Read on Twitter
Today @tylercowen has a nice wrong article on centralized quarantine where he (wrongly) states the only recent quarantine program has been in Cuba and (wrongly) states CQ requires family separation and (wrongly) says the scale of CQ would be too large. https://www.bloomberg.com/opinion/articles/2020-05-13/forced-quarantines-are-not-the-american-way?utm_source=twitter&utm_medium=social&utm_campaign=socialflow-organic&utm_content=view&cmpid%3D=socialflow-twitter-view&sref=htOHjx5Y
It's difficult to track down exact procedures used in previous pandemics. Especially in the case of influenza pandemics, drastic procedures were often not taken due to the availability of not-totally-useless antivirals, and vaccines.
Meanwhile, quarantine is not very useful at all for dealing with pathogens that spread through vectors like mosquitoes, or through the water supply. So quarantine just isn't helpful with e.g. cholera, Dengue, Zika, etc.
For infectious diseases like HIV, quarantine can theoretically be useful, but since HIV is not contagious without VERY close personal contact, and it is relatively straightforward for a person to know at least in theory how to avoid that contact, quarantine is not...
The first-best policy solution. It's not even the second-best.
Quarantine only makes sense in the case of a disease which is PRIMARILY contagious through everyday forms of contact on a person-to-person basis, and for which treatments/vaccines are not available at scale.
Quarantine only makes sense in the case of a disease which is PRIMARILY contagious through everyday forms of contact on a person-to-person basis, and for which treatments/vaccines are not available at scale.
If you have treatments or vaccines, quarantine is not your best solution. If the disease is not spread through personal contact, quarantine is not a solution at all.
How many diseases have we encountered for which quarantine is even a HYPOTHETICALLY PLAUSIBLE solution?
How many diseases have we encountered for which quarantine is even a HYPOTHETICALLY PLAUSIBLE solution?
Well....
SARS and MERS are two. And we already have laws providing for forcibly quarantining people at the border if they have these. COVID is now on that list too. We just don't enforce these laws, because we have puny state capacity.
SARS and MERS are two. And we already have laws providing for forcibly quarantining people at the border if they have these. COVID is now on that list too. We just don't enforce these laws, because we have puny state capacity.
Tuberculosis before widespread BCG and modern treatments is another.... and the dominant strategy to deal with tuberculosis was sanataria! And they worked!
These were MOSTLY voluntary, but that wasn't because society was unwilling to be coercive (consumptive people OFTEN faced MANY forms of coercion in their life), but because the state lacked the fiscal capacity to support large numbers of patients.
We can also look at the case of flu pandemics before vaccines. Flu vaccines became available after the 1943 flu season. But the horrible 1918, 1929, and 1943 epidemics all saw huge deployment of forcible quarantines. MOSTLY for the actually-sick, not close contacts...
But this was, again, mostly due to fiscal capacity, not unwillingness to quarantine protectively.
Even after vaccines were available, quarantine programs continued. The 1976 swine flu near-outbreak, for which a vaccine was not ready, was contained through quarantine.
Even after vaccines were available, quarantine programs continued. The 1976 swine flu near-outbreak, for which a vaccine was not ready, was contained through quarantine.
Granted, it was an outbreak on a military base and so only military personnel were impacted. But it shows that at least as of 1976, when faced with a non-vaccinatable, not-very-treatable novel influenza, quarantine was the DEFAULT RESPONSE.
Keep in mind, domestic military bases ***have families on them***, so this quarantine probably did involve some degree of family separation and civilian impact, so this is not "just" a military personnel situation.
Tracking down details on responses to the 1957 and 1968 pandemics is harder. Most documents just ASSUME that we understand what is meant by health agencies fighting them.
But it does seem to me at least in 1957 that quarantine orders were used. 1968 I am less confident of, but it seems likely.
To this day, novel influenza strains are legally quarantinable at the border.
