice to see some open volunteers for taxing 90% wealth tax plus a heavy personal loan to pay for all this... If you and Kevin want to publicly make that donation, I'm sure it would be credibility enhancing.

It starts with the draft and the idea of public service. Completely replacing social capital with financial capital is foolish idea. When the Greatest Generation won the war, they came back with a shared purpose and a comradery that lasted until both feet were in the grave. Every American stranger you met was like a distant relative you could trust. We all saw and felt the same things. Everybody was familiar. We had each others' backs. The Boomers never got this nor any generation thereafter. This was a bad mistake. You know why Taiwan, South Korea, and Singapore beat this bug? Because they still have national service. Every able-bodied male serves and learns how to get things done as a team. No amount of "price signals" or "contests" or other academic wankery will replace good old-fashioned patriotic service.

Take it or leave it. That's how I see it.

So you wouldn't pay.

sorry don't mean to intrude but how did the conversation move from tyler's "it gets worse" scenario to kevin's "so we're giving up?" to col's "libertarians are quitters" to volunteer's "so we need a 90% tax". the first three make sense as a conversation but the last one seems random? if we pay a larger tax then what exactly, that libertarians won't quit? i'm not following.

Those who are not "quitters" should put their money where they mouth is and foot the bill for the "war" in which they declare themselves armchair generals. Not a difficult point. You want a war, you pay for it. Calling people quitters without volunteering to foot the bill yourself is pretty pathetic.

You are at "war" with the coronavirus with or without the keyboard warrior agitation. Arguing over the bill is nonsensical.

Well, no, you aren't practically mobilising a war economy no matter what happens. There are policy choices, and those that advocate for a stronger change from business as usual should pay for them, especially if they are the ones making copious rhetoric about sacrifice for the common good. Your replies show decreasing levels of rationality.

It's a truism that those that agitate for war should generally pay for it. In the spirit of that, I suggest "Colonel Boyd" who is indifferent to whether anyone can make their mortgage payments, should sell his and donate this, with all his assets to the "war effort" on which he is so keen. Be a lot more convincing he isn't an armchair general demanding sacrifice from others...

You make less and less sense with each comment. You too are demanding a sacrifice from others. A nurse in NYC is dead from the coronavirus for wearing a trash bag because they ran out of PPE. What sacrifice did you make for her death? Let's take 90% of your property to pay for her funeral and take care of her loved ones. See how retarded this is.

No, I did advocate any cause of action that is a change from "business as usual" which led to that.

Not a consistent extension of the idea that those that advocate wars and initiatives should volunteer to pay for them.

Your understanding here is "retarded".

And not only would you not, you'd claim it's good for young people (not yourself of course) to be economic cannon fodder because it builds a sense of shared heroism in them, makes them almost "comrades" (but not in a Communist way of course).

"We need a war every generation to stop the kids getting soft!" type stuff...

Easy to volunteer yourself to be a Colonel of course (and so self aggrandizing!), somewhat more credible to volunteer to be the economic cannon fodder.

My motto is "No shutdown" without compensation (at any margin). See: https://priorprobability.com/2020/03/25/a-takings-model-of-the-economic-shutdown/

"Your money or your life" is your compensation.

True, but if we truly value liberty, then people must be free to assume whatever risks they want. In the alternative, if we (the gov't) is going to force us to stay home for our own good -- and for the greater good (i.e. to slow down the spread of the virus) -- then I must be compensated for my losses, right? See: https://priorprobability.com/2020/03/26/how-much/

Then you must pay people who die from this virus. For example, the person in a nursing home who is exposed by a caregiver because whose husband works at a non-essential but ride or die type of business and brings the virus home.

But your scenario is imminently solvable. There are 30 minute tests coming out in Europe, and probably available here soon. A hand wavy 'everyone needs to be tested', which is nonsense, changed to 'a sizeable number of critical choke points in the transmission need to be tested' becomes a solvable problem in a much shorter time frame.

Considering somewhat exponential curve of growth, I would say that as soon sigmoidal curve approximation shows the peak higher than 50%, any further fight is completely theoretically impossible. But that is a rather late cut-off. When do you start? I would say , if 1% of general population (across the country, not just in one quarantined city) is ill, it is probably meaningless to fight and you can just accept that everyone will be ill, mortality be damned. However you can then expect mortality of up to 50% (not 3% like in controlled disease cases). If such a case happens in any big country (especially USA), we, as humanity, have failed, and it will be the biggest extinction event for humanity. A momentous failure, for sure.

Mortality of up to 50% would be several times worse than Spanish flu (which had a higher case fatality rate among the 0-60 demographic and was before fancy medicine), which seems unrealistic.

It's a crazy number based on what we've seen from this disease.

About 80% of people have mild cases and recover without any medical intervention. It's like needing to stay home with a mild case of the flu or a bad cold. The 20% who are hospitalized won't simply all die without hospitalization. It's a small fraction of those people who need ICU care or ventilators. In Wuhan, about 5% of confirmed cases got ICU care and about 2.3% got ventilators. South Korea has done the most extensive testing and currently has a 1.4% case fatality rate.

It's also possible that missing mildly symptomatic and asymptomatic cases means that all of these rates of high-level care and fatalities are overstated because the denominator should be bigger. South Korea has done a good job of testing, but they apparently still haven't found everyone because they keep finding about 100 new confirmed cases each day ( https://www.worldometers.info/coronavirus/country/south-korea/ ).

There was a column by two Stanford doctors in the Wednesday WSJ that arguing for a much lower case fatality rate (based on large numbers of unidentified cases), perhaps even lower than the seasonal flu - https://www.wsj.com/articles/is-the-coronavirus-as-deadly-as-they-say-11585088464?mod=trending_now_pos2 . Not sure I'd go that far, but the case fatality rate could be well under 1%.

Said a bit differently, the 7,500 deaths through March 25 in Italy might be from millions of cases: in other words, from some huge multiple of the ~75,000 confirmed cases. That still implies a relatively serious infectious disease, but one of the defining features would be that it overwhelms healthcare systems because it spreads so much faster than the flu in a population with no prior exposure, not that it's necessarily a lot more virulent than the flu.

50% is pretty realistic right now. 3-4% is the mortality rate, when everything goes according to plan. Right now Italy has 7500 deaths our of 17000 cases that ended (most still remain unresolved) and the death curve does not seem to move in the direction of 3-4% mortality like it did in China or how it just became in Germany. Right now USA has 2 times as many dead from Covid-19 as they have people recovered from it. Those are not crazy numbers, those are the realistic numbers, when the pandemic goes out of control, sadly. Hopefully, it will not happen in future. But that hope is thinning by the day, because USA does a bit little and a bit late to prevent the bad case scenarios.

It's totally unrealistic. I would gladly bet you all the wealth I have that mortality rate would be below 50%, if there was a safe mechanism to do it.

50% is totally unrealistic based on current numbers.
We know that true asymptomatics are at least 18% of all infected (Diamond Princess).

We don't have any reason to presume that anyone not needing hospitalization (~80% of identified cases, probably 85+% of overall cases will die).
Also, I highly doubt that even with total lack of hospital care more than 2/3rds of those needing hospitalization will die.

So, overall any predictions of mortality higher than 10% (and that is with full health care collapse) with the current data are totally unrealistic.

I also call BS on the whole idea of "it's actually millions of cases". Iceland has ran tests on 10000 out of 300000 of their population. Not just the people with symptoms, but randomly selected as well. And they saw only 50% of carriers to show no symptoms. So at most, that makes the number of cases double the reported number. That makes my assumption of 50% mortality into 25% and frankly, that's shit as well.

Iceland seem to have hospitalized about 2.5% of their cases, when they they have identified mostly from symptomatic cases (see - https://nordiclifescience.org/covid-19-first-results-of-the-voluntary-screening-on-iceland/ - 12/473), before random sampling added those asymptomatics (though not capturing near all of them in the gpop).

If generously half are asymptomatic, you can guess 1% infections need hospitalization.

About 1/2-2/3 of those hospitalized probably just need beds and monitoring, and not respirators or ICU, so perhaps 0.3-0.5% of infections may need respirators.

Healthcare system overload may be a quite remote possibility with even fairly modest measures to up overall capacity. I previously guessed about 12 respirators / ICU per million in the US. Even if it's 120 its not really too much hard work.

This thing is strongly age dependent, and there aren't enough old people to get to 50% or 25% mortality from this thing. Insanity.

There will be some middle ways emerging on this thing; shelter the elderly, reduced contacts and staffing and opening hours, workers must wear masks and get seralogical tests, everyone who can WFH does so 4 days out 5, limitations on size of gatherings.

Not going to do yo-yo lockdown or perma lockdown until vaccine.

Have you told Trump? Because in the U.S., Trump wants businesses open and churches full in two weeks.

Not gonna happen.

You are right. Though Trump may desire it, the U.S. has a federal system where the president is not allowed to interfere with a state's choices of how to deal with a public health crisis.

The problem is that, sadly, it is imaginable that in states where politicians have indicated a desire to die instead of allowing America to stay locked down, that a governor will not keep churches closed at Easter. Though what is occurring in NY might be enough to change the minds of someone like the lt. governor of Texas in the next couple of weeks.

I as a young male will be ok. Since the disease kills more men than women, this will leave a gender imbalance in my favour. Since the US will be depopulated, we can advocate for polygamy along with importation of hot young Russian women. We can then go on to repopulate the country Paraguay style (after the War of the Triple Alliance)!

There’s a flaw. You don’t really want hot Russian women as either “repopulaters” or as a major demographic.

They are feckless, reckless, improvident and as narcissistic as a Kardashian.

Is Paraguay the model state now? Paraguay?

@K R - "I would say , if 1% of general population (across the country, not just in one quarantined city) is ill, it is probably meaningless to fight and you can just accept that everyone will be ill" - source? This assumption is itself meaningless. See my analysis above on cost-benefits; it pays today to have a 2 month shutdown.

No source. I am just thinking in graphs. With each passing day and each new patient it becomes increasingly hard to control the spread of the disease. And the sigmoidal curve will stop at 60%-70% of the population. In my internal idea, at a several percent, the likelyhood of further spread is too much, we cannot stop it be any measure. But that's just my gut feeling, you need proper models for that.

Well your other "gut feeling" expressed above is that this virus is almost as deadly as Ebola...

Paul Romer has proposed a sane exit strategy for quarantine that improves the economy. Others I have seen seem to be constructing (sometimes elaborate) epidemiological models, suggesting a big push quarantine, and neglecting the exit strategy.


All the negative responses I have seen to the now 3 part blog post seem to misunderstand the innovation aspect in a fundamental way, that the point is to invest now in testing and PPE to prevent a crisis a month from now (which is a *much* smaller investment than fiscal stimulus). The stakes are letting people back to work or forcibly quarantining them for months on end (think of the externalities). If there is a serious criticism to be had, please make it.

Can people rally around "pay by the test" (more for lower false negative rate)?

Less than a tenth of people have died from this compared to the flu so far. So we are a few weeks before the pain of 10-20% of the country infected at once. If it gets that bad, then we can talk about a shift in strategy. Until then, there are opportunities to prepare.

1. You might want to consider whether marginal analysis is the right way to look at it and instead pose it as a problem that involve networks and network analytics for the solution. That is, what are the critical tipping points, or, how many rods in the reactor essentially will determine that the reactor will not explode.

2. Again, using network models, you might want to recognize that different parts of the network have different levels of infection in their "community". Rural South Dakota is different than urban New York, and there are few travel links between them; but, on the other hand, parts of rural upstate NY may have many links between NYC the Finger Lakes region. So, you would want to look at how you treat the Finger Lakes region, exposed to travel from NYC residents, different than S Dakota, even though at one time each may have the same level of infection but different levels of potential infection if limits are lifted.

I don't think marginal analysis is the way to look at it. Examine it as a network analysis problem: think: there is a train running up and down the east coast. What if you lift one restriction here, what does that mean for cities along the route or nodes along the path that connect to other parts of the system.

3. We could see a model of "every man (or state) for themselves" in lifting of controls, where one state doesn't take into account how it is affecting the health of another. Think Mississippi or Florida, where after spring break, infected kids return to an area which has successfully reduced spread in the local community. Again, marginal analysis may not work if parts of the network are the mixing bowls of contamination of other parts of the network.

Thanks for this post.

"And don’t you find it strange that no one has been willing to raise this point before? "

I hang around with some hysterical people so I've already dealt with this. No we're not going to fail. At worst, we do like the Asians and the Italians did (lockdowns, triage, military) but we're not going to get there because as slow and lethargic we've been there was enough advance warning and our lead-from-behind leadership class still did enough to prevent the worst. Hard to stomach for those that want a disaster movie.

The only change I'd like to see in the future is that the stock market needs to crash sooner. Americans don't do a damn thing until it hurts their wallet. I remember comments from people saying the virus isn't a big deal because the stock market is still going up. We shouldn't take cues from there but we do.

USA already is already in the first place in the world for active cases of Covid-19. Right now it has more than China did at the peak of their epidemic in the mid-February (when the lockdown was already in effect). And USA has much more lax attitude towards complete lockdowns than China. 1000 dead vs 400 recovered in USA does not paint a great picture for USA health system capacity as well.

Italy has only been testing people who arrive very sick at the hospital.

Their own medical experts speculate that actual prevalence is up to 10x higher than confirmed cases.

I would add to Andre's point that if you look at Italian confirmed cases by age cohort, there is a massive bias towards identification of the virus in the elderly that does not match Italy's underlying demographic age distribution. The obvious conclusion (since we know from countries with more comprehensive testing that the disease spreads as easily in the young as the old) is that younger people who are infected but have not been tested because they exhibit few or no symptoms is considerable.

First question is whether lockdowns are working. Italy seemed like cases leveled off after two weeks, so we will see in the US soon. If the virus keeps spreading through lockdown, we have failed, and the lockdown can let up basically once medical production like ventilators is ready. In this scenario we have another yearly flu which is way deadlier and the world sucks more now, sorry, chalk one up for the “great stagnation” hypothesis.

If lockdowns do work the question is how to let up. I’m sure there will be some sort of testing system put in place that many public health authorities express confidence in. The bad case is that it doesn’t work, we discover that in a few weeks. We can then either return to lockdown and plan to repeat, or just give up. But we will have a lot more information at the time.

Basically the problem is cyclical. It doesn’t make sense to go from “lockdown” straight to “give up”. “Lockdown” and “attempted relaxing of the rules” can go back and forth, and if at some point we don’t want to do more lockdown we can then declare victory and say we are done.

It’s quite hard now to determine triggers for when to do a second lockdown, because that’s a ways away. That’s the real time we might consider giving up, though. We should certainly have more information then.

Right. There are multiple strategies more promising than Give Up.

And we are learning important knowledge at the fastest rate in history, so the strategy we adopt in the future ought to be better than the strategy we would adopt now.

The experiments done back in 1918-1919 show that ending the lockdown prematurely gets one a much higher spike with more deaths overall. Exponential processes have lead times. Experts have modeled them. We have actual data on handling epidemics. We are nowhere near close to having tamped down COVID-19.

When we run out of toilet paper.

"everyone's got a plan 'til they get hit in the mouth"-Mike Tyson

I suppose it's obligatory to throw in an empty comment about "Trump's incompetence" if you want to maintain cachet in the right circles, but it is disappointing in a supposedly analytical blog. What exactly did he do that was incompetent, given the information reasonably available to him at the time? Sure, we can look back in time and identify things that could have been done given what we know now, but what exactly did you or experts you respect recommend at the time? I'd think that anyone who called President Trump racists or xenophobic for limiting flights from China on January 31 would have the decency just to shut up now. As with any surprise event, the most thought-provoking analysis is Roberta Wohlstetter's classic--"Pearl Harbor: Warning and Decision." She shows that, looking back it's possible to find lots of signals about the coming attack. But, there were lots of signals of other coming attacks, which never happened. The cost of responding to this ambiguous warning made it impossible to protect the Pacific fleet. It's true that there was lots of warning signs about the corona virus in early February, but they were ambiguous and the cost of responding was high. I guess it is possible to argue that FDR's incompetence was responsible for the Pearl Harbor disaster, but Wohlstetter showed the absurdity of that argument.

What did he do wrong?