So ON THE FACTS, @tylercowen 's pejorative limitation of CQ to Cuba is foolish. The reason CQ is not more widespread is because it's not a hammer; it's a precision tool only applicable in certain cases. Those cases don't come up often.
But when those cases HAVE occurred in America, we've used quarantine as a strategy. Not always on the scale as is currently being deployed around the world, but this was very clearly due to lack of state capacity, not compunction.
Now, on to family separation.
Centralized quarantine does not have to require family separation. I agree especially in the case of children we should not separate families if families want to stay together.
Centralized quarantine does not have to require family separation. I agree especially in the case of children we should not separate families if families want to stay together.
Assuming we're using hotels, there are family-size rooms to be used. In the enabling legislation or order, just specify very clearly that children can't be separated.
But @tylercowen 's "150 million Americans could be quarantined" is an absurd misrepresentation.
Over the course of the next year? Sure, maybe 150 million people would each spend a few days in a hotel while they await a negative test result.
Over the course of the next year? Sure, maybe 150 million people would each spend a few days in a hotel while they await a negative test result.
The US has over 5 million hotel rooms that run at something like 75% capacity each night which suggests something like 2 billion person-hotel-nights in a normal year. So we'd increase hotel demand by... 35% or so?
As far as the morality of separating people....
Serious question for @tylercowen : do you know anyone in the hospital with COVID?
You often are not allowed to visit them anyways!
Serious question for @tylercowen : do you know anyone in the hospital with COVID?
You often are not allowed to visit them anyways!
It's not like there's a lovey-dovey alternative here!
(gotta pause for a bit while I do something else)
(gotta pause for a bit while I do something else)
Back for a minute.
Again, treatment is already in isolation, so it's not like CQ is somehow denying families care options they normally would have had. https://twitter.com/hyperplanes/status/1260738009818857474
Again, treatment is already in isolation, so it's not like CQ is somehow denying families care options they normally would have had. https://twitter.com/hyperplanes/status/1260738009818857474
Now then:
IFFFF we needed to quarantine 150 million people, it would be a ***manageable shock*** to the US hotel sector.
But we don't need to quarantine 150 million people.
IFFFF we needed to quarantine 150 million people, it would be a ***manageable shock*** to the US hotel sector.
But we don't need to quarantine 150 million people.
Suppose R0 is 2. And suppose, through mask-wearing, large-assembly limits, more WFH, and other very modest social distancing measures, we can lower R0 to 1.5
How many people need to be quarantined to get us to 0.7?
How many people need to be quarantined to get us to 0.7?
The answer varies based on the distribution of spread-behavior among infectious people. If most contagion is driven by a few super-spreaders and super-spreading is driven by person-level traits, we could get R0 to 0.7 quarantining as little as 5% or so of infectious people.
I don't think that's a plausible model!
But some people do.
Israel has gotten R0 down quite low witha quarantine program that has just 30% compliance rates; probably quarantined less than 15% of infectious people.
But some people do.
Israel has gotten R0 down quite low witha quarantine program that has just 30% compliance rates; probably quarantined less than 15% of infectious people.
But given that SOME degree of heterogeneity in spread behavior associated with person-level traits exists, and given that super-spreaders are by definition easier to ID due to more network nodes connected to them, there is no model in which it's necessary to quarantine 50%...
Of infectious people in order to reduce R0 50%. Depending on your assumptions, R0 will drop 50% with somewhere between 20 and 45% or so of infectious people isolated within 2-4 days of infection.
So how do we get to there?
Well, in an ideal world, we have fast testing and good contact tracing, and something like 50%+ of our quarantined people test positive.
That ideal world is not America in 2020.
Well, in an ideal world, we have fast testing and good contact tracing, and something like 50%+ of our quarantined people test positive.
That ideal world is not America in 2020.