* He trusted Xi
* He treated it like a political problem rather than a health problem to be solved
* He lied and lied and lied and lied about how serious it was
* He has built a culture of sycophancy that makes it very hard for him to know the actual facts
* He doesn’t read briefing books
* He didn’t follow the NSC’s checklist for pandemics

Producing a meaningless list of buzzwords just emphasizes that you can't actually identify any concrete actions not taken.

Looks pretty concrete to me.

Despite trumps many shortcomings (in my opinion of course), his overall response in the lead up to this does not seem to be notably different than most of the western world. Trump's actions, much like the fact that he is president at all, seems more symptomatic than causative.

Right now states and localities are bidding against each other and the federal government for supplies. This is something the federal government is supposed to prevent so that medical supplies are available and affordable.

We have a shortage of medical supplies. The federal government has long had the power to mobilize manufacturers, transport and distribution to deal with this during emergencies. It has not done so.

We waited forever to shut down international travel. Unlike many nations, the US has physical barriers. We're were building a wall down in Mexico rather than doing testing at airports. If Trump wanted an excuse to keep out folks from Mexico, coronavirus was an opportunity.

We used to have a staff of disease experts ready to coordinate the national medical response. This was disbanded, so our response has been patched together by governors and local health experts working as best they can.

Now, the guy who is supposed to be providing us with some leadership is talking about declaring the plague over on some arbitrary date and devil take the hindmost.

We used to be proud that the US could do things as well or better than elsewhere. Now, we seem to be happy with "not massively worse than elsewhere".

I went back through the posts at MR tagged Medicine and one of the first directly relevant to the Coronavirus outbreak was February 8,


The main thrust was how Trump’s travel bans burnish his nationalist credentials. By that time, in addition to the bans, a CV task force had been established. The reactions - Trump was a xenophobic and his TF was not diverse enough (CNN). He has been way ahead on this. He discovered the utter uselessness of the CDC and the FDA
You advise going “full Singapore” but I don’t see where or when you advised that early on.

The current trope is that Trump acted slowly. That is a complete lie pushed by the Trump Deranged and in constant repetition here at MR and many other places.

Trump did act slowly. It doesn't take a genius to connect the carnage in Hubei with a global pandemic. As the leader, you are expected to lead. Instead he played it down as a Democrat hoax and wasted his time writing "i miss u" letters to Kim Jong Un.

See the DeBlasio tweet in March urging New Yorkers not worry and to go out to the movies.

Trump has done Ok, in spite of the resistance. The press has, as usual, covered themselves in crap.

Well they have priorities right, oh wait..

"G-7 failed to agree on statement after U.S. insisted on calling coronavirus outbreak ‘Wuhan virus’"

It's like you brought a MAGA to a pandemic fight.

The stock market tanked. The job market tanked. People are living in fear of the pandemic. People are living in fear of their next paycheck. The Fed is bailing out the credit markets. The Federal government is bailing out corporate America. Trump has done ok? Don't try to make this cute. This is a shit job and we don't have enough toilet paper to wipe the stain between America's cheeks.

MAGA 2019: Sure Trump is not the sharpest knife in the drawer, but what's the worst that could happen?

MAGA 2020: Sure the worst happened, but that has nothing to do with him! Or us supporters?

Yea, if only Joe "Aqualung" Biden was President ... things would be perfect.

I have already said I'd be happy with Pence, and that is the message that should really sink in.

The Republicans had an out and they didn't take it.

+1 aqualung reference
bet the demos gonna do a bait&switcheroo before november

Oh right.

One tweet from DeBlasio justifies all Trump's lies and inertia and failure to act.

Trump's better than De Blasio...don't set the bar too high

From Slate:

President Donald Trump slammed Democrats for the way they are reacting to the coronavirus threat, saying that their criticism of his administration’s handling of the outbreak is “their new hoax.” At a campaign rally in South Carolina a day before a critical Democratic primary, Trump patted himself on the back for the White House response to the coronavirus. “Now the Democrats are politicizing the coronavirus,” Trump said. “They have no clue, they don’t have any clue, they can’t even count their votes in Iowa.”

Of course, the clueless and biased misread this as him calling the virus a hoax. He was way ahead of them.

Today, the U.S. has one of the highest growth rates of coronavirus cases in the world. Its peers in terms of response are not Japan, Singapore, South Korea or Northern European countries but rather Spain and Italy. Coronavirus growth rates are looking more and more like they are correlated with levels of corruption and basic institutional competence within the government and the leadership of the U.S. government, including Trump himself, needs to be held accountable.

It seems like the people criticizing the federal government's response to the looming threat were right after all.

You really are clinging to this Chinese virus hoax, aren't you?

Cases per capita in the US are currently well below those of the Nordic countries.

Wrong. Finland has less cases per capita than the US.

Ok, Cases per capita in the US is currently below Sweden, Norway and Denmark,

Do you just not know?

Before Trump’s inauguration, a warning: The worst influenza pandemic since 1918

It would be sad if you "TDS" dudes just don't know, but to me this adds up as "forgive anything," and "the buck stops nowhere."

I mean seriously, just think about it. If any error had been made at any level of the Obama government, how much would Republicans lay at his feet? Would they consider that the chief executive had ultimate responsibility?

But now, daily self-contradicting gibberish is treated as some new American exceptionalism.

MAGA 2020, for anybody who still has the stomach for it.

Why can’t we blame everyone responsible? Why must it be a partisan shitshow Banana Republic response every single time? Blame Trump, yes. But save some for the Obama admin, FDA and CDC:

This is the Obama administration that depleted the entire nation’s strategic reserve of hundreds of millions of respirators and masks....in 2009. And did nothing for years to build it back up.

The rational non-insane response would be: Obama had 7 years and did nothing, and Trump had 3 years and did nothing. Both get an F.

Trump owns the failures of the CDC and FDA in 2020 since he’s the Exec.

Full congressional Inquiry, name names. If Trump was in any way involved in the CDC threatening Dr. Helen Chu then begin immediate impeachment proceedings.

If it was a series of career employees then negligent manslaughter charges need to be filed against everyone involved. Ditto for the FDA employees who banned PCR testing for 5 weeks.

Don't lean too heavily on that time machine. The Obama administration made pandemic preparation a top-level national security concern. Trump let his subordinates cancel that and then said "it has nothing to do with me."

Now in terms of stockpiles, sure for historical purposes you can look back at that. How open was the Congress to spending in 2009? You would think it was a prime time for fiscal operations. you think that in the great recession people would not be too interested in deficits.

So did you or anyone you know call for investment in that preparation?

Or were you actually spending your time calling for austerity?

Ah, we do in fact know who you should visit with your time machine:

"For Maine Sen. Susan Collins, a lifelong Republican who believes in limited government, funding for pandemic flu preparations didn’t belong in the Obama administration’s economic stimulus plan. She lobbied hard to kill the money, and as one of only three Senate Republicans supporting the package, she had a lot of leverage with an administration desperate to show bipartisan support in a global economic crisis. "

It's anonymous. He's as partisan as they come. It's not in him to admit that their is widespread fault, because then he can't portray the other side as being uniquely bad.

Some of you treat the CDC like it is a sovereign nation, unbeholden to anyone, and certainly not the president of the United States

to which it reports.

You maintain this bizarre cognitive dissonance that a man can be the President of the United States and also an innocent bystander.


this was 19 days ago...what makes anyone think Trump has a handle on this now?

Let's bookmark this comment for the next time anonymous flips his position: Today he's complaining that Trump did not overrule the career experts at the CDC and gone with his gut instead.

Note too that anonymous is still unable to say exactly what Trump should have done differently.

I did not remotely say that.

In fact if you follow cove77's link you'll see Trump corrupting the CDC's message in real time.

"Q But how can hospitals be preparing if they don’t know how many people they’re going to have to deal with?

THE PRESIDENT: Well, we’re prepared for anything. We’re prepared. We are, really, very highly prepared for anything. And in a short period of time — I mean, what they’ve done is very incredible. And I’ve seen what they’ve done back there. It’s really incredible."

That was March 7th. Prepared for anything.

"Note too that anonymous is still unable to say exactly what Trump should have done differently."

Huh on this?

Follow the plan.


"Executive" departments are only partially under executive control. That has been true since the establishment of civil service protections in the progressive era.