More realistically, we can probably assume that we will have a lot of misses. So say that we have to quarantine 5 people to get one positive. That means, we need to quarantine (Infectious People / 3) * 5 to get R0 to very-far-below-epidemic levels.
BUT WAIT!
Many positive tested people ARE EACH OTHER'S CLOSE CONTACTS!
So we can't just multiply positive tests, because a lot of those folks have duplicative contacts.
Many positive tested people ARE EACH OTHER'S CLOSE CONTACTS!
So we can't just multiply positive tests, because a lot of those folks have duplicative contacts.
So like if you have 5 people testing positive in a household, they don't EACH have 10 separate close contacts. Those close contacts will have a lot of duplication!
I'm unsure exactly how this math pencils out.
But.... it does not add up to 150 million people.
But.... it does not add up to 150 million people.
We've seen CQ programs conducted in many countries now with very good effects and nowhere has needed to quarantine half of its population.
IF we needed to, we could (again, for a few days).
But... we aren't gonna need to.
IF we needed to, we could (again, for a few days).
But... we aren't gonna need to.
Gotta pause again, be back in a bit.
This is a good point.
If we get 30-minute tests, people might only be quarantined for a single night (best to get 2 negative tests a day apart). https://twitter.com/bobbywarnsiii/status/1260748979534348290
If we get 30-minute tests, people might only be quarantined for a single night (best to get 2 negative tests a day apart). https://twitter.com/bobbywarnsiii/status/1260748979534348290
The version of CQ I espouse (families can be together! it's just a hotel stay! it's very short if we have tests!) is so "mild" many people ask me: why bother? Wouldn't home isolation be better?
No.
No.
First of all, MANY families will REQUEST separation, to protect their family members. We should accommodate that request and give people the resources to protect their families.
Second, at-home quarantine is extremely porous. Household members run errands. They receive visitors. What makes CQ effective is that it cuts household-to-household transmission ***to zero***. Even strict lockdowns don't do this.
New York has been in lockdown for weeks and yet 40% of cases are still coming from community transmission, and 60% from within household. Giving options to isolate from family reduces that 60%, and central isolation obliterates that 40%.
But my sense is @tylercowen knows all this. He doesn't dispute the effectiveness. He disputes the morality of coercion.
Gotta pause again and then I'll come back and discuss that.
Gotta pause again and then I'll come back and discuss that.
Okay, before I go on, two comments:
1. EVERYTHING is more effective if done early. But that's not a justification for fatalism today. Yes, earlier action is better. BUT TODAY IS THE DATE FOR EARLY ACTION FOR THE NEXT WAVE! https://twitter.com/tylercowen/status/1260754199073173504
1. EVERYTHING is more effective if done early. But that's not a justification for fatalism today. Yes, earlier action is better. BUT TODAY IS THE DATE FOR EARLY ACTION FOR THE NEXT WAVE! https://twitter.com/tylercowen/status/1260754199073173504
2. The first centralized quarantine sites in Hong Kong were attacked by protestors with firebombs. Opposition to CQ in Hong Kong was intense and swift and resulted in quite intense confrontations. People here DO NOT respect the state. https://twitter.com/tylercowen/status/1260754199073173504
The reason CQ was ultimately accepted was because ***it works***. Many people opposed it at first; nobody trusted Hong Kong's government to not abuse the CQ power.
But urgency overruled the opposition.... and it hasn't been abused (yet, AFAIK).
But urgency overruled the opposition.... and it hasn't been abused (yet, AFAIK).
Will CQ save New York? Tricky to say. Will it allow North Dakota to reopen soon and safely? Absolutely.
OKAY, on to the moral question.
What is @tylercowen 's argument for why CQ is a bad thing?
He.... actually doesn't make one. At all. He doesn't make a single moral argument.
What is @tylercowen 's argument for why CQ is a bad thing?
He.... actually doesn't make one. At all. He doesn't make a single moral argument.