The advantage is the that corrupt politicians don't get to give millions of jobs to cronies and donors. The disadvantage is that the elected leaders (and, by extension, voters) have very limited ability to make changes that go deeper than very top level policy or even to fire incompetent people.

Still seems like we need the Hammer and the Dance policy, as the capacity constraints will be breached at ICUs even if the country (this applies to the US, UK, Germany, France) went into Italian style lockdown (even two weeks ago).

So you need czars appointed to stand over the FDA and build the capacity.

Additionally, the latest data over the past three days (certainly not a reliable trend) is that confirmed cases in Washington State and NY and Florida are not going exponential. Certainly not symptomatic cases. Additionally, the HCX treatment, while not confirmed by a trial (will take minimum 4 weeks to do so) is anecdotally working to reduce fatalities.

Furthermore, there is more credence to the argument that a model population has about 80% either asymptomatic positive test or immune, looking at Diamond Princess, Iceland, the recent Oxford report.

Tyler, at what point do you update your priors? The data is changing. Look at the segmentation of cases amongst different regions in Italy. It's clustering around Lombardy. Some molecular biologists are suggesting this could be due to a higher portion of a certain genotype in Toscana compared to the model population around the world.

The data is changing every day, we should be prepared to update positions once new trends are established. Even if the trend doesn't change, if we do nothing but just sit in lockdown, those deaths are going to come anyway unless ventilators and/or drug treatment is solved. That involves a war time organisational effort, not being global voyeurs to this human tragedy.

If by Oxford "report" you mean "that draft on dropbox from someone claiming to be affiliated with the Zoology department at Oxford that isn't even a pre-print" you need to raise your standards for what constitutes a "report".

Hong Kong is a small district, good Public Health Service, and they are stuttering between shutdowns not shutdowns and shutdowns again, against Covid 19. This is been going on for months.

Once the lockdowns end in the US are we going to go on a wild goose chase every time someone sneezes? Surely, a nation the size of the US cannot do any better than Hong Kong.

I don't see how we can go back to lockdowns after we have stopped alockdowns--- which makes one wonder at the purpose of a lockdown anyway in a country as free and open and large is the United States.

It is okay to change your mind. The lockdowns are a very expensive failure, and the longer they last the more expensive they will be.

"one wonder at the purpose of a lockdown anyway in a country as free and open and large is the United States."

To stop the virus.

See Hong Kong. It won't work.

Besides, thanks to open-borders enthusiasts, we have perhaps 1 million people a year coming across US borders without so much as a light-beam on their forehead....

Trump is the boss of ICE and Border Patrol not Bryan Caplan. It's already Trump's 4th year in office and you're telling me the border is still Swiss cheese?

Tyler, I think you identify why no one (or at least no one that you cite--certainly lots of people are talking about it quite regularly, including one Lt. Governor) is talking about when to "let it rip", as it were: the belief that "we are not facing a dollars vs. lives trade-off, rather the better solutions will improve both variables."

But that's exactly the sort of thinking you would criticize the rest of the time--that is, that there are not trade offs. There is a trade off here: economic activity is valuable and many people engaged in it will be just fine, while some people will get seriously ill and die.

There is a trade off between money and lives. There just is. Unless--and I really hope this is not the case--that all the preventative methods fail, in which case we both have economic activity halted *and* the loss of life. But the idea that there is likely some policy that involves more economic activity *and* flattens the curve at the same time seems like avoiding the kind of thinking you and Alex spend an entire textbook dispelling.

On a side note: I have been playing with a simple "value of deaths avoided", like many others. Using hospitalization (both "regular" and "acute" values) and death rates, applied to the entire population (i.e., assume everyone eventually gets COVID, and the rates of illness and death we have seen are representative--this should be close to an upper bound, it seems), and using value of a statistical life rather than QALYs saved, I get a value of something around $35.3 *trillion* of avoiding the pandemic in the US. Thus, measures that avoid *half* the pandemic are still very valuable. So I would err on the side of "everyone stay home" for quite a while, maybe even four months.

I should mention some caveats though that drop that number like a rock:

a) using QALYs. Not even quality-adjusted, but just valuing life-years saved at $150,000 per year using Social Security actuarial tables (and taking the most generous number of life years for each age decile) drops the value of avoiding the pandemic down to $10.6 trillion. Again, preventing even half of that is worth shutting down for a quarter (if it works).

b) adjusting the *infection* rate really drops these numbers. I tried applying the infection rate from the Diamond Princess (since it's the only population-level data--are there any better estimates?). Obviously, this cuts the number of illnesses, hospitalizations, and deaths significantly. $35.3 trillion drops to $7.7 trillion and $10.6 trillion drops to $2.1 trillion. Not nothing, but it makes the cost of each month of lost economic activity a lot more important in the cost-benefit weighing.

Some additional points (this is getting long. Sorry):

i) I include hospitalizations--both "regular" and "acute"--using a cost of about $16,000 per "regular" stay and $32,000 per "acute" stay. Don't have the citations on me, but once I get my distance learning lectures recorded, can put this together if anyone's interested.

Hospitalizations, using those values, are just trivial in dollar terms compared to deaths. In the most dire scenario, we're talking $0.5 trillion.

ii) I think Noah has a good point on long run reduction in quality of life. I would add to his quality adjustment the spillover in deaths and ailments from other preventable ailments if hospitals are overrun. It is my understanding (please correct if untrue) that those additional deaths are *not* included in the mortality figures from others' projections. Factoring those in requires merging these sorts of calculations into something like the agent-based model that changed so many people's thinking.

iii) In general, I think any inclusion of the tail risk that this gets much worse should also consider the tail risk that we are currently doing more than is needed to avoid more deaths and illnesses. But that just means that the whole estimate has higher variance. In that world, which do you think matters more: trying to save existing people today, even if it reduces economic activity (and possible economic growth for a year or more), or allowing for greater economic activity today (which may affect future generations through compounding)? It seems to me that, based on Stubborn Attachments, you would favor the latter, and thus want to "let it rip" sooner.

I think I favor the former. I started putting those numbers together to say "I think lockdowns are dumb, I think we should live our lives." And while I think the lower-lives-saved scenarios are probably closer to reality, we just don't know. Let's wait a few more weeks, just in case.

'Tyler, I think you identify why no one (or at least no one that you cite--certainly lots of people are talking about it quite regularly, including one Lt. Governor) is talking about when to "let it rip"'

I was going to make a similar comment. In addition to Texas's Lt. Governor, Trump is implicitly saying that the economic costs of all of this physical distancing (the WHO is now telling us to call if physical distancing instead of social distancing) is becoming too high, compared to the lives saved.

And they might be right but I agree with your basic estimates that based on our current knowledge, we are correct to continue the physical distancing measures and should keep doing so for at a minimum two weeks, probably four or more -- while collecting data and tracking the numbers as best we can to try to gauge where things stand.

Maybe even four months, but I'd want to see the numbers after the first and second month and then see the modeling to determine if four months provides more benefits than costs.

There will be absolutely no further progress against Covid-19 after two more weeks in the NYC region. And based on all previous experience, there will be absolutely no way to ignore what is happening with various what-if exercises.

Unless it is along making further decisions along these lines -

Hospitals on the front lines of the pandemic are engaged in a heated private debate over a calculation few have encountered in their lifetimes — how to weigh the “save at all costs” approach to resuscitating a dying patient against the real danger of exposing doctors and nurses to the contagion of coronavirus.

The conversations are driven by the realization that the risk to staff amid dwindling stores of protective equipment — such as masks, gowns and gloves — may be too great to justify the conventional response when a patient “codes,” and their heart or breathing stops.

Northwestern Memorial Hospital in Chicago has been discussing a do-not-resuscitate policy for infected patients, regardless of the wishes of the patient or their family members — a wrenching decision to prioritize the lives of the many over the one.

'So pick your “no return” point' - 6 weeks ago? 8?

After this initial spread and its effects recede, discussions based on a lack of experience with a novel disease will much easier to ignore. So much time spent in speculation, not preparation, even now.

Viruses don't care. And for those thinking that being spread out helps, one of the worst effected areas of France is much less dense than the entire DC metro region, though it includes cities such as Strasbourg or Muhlhouse.

The question cleverly left unasked, because those who agree with the post will ask it any way, is "What if according to Trump's estimates we are already past the point where the virus needs to burn through"?

How many have noticed that Tyler, while far to the left of, say Tabarrok, has been softer on Trump than most libertarians?

There is a minority of leftists who support Trump because they see him as a pacifist, but there is a still smaller minority of leftists who go soft on Trump, not because they consider him a pacifist, but because they are seduced by the proposal that Trump may be a genius.

Wait, you're saying TC is a leftist?

Is it imperetive to the point at which you start letting people die now?

I suggest caution with this..They're could be costs to this too. Human emotion and desire to save lives isn't just wishy washy soft heartedness. It prevents for the "greater good" attrocities like those committed by nazi's, communists and the British Empire amongt others.

Trying to be hard headed has slowed Britain and US's response and will cost lives and will not be looked on well by history.

Maybe there is a point, but be flexible, wait, and add a bit to whatever you think it is.

>  will not be looked on well by history.

I'd take that bet.

When we actually have the data. That's when I change my mind. Get the data. Who has it -who has had it- are they immune now- what is the actual mortality rate- how effective is what we are doing in trying to save lives.

We're in a "fog of war" situation where urgent decisions nevertheless have to be made. You go to war with the data you have, not the data you might want or wish to have at a later time.

Well since you are a 'captain', wouldn't you slow your ship down if you were in the 'fog' and had really inadequate data telling you where you were steering it? If not, I would demote you to 'crewmember slime'.

What if your "ship" is actually an airplane and you don't have the luxury of hovering in mid-air until the fog clears?

A responsible captain would try to keep the passengers as safe as possible even if that meant diverting the flight to a less desirable airport where they had to wait and board buses to continue their journey.

Sometimes you just shouldn't worry about the schedule.

The numbers thrown around here are preposterous. The actual infection rate is at least an order of magnitude above what is confirmed. That also means the mortality will be at most .5% and probably a fifth of that. We are flying blind without a blood antibody test on a large and random sample.

Check out the Oxford model which says half of the UK has already been infected. It's credible. If it's true this will be rememoas the great hysteria of 2020.

What Oxford model? I've seen a draft that is allegedly from people at the Zoology department at Oxford that isn't even a preprint. The websites of the working groups involved don't mention it anywhere. The email address included for further inquiries is not at oxford but at some sketchy website. I didn't read the full thing, but it sounds like something along the lines of "we should test more to make sure this hypothetical scenario isn't what is actually ocurring", and I don't see anyone comparing it to the actual new data coming out of e.g. Italy (if half of Italy were already infected, the social distancing stuff wouldn't have worked).

It's quite a speculative model, and the point of it is really to show how the same outcomes can be consistent with a range of underlying true models, but they note 50% is an extreme.

This is kind of a pathetic response tbh. The Zoology department is where they have their faculty who understand and model the spread of disease. It is led by their Professor of Theoretical Epidemiology. It has contributing professors from their Nuffield Department of Medicine.

Re; Italy, the cases are regionally clustered. Social distancing may work better in places where the infection isn't like 10-20% or something.

South Korea has continued extensive testing of their population and is up to 341,000 negative tests run. On the other hand, they have slowed the spread of covid-19 dramatically and therefore provide one of the best estimates of mortality we have. That estimate is currently 131 deaths out of 9,241 cases or 1.4% mortality and that ratio has been relatively stable.

Given that the U.K. has 465 deaths from covid-19 already, if one half of the population managed to get infected without anyone noticing, that probably means we are 5-10 days away from what will be a catastrophic loss of life and collapse in the health system there.

They are testing for the virus. Not the antibodies. Check out Medcram 43 on YouTube if you need to understand the difference.

It pays to be more humble. I agree more with Jerry Brown above that we don't have enough data on prevalence. You seem so sure that the mortality rate will be that low and that everyone has already been infected. What happens if you're wrong?

The guy who made the model for UK 500k deaths that made them go on lockdown has flipped. Now says it will be less than 20k - half of whom would have died within a year anyway.


Did we just crash the world economy and cause 2M+ extra hunger deaths among the world's poorest?

Sure looks like it.

What exactly is the "Democratic Party public health establishment"? Are you claiming that public health experts are part of the Democratic party?

At some point we need to talk about geopolitics.

China has beaten the epidemic. No, they aren't faking that. Their economy will now roar back to life. Meanwhile Western economies will collapse at some point without a "premature" reversal of policy. Stimulus is a stopgap, not a cure.

Geopolitical power goes hand in hand with economic might. The Soviet Union stopped being a military superpower not because they forgot how to build weapons systems, but because their economy fell apart.

At some point we might blink and wake up in a world where China is the world's sole superpower. I don't mean decades from now, but before this decade is halfway through. After all, the Soviet collapse was pretty abrupt.

"Lockdown forever" is just a modern version of "better red than dead". But historically, enormous bloody sacrifices have been made to avert unfavorable geopolitical outcomes. Indeed, it's practically the sole reason wars are fought.

Having said that, I for one welcome our new Chinese overlords. Abundant spare time offers a wonderful opportunity to brush up on their language.

PS, we'll all die in the 2037 super-plague regardless.

+1. have been thinking there is a good chance we're headed for post colonial status very soon

Just yesterday, though it seems much longer, I wrote:

"I think everything Tyler says it's true, and my take away (whether or not it was the Straussian message) is that we are screwed."

That's because speed was of the essence and I saw too much confusion in the response. That includes premature calls to end the shutdown, and even worse, "let the olds die for the sake of the economy." That second idea made too many appearances yesterday for comfort.

Even if those bad ideas do not take hold, they must dilute the effort.

So I am not sure I am ready to accept Tyler's optimism they were even starting from a full push for social distancing.

Perhaps a bigger pessimist would wonder how long we can support a messy and partial shutdown, before saying to hell with it.

Note that a lot of good things are happening, with protective gear, testing, treatments, and vaccines. But they are in a race condition with the onslaught. It was not all in place for New York. YMMV.

human kind
Cannot bear very much reality.
Time past and time future
What might have been and what has been
Point to one end, which is always present.

T. S. Eliot, "Burnt Norton"

New cases confirmed today were contracted, on average, around two weeks ago - add up the time from infection to onset of symptoms, the time from symptoms to testing, and the time from testing to results. The backlog of test availability is further skewing the "confirmed cases" growth curve in the US.

People are reacting to the daily data as if it's a reflection of current policies, when in fact it's a reflection of policies and behaviors from two weeks ago. The experience of other countries with similar policy shifts enacted earlier than the US's suggests that the case growth in the US will greatly dwindle by two weeks from now, which means that the *current* infection rate is already substantially reduced. But people can't perceive that because they're looking at the daily new cases as some sort of scoreboard and referendum on how we're doing.

That's a good reminder. If we take 14-day quarantine as the safety limit, then we might take 7 days as average for onset? I don't know if there's better date on that.

But regardless, a week ago much of the nation was still partying on. We should see some moderation a week from now.


Most estimates are at 5 days median from contraction to symptoms. Some very large percentage fall in between 3 and 7 days.

There have been outliers in excess of three weeks but with self reported date of “onset of symptoms” it’s hard to know.

The new cases were generally not contracted two weeks ago. The infection to symptoms curve is probably a log-normal curve with with a G-M a little less than 7 days. Most of people with severe acute symptoms were infected 10 days ago or so when everyone was partying before the recommendations and controls for social distancing went into effect. How many of the NYC firemen and policemen who are now infected were in a crowded bar on St Patrick's day or a few days before?

considering how incomplete and inconsistent the testing data is, I'm not sure we should put much stock in it at this point. If we see the trajectory of deaths start to change, that will mean something. But even that will probably be deceptive because the death rate will increase as health care capacity gets exceeded. Random testing seems paramount. Arguments that we can't do it because of limited testing capacity are short sighted. If you *think* you are infected, you stay home, which most people are already doing. If you are in hospital you will be treated based on condition, not on diagnosis. There are probably cases where some targeted testing is still warranted, but it seems like the best use of testing capacity at this point is random sampling.

The health establishment seems to purposefully trick people. Their intentions are probably good. They want to "flatten the curve" and get some control over the pandemic.

But they're misleading people into believing that the virus can be suppressed out of existence or that a vaccine is just around the corner and once we have that, we won't have to choose between an economy and lives.

No experts believe a vaccine will be ready in less than a year and it's possible a vaccine will never be available. Even if we get a vaccine, it may not be even 50% effective or only good for a couple years.

So even if the mitigation strategies now being employed are successful, it's likely the lesser of two evils will need to be chosen at some point.

Personally, I don't think it will be an explicit choice. I think people will just begin to behave differently and the limits of risk people are willing to accept will change.

No one will actually face their own hypocrisies. They'll just social media shame people differently.

Nurses on shift in New York hospitals are wearing trash bags as protective gear, because that's all they've got.

Damn right they want to flatten the curve.

Tyler writes:

"But we are economists, not mood affiliators, and so we must address the classic question of 'at what margin?'"

The "must" means:

I affiliate with more meticulous reasoners, and so should you!

The relevant margin is not mood affiliating versus not mood affiliating. The relevant margin is affiliating with X versus affiliating with Y.

What coronavirus strategy?

A problem with half measures, in this case a partial lock-down, is that for those in early lock-down, the duration is much longer, the potential economic suffering much greater, and the tolerance for it greatly diminished. All of this is compounded by reckless, irresponsible behavior. It has always baffled me when there's a blizzard the thousands of folks stranded in airports. Did they not read or hear that a blizzard was on the way? And now the thousands of people stuck on cruise ships or in foreign countries because they chose to ignore the risks during a health crisis. Are people just ignorant, blithely going about doing what they want to do without regard to the consequences, to themselves and others? I'm not the first to point this out, but it seems young people (twenty somethings) and seniors are most likely to ignore the risks of social interaction. Sure, young people feel invulnerable, but what's the explanation for seniors? My point is that adherence to a mandatory or voluntary lock-down varies by cohort, which makes the lock-down's effectiveness difficult to predict (although I'm sure some predicted that urban centers filled with young people were at greater risk).

A few hypotheticals:

1. What happens to the economy, if during a burn-out strategy, health care workers walk off their jobs in mass. Does anyone think that the markets or economy will respond favorably? What about everyone's relatives being treated for cancer, rental failure, stroke? Does anyone really think that the rest of healthcare functions independently of what is happening in the emergency department?

2. What happens when some employee shoots their boss rather than work in a dangerous situations? Telling everyone to go back to work does not mean will meekly accept the proposition. Many employees will rebel and go on social media to retaliate against their employer. See Gamestop for an example.

3. What happens when the plaintiff's attorneys start suing every company that fires employees who refuse to work in dangerous situations or even just force their employees to work in dangerous situations?

4. What happens when the rest of the world refuses to do business with the U.S. after a burn out strategy is adopted? Can a multinational corporation really treat employees in Europe with concern and employees in the U.S. will callousness?

5. What happens to the companies who embrace the burn out strategy while irritating many of the potential customers and alienating their employees. What happens to a company's reputation when an employee commits suicide rather than be forced to work in an unsafe situation?

What is absolutely amazing is how economist seems to refuse to think of any negative consequences of their burn out strategies. They seem to see every down side of public health and none of the benefits.

My sister was diagnosed with terminal cancer. When sharing her news with someone she would often get the “ you can get run over by a bus tomorrow” response. She told me she was tempted to rent a bus and run over those people. A lot of this “let it burn out” nonsense assumes that people won’t rebel. If I get the disease and am refused treatment because of my age, as is implied, I’m getting a shotgun and looking to find the first economist to take with me on the journey.

Though the actual change is from March 21, since that date all patients older than 80 are only being provided palliative care before they die.

A strategy involves looking to the future and making choices. As of March 21, the French were completely out of choices in dealing with the number of patients arriving at hospitals. They did not change strategy, they have only reacted. One assumes further changes over the next days and weeks, as the number of cases continues to increase. The increasing number of deaths over the last few days in France certainly includes those who received essentially no medical care all, and thus died promptly.

The damage the virus causes overwhelms any attempts to make a strategy in the middle of reacting to it.

Good pount. My personal time is 1 month from today, and gave enacted extreme social distancing for the past two weeks. At that time, I will proceed normally through the world, and no longer choose to work remotely. 6 weeks self isolation is sufficient contribution to the social good.

Who knows if 6 weeks is right, but I don't believe anyone should set a calendar date now.

What you should do is define conditions.

Will you go out if the death rate in your city is still climbing? Will you go out if doctors and nurses in your town are still dying?

This is still a collective action problem. "We" what hospitals to operate smoothly and safely.

Could it be that President Trump actually deserves a kind of credit for rather impolitely, brutally, and vulgarly thinking ahead and then saying out loud what the intellectuals and the experts are unwilling to think and say? Could it be that politicians, for all their many flaws, are actually better decision-makers than experts in a crisis? Could it be that two New York politicians willingly fighting with each other in public (e.g. Trump and Cuomo) is the best decision-maker of all?

Remember when FDR was considered a kind of amiable simpleton, a "Gentleman's C" out of central casting, by the New Deal Harvard policymaking elite? How did that turn out?

For 10,000 years, Man has been tempted to treat gibberish as Oracle. And sometimes we fall for it.

Indeed, the Obama years were full of such foolishness.

Tyler I suspect that you are committing the hindsight bias as well as failing to grasp implementing a policy in the real world is more than dreaming it up when suggesting that the Government should have followed the Singapore model (better approach Taiwan). You fail to appreciate or are willfully failing to acknowledge that the "Singapore" policy was developed long before this outbreak. It was merely implement for the nth time a few months ago. We see the same pattern in Taiwan, S Korea and Japan. Bashing the administration is simply childish.

I would also suggest that government policy should be framed as decision making under uncertainty. This means we need decision objective (however specified), a set of decision variables, and constraints. What complicates matters in this situation is that we do not know the states of the world we potentially going experience, nor the choices we have available to us. This makes assessing choices difficult and formulating a policy challenging or nearly impossible to do without it appearing ad hoc.

I would suggest that the correct decision sciences approach is see the current set of options as one-off decisions in a world of extreme uncertainty. In these cases, minimizing risk of ruin is a real consideration.

"You fail to appreciate or are willfully failing to acknowledge that the "Singapore" policy was developed long before this outbreak. It was merely implement for the nth time a few months ago. We see the same pattern in Taiwan, S Korea and Japan. Bashing the administration is simply childish."

It is easy to find articles going back at least the past twenty years warning about pandemic risks and the need of governments to be prepared. Governments need to prepare for pandemics the same way California needs to prepare for earthquakes or Florida needs to be prepare for hurricanes -- these risks are part of a common body of knowledge with which people in leadership positions should be expected to be familiar. Another related risk we seem to be doing nothing to mitigate: antibiotic resistant bacteria.

With that said, the solution was easy and remains so: call up the SARS and H1N1 veterans in these countries or any number of American infectious disease experts and ask for advice on putting a proper pandemic risk mitigation system in place. The good news is that some countries and people clearly have the answers and the U.S. government just needs to be run by people smart enough to listen.

I'm sure the lack of preparedness predates the Trump Administration but you know where the buck stops.

This 2009 plan does not mention ventilators, but it does contain these words:

"Enforce distribution and use of PPE and
other personal protective measures
throughout the workforce."

You can find it as a PDF, the title is:

"Centers for Medicare and Medicaid services

Pandemic Influenza Operations and
Response Plan"

I think we had plenty of plans within our own government that we didn't use.

" . . . (but of course offer him health care)." At a price? Doesn't that go without saying?

Has anyone else noticed that Tyler's comments seem to have a shrill tone these days?

He and a lot of others should have done this in 2019, and stood for impeachment and removal.

A surprisingly large number of smart people thought they could whistle past the graveyard.

(In case you haven't noticed, today's mood is "no quarter for apologists.")

I'm seeing urgency but not shrill. Why do you say shrill?

We have to fight until we have random testing of the community. How are individuals supposed to asses their risks and make appropriate decisions without knowing the risks.

We are buying time... time for what. Testing and treatments. Right now we are just going about this blindly so governors are making risk assessments for us. Clearly the is suboptimal, but without knowing our risks, how can I make informed decisions?

Possibly, just possibly, a 58 year old man is starting to weigh the odds that the system he has supported so diligently just might be in the least concerned if he died.

Especially if he is reading the comments here.

It's not obvious to me why the USA strategic response shouldn't be based on "local/regional" data. It seems obvious that if Podunk, WO is 'virus free' that physical distancing is a waste of time there. There's at least one recent paper claiming R0 of 4. Cutting this in half would only slow the pandemic (cet. par.). It seems obvious to me that timing is everything here. Respirators have been id'd as a bottleneck, but so have PPE, swabs, medics, etc. etc. etc. Ramp-up of mfg (of tests, of PPE, of quarantine & treatments centers) are unlikely to be exponential (at least, not for very long). I read that we don't have a vaccine for SARS even after all these years, and I read that the potential side-effects of a "bad" vaccine could not surface for years - which to me suggests that any "short-cuts" to clinical approval could easily do more harm than good (in the long term). We've got an exponential bug, and our responses will likely mostly be linear; wanting "better data" is like wanting to know the results of the Sunday game on Saturday. I wonder why similar viruses haven't been considered useful like cow pox was for small pox, (that would seem something we could try, especially on some healthy youngsters...).

I should also add that the implicit assumption here seems to be that the "current strategy" consists of just a single phase (physical distancing and mfg. ramp-up). This seems to be a childish (naive) assumption, and suggests that the thinker is an outsider looking in. Adults know that timing is everything.

It might be the right time to reconsider the axiom of American empire: that economic growth is the end-all and be-all.
For one example, I don't think the decades-long shrinking of Japan has made many of its citizens less happy or even less well off.
And to answer the question posed about 'when': the post by Tabarrok following this one answers it quite nicely.

Per capita Japan has been growing quite nicely. Still there are serious tensions.

Thank God we can drop the "Straussian" trope for a while at least! It always seemed to me to be intellectually dishonest: a pundit could predict X and then if Y came true, assert "My prediction was Straussian, I was actually right!" We need CLEAR and DIRECT communications, maybe not just while in this viral foxhole, but going forward generally. I am getting really really tired of "dog whistles," "Straussian," "coded," and all the rest. Just speak what you believe.

Um, the BBC's headline on its website - today. Right now - is "Virus batters US economy as Europe deaths sore"

Checked the numbers. In Spain, Italy and France, new deaths were down from the previous day. Italy is now clearly over the worst. Moreover, in no country, anywhere in the world, are covid-19 deaths even approaching what they would be in a typical flu season. Italy seems on course to match it, but Italian deaths are skewed by the fact that many died for lack of treatment (insufficient number of ventilators).

Right now is perhaps not a good time for a shift (except in having headlines match the event), but in one week you can expect to see a steep decline in European deaths. That'll be high time.

You really don't stop, do you? https://www.worldometers.info/coronavirus

Italy -
March 25 - 680 deaths
March 24 - 740 deaths
March 23 - 600 deaths
2000 dead in 3 days, and deaths have not declined between March 23 and March 25, they have increased.

March 25 - 650 deaths
March 24 - 680 deaths
March 23 - 540 deaths
Almost 1900 dead in 3 days, and deaths have not declined between March 23 and March 25, they have increased.

Italian deaths since March 20 have been trending downwards, not upwards. Anyone who is remotely numerate can draw a best fit line for daily deaths in Italy, and see a downward slope since March 20. That's a full week, and not your 3 days.

I confidently predict that Spain will not see another day as high as 680 deaths (only 442 today, by the way). If it does, I'll eat my words. If it doesn't, you should eat yours.

I have not any predictions at all, just quoted the death total of the last three days in Italy and Spain. Three days which just happen to represent around 8% of Epstein's predicted total global death toll of 50,000.

The cresting of Italian deaths isn't a prediction. It is visible from the data. Italy is always harder hit by the flu than any other country on earth, and the deaths this year were compounded by the fact that they were concentrated at one time, preventing treatment of the critically ill. Neither of those factors obtain in the US.

And, for the record, leaving aside Epstein's prediction for total global deaths, the seasonal flu kills half a million people each and every year, and we pretty much deal with it. The current coronavirus death toll (3 and half months since it started) is just over 20,000.

Can we agree that there is no obvious crisis until global deaths are at least one fifth (100,000 deaths, that is) of the toll taken by seasonal flu? I'm prepared to say it shouldn't be a crisis until global deaths actually exceed those from seasonal influenza. Call me crazy.

Global crisis? Who knows. A crisis in Italy, Spain, and France right now? Absolutely. A likely crisis in Germany and the U.S.? Sure, that seems reasonable.

Are you seriously suggesting that somewhere like the UK should be ignoring what is going on around them since their number of flu deaths is pretty much the same as every year, while the number of UK covid19 cases and deaths continues to climb.

Not "ignore." However, the extreme measures taken in Italy are not warranted everywhere. Public health information (wash your hands frequently), together with limiting "large" public gatherings, as well as targeted measures for the elderly, combined with extensive testing and isolation of positive cases, seems like it could do most of the work.

"Public health information (wash your hands frequently), together with limiting "large" public gatherings, as well as targeted measures for the elderly, combined with extensive testing and isolation of positive cases, seems like it could do most of the work."

Yeah, "seems like". Thanks.

"Italian deaths since March 20 have been trending downwards, not upwards."


Mar 20: 627
Mar 21: 793
Mar 22: 651
Mar 23: 601
Mar 24: 743
Mar 25: 685
Mar 26: 712

I don't dispute the numbers. I relied on them to draw my conclusion. Today's deaths were lower than 2 days ago and much lower than 5 days ago. That, I'm sorry to tell you, is trending down. It is easy to see just by looking at the bar graph on worldometer website, from which these figures are drawn.

The "best fit" line for the last week slopes down, not up.

Here is evidence that European leaders are also beginning to ask these questions:

"The sky is falling. The sky is falling."
There's a reason we read Chicken Little and a lesson to be learned. Get a freakin' spine!

The virus is now endemic, here to stay, at least for a couple of years until a total vaccine is available. The constraint is hospital beds for those with acute respiratory stress. The unknown variable is immunity and incubation period seems to be a couple of weeks. I am not hearing about re-infection, so I am taking a guess that immunity is seasonal..

Then I apply abstract tree, the virus and anti-bodies want a round tree trunk in their trading pit, what is the optimal number of immune and infectable to keep both anti-body and virus the most adaptable.

Because of the mismatch between incubation time and immunity time my model predicts outbreaks of the virus every so often, it will have to move about to keep the tree trunk round and make a flourishing tree, virus the branches, antibodies the roots. It will show up in neighborhoods now and then until we have a vaccine.

How much now and how much then? Wait a few days watch for re-infections, and watch for better stats on incubation time. The virus is search for spots where immunity has worn, the anti-bodies look for spots where the virus can be attacked. A self sampling system, an abstract model works.

In betting or trading [pits we use the no arbitrage condition to find equilibrium. Do the same with the virus vs antibody. Whenever there is a synchronous event, like regular vacations, the virus will find spots where most immunity in a small group exhausts. But then two weeks later it is restored.

The anti-body and virus sample each other, each looking for arbitrage where an opening for an influx exists, making round tree trunk, or in finance, making the velocity equations work.

Assuming seasonal immunity, I predict the virus will show up in one out of twenty neighborhoods, mostly those undergoing some synchronous redevelopment that causes mass movement of people.

Given my analysis and the infection rate in New York, I expect New York to reach this equilibrium in eight weeks.


When China got its foreign travelers home, it a mostly within a single week on charter flights, a synchronous event. That event left an arbitrage spot for the virus, and it attacked and now is followed by a spike in seasonal immunity. Incubation and immunity times are the key ratio, they are ill matched and outbreaks will take an asynchronous orbit.

I should note.
I am optimistic that incubation times can be cut back to a week, that is a week transition from infected to immune. This is with off the shelf medicine. I also think immunity periods can be boosted using off the shelf immune boosters.

Consider the dynamics of the anti-body to virus 'market'. The paradox is the model equilibrium occurs when both antibody and virus see the greatest arbitrage moment, flocking occurs. But the immune period is much greater than the infection period, the virus is soon overwhelmed and extinguished in at the outbreak. But a leaker escapes, and random patterns eventually find another neighborhood that has been isolated too long, the outbreak moves around.
At equilibrium, then, the anti-bodies and virus are most active, most robust. We know that point, we know the no arbitrage condition as an equilibrium. It is never reached but the virus and anti-bodies always seek the optimum ratio. This is a self sampled system, new math.

This is positioned as what's our line if the defense fails, but I'm also really unclear what people see as the line if the defense succeeds. That is, I see actually nobody talking about conditions to end economic shutdown without being castigated as stock portfolio pumping denialists of some sort. You can't lock down the US economy for months on end. It'd be like taking the GDP of Canada and throwing it in a volcano.

The rosy scenario was that families with members less than say 50 come out first, and olds stay home. But if many hospitalized are < 50 that doesn't work. Then you have to consider much more constrained versions of "come out." Like, "everybody wears masks and gloves in public."

Hmmm ... at what point do we keep The Federalist and ban twitter? I think we passed it already.

When medical advances, drugs or treatments change the situation with regards to ICU capacity vs cases.

Identify the issue here. It isn't people getting the virus. It is the proportion of people getting the virus who need intensive care to survive.

How many people die every year in the US or Canada? In the US 2,800,000 per year. That is the base rate. How far off of that are we?

The problem isn't that people are dying. It is that the care facilities are overwhelmed.

A story. In the 90's when Canada was sorting out how to pay for the government we had, the medical system had less capacity than needed. Lots of mistakes, misjudgements, bad policies, etc. The point of contact with the medical system was the emergency ward for almost everyone, they were packed full, and there wasn't hospital capacity behind it to take the people coming in, so the stacked up in the emergency wards. Ambulance personnel couldn't hand off their patients, so were sitting for hours waiting. One impatient and rational ambulance driver decided to park his vehicle inside the hospital waiting room, going through the windows and wall.

The issue is capacity. Fix that with medical treatments that keep people out of ICU. Fix it with testing and protocols to prevent infections of nursing homes.

The economic shutdown is only to give time to the medical people to come up with a better answer.

Sadly, that will only be clear after the U.S. experiences it, just like Italy and Spain and France.

1. Assume national lockdown began on March 23rd, so it should end by Friday April 3rd to catch the first wave of cases and give people time to settle down.
2. Mandate companies allow all individuals over 65 or who are considered part of the at risk population by guidelines or a doctors note, or who haven't been allowed to return per point 3 below to work from home or claim full unemployment benefits from now until we have a vaccine.
3. Allow all businesses where social distancing is possible to reopen on April 6th with a specific set of sanitary guidelines and allow people to return to work by randomly drawing first letters of last names (maybe in batches of 9?)
4. Have the government purchase 2 year term $ 1 million life insurance policies on behalf of everyone in the at risk population in (1) to compensate them for their risk and as a gesture of good faith.

Wash your hands, cover your mouth, get back to work, and end this totalitarian experiment?

Whatever we do, a deep or prolonged depression carries with it so much potential ugliness that those who pretend that it's isolated to the S&P 500 or unemployment are kidding themselves. This is the condition under which we could see the further advancement of authoritarianism, increased likelihood of violence and war, and even greater health risks. These are very bad outcomes and it would be truly tragic if they come from a self-inflicted wound.

>You can’t just condemn Trump and stop thinking about it.

Especially because if you are condemning Trump, you're not thinking at all.

I question the premise "what if we have made no progress against corona" - we are making progress every day, even if small and not showing up in our favorite log graphs of total cases.

Homemade masks aren't 100% effective but they are not 0% effective either. We're learning how to socially distance and perhaps we will be better at it in a week or two, so we can resume economic activity. More companies will adapt to remote working and delivery. (As a technology worker I have been busier than ever these last couple weeks with nonstop zoom calls. In fact my company is still hiring.)

In the "long term" - 2+ years, an eternity - we will clearly have herd immunity, either from exposing the entire population at some variable high cost in lives, or through a vaccine or effective treatment.

Every day we delay, flatten the curve, we learn more and get better. Technical change is endogenous.

The homemade mask thing is completely stupid. 3M can make millions of masks a month, far more than a bunch of volunteers sewing at home. And they will, now that congress has removed the regulations that prevented them from selling industrial use N95 masks into the medical market.


My less intelligent and barely related take on this: I think all the "sorry grandma, we have to grow the economy" snark is dangerous because it misses an important point: human costs cannot simply be measured in how many people live and die. There are tons of costs to lowering the body count which need to be taken seriously: joblessness, homelessness, closing businesses, depression, etc...

I was making this exact point in Alex's thread above.
Thanks for making it more succinctly.

I'm totally infuriated by all the Facebook memes going around basically saying Republicans think we should let millions die for the sake of profits. Just take seriously the idea that "the economy" isn't just a bunch of corporate profits and it actually has human impacts. Sheesh. YOU PEOPLE USED TO KNOW THIS.

Dummy here. Can someone explain what "straussian" means in this context? I've seen it mentioned a few times on MR and am still confused after a google search.


The GDP argument might have made sense 50 years ago, but every year the value of an hour's work as a share of GDP has fallen. It's down about 40% since the 1960s and keeps falling. Given the irrelevance of the GDP, any economically rational American would figure that scraping by with a crappy, low pay job with no benefits is not worth all that much more than dropping out of the economy entirely and just scraping by catch as catch can. The GDP could double and they'd be lucky to get a ten cent an hour raise. If it fell by half, they would only be ever so slightly worse off. They have no real stake in raising the GDP. Why should they risk their lives and health? It's not as if they can afford the co-pay and deductible now. Their only hope is to try not getting sick.

"If [GDP] fell by half, they would only be ever so slightly worse off."

That's absurd. We're seeing the impact of a decline of less than half, and it's mass unemployment. Given our particular circumstances, it's concentrated in industries tied to travel (hotels, airlines) and face-to-face services that aren't emergency healthcare or selling food in a way that's consistent with social distancing (bars, restaurants, barbers and salons, non-food retail, etc.). The people who are hanging onto jobs in these industries have them based on the promise that we'll see a sharp rebound in economic activity after the worst of the pandemic passes.

If this decline was sustained, it would also feed through to manufacturing of all durable goods as well as the professional jobs that people are now doing from home. Much of the tech industry, for example, consists of either ad-supported consumer services (at risk as advertising spending has tended to stay in a narrow range as a percent of GDP for decades) or software with a large percentage of sales to business (Microsoft, Oracle, etc.).

Much about little. Let's destroy the economy because people don't understand math. The mortality rate is not high enough to panic as much as the government is panicking. Go look at the Italian data and you wee more people over 90 died than under 60 and that three quarters had two or more serious conditions. If someone who is undergoing cancer treatments and dies has the virus in her system, what was the cause of death?

Given how infectious this coronavirus is, herd immunity would require at least an 80% immunity rate. (It's probably higher than that, but this is just an approximation.) That's 240M+ people infected. Assume 0.5% of those infected die. That's 1.2 million deaths to get herd immunity. (It actually would be higher than that, especially if our medical system is saturated.) If 5% of those with the disease have to be hospitalized, that means 12M hospital admissions. (Again, it is probably higher than that, assuming our hospitals could actually admit them.) Herd immunity means a lot of dead people and a lot of seriously sick people. I'll gloss over the likelihood that many survivors will never recover proper lung function. Polio didn't kill everyone who got it, many were paralyzed.

How long would it take to get to 80% for herd immunity? The number of cases is doubling twice a week right now with our current measures. We have about 80K cases, so we need 3,000X or about six weeks which gets us into mid-May. This is a sloppy analysis, what physicists call a Fermi solution, but if our current isolation methods don't work as they have in past plagues or we drop them, we are looking at possible stabilization some time in May with over a million dead. Not all of them would be old people.

For many on this forum, this is a glorious prospect. Think of all the graves to be dug, services to be conducted, estates to be settled and so on. It would be a boon to the GDP. Then again, many people watched the recent Marvel movie Endgame and were rooting for Thanos.

The way I see it, the cat is out the bag now and we're all going to get the virus. Well, a lot of us are. All the social distancing isn't about stopping the virus anymore, but slowing it so as not to overwhelm the healthcare system. So, to the extent possible, we should control the speed of the spread to keep within the bandwidth of the healthcare system, but not slow it so much that there is too much excess capacity. And when the models predict that people going back to work will not overwhelm the healthcare system, we should do so. Even if more people get the virus than otherwise would.

This approach would maximize the number of lives saved while minimizing the economic impact.

Comments for this post are closed