He says quarantining healthy prime-age people in hotel rooms where they have their own facilities and no contact besides opening the door to pick up a bag of food is similar to the hands-on assisted care of nursing homes (wrong)... but that's not a moral argument!
That's simply an argument about efficiency! And a wrong one at that.
He says it would be an indefinite detainment because nobody knows how long contagion lasts.
True, we don't know. But we do know people who test negative (ideally twice) are not infectious for a disease they don't have. Test negative and you go home!
True, we don't know. But we do know people who test negative (ideally twice) are not infectious for a disease they don't have. Test negative and you go home!
Because quarantining people is expensive, states are likely to place a high priority on getting quarantined people tested ASAP so they can be sent home and taken off the budget rolls.
When you think politics in AMerica are more turbulent and extreme than in Hong Kong lol https://twitter.com/Gws130/status/1260759450257895425
AGain, @tylercowen is worried about false positives: but.... why? If a person tests positive, they'd be quarantined in a decent room with a daily compensation for their trouble. And then they'd be tested again. After two negative tests they'd go home. Tops 7 days.
Again, if we get faster tests, they could be out in 36 hours.
There's no way states are going to just maliciously hold on to random people for funzies when it's running them hundreds of dollars a day and that person is an enfranchised person with full legal rights and internet.
There's no way states are going to just maliciously hold on to random people for funzies when it's running them hundreds of dollars a day and that person is an enfranchised person with full legal rights and internet.
So the scale of harm here seems to be very small. I suppose if you're an extremely strict deontologist you might oppose it on that basis, but that person would also oppose public assembly bans, mask requirements, or *any* epidemic response, tbh
"Indefinite" lol.
People acting like "until you get two positive tests" is equivalent to Gitmo. https://twitter.com/eznark/status/1260761230236626946
People acting like "until you get two positive tests" is equivalent to Gitmo. https://twitter.com/eznark/status/1260761230236626946
It's a very definite rule: until you're not a walking murder-germ-bag. Which in the huge majority of non-hospitalized cases takes under 30 days, and for cases that aren't positive will be 7 days or less.
Then @tylercowen also says CQ is immoral because 40-60% of people might get infected.
This is the most absurd bit of the whole piece.
Yeah, 60% of people will get infected.... IF WE DON'T DO CENTRALIZED QUARANTINE.
This is the most absurd bit of the whole piece.
Yeah, 60% of people will get infected.... IF WE DON'T DO CENTRALIZED QUARANTINE.
"We can't take measures to reduce the spread of the disease because the disease will infect too many people."
The appropriate idiom for my feelings on this is, "I can't even"
The appropriate idiom for my feelings on this is, "I can't even"
Alternatively, "facepalm"
This defeatism is precisely what's been so wrong with Western policymaking since Day 1 of this, and by Day 1 I want to be very clear I mean December 31 when the first English-language news of a new SARS broke and non-idiots who pay attention to Chinese news were like "NOOOOOOOOO"
Then, his last argument is that CQ is bad because it will make Americans want to own guns.
....
I'm 100% on board with Americans wanting to own guns? What's.....? What's the argument here?
....
I'm 100% on board with Americans wanting to own guns? What's.....? What's the argument here?
"THIS POLICY IS BAD BECAUSE IT WILL MAKE PEOPLE APPRECIATE THEIR 2ND AMENDMENT RIGHTS!"
wut
wut
Note the argument being made is not that there would be ACTUAL VIOLENCE. I AM worried about that! But that's not @tylercowen 's argument!
His argument is it'll make people.... more polarized.
His argument is it'll make people.... more polarized.
We can't save the next 80,000 lives because doing so might make people want to own guns.
Then @tylercowen asks "what if scientists are about to have a breakthrough!"
Great! Once there's a breakthrough, we can stop quarantining people!
But we can quarantine people will we wait for a breakthrough.
Great! Once there's a breakthrough, we can stop quarantining people!
But we can quarantine people will we wait for a breakthrough.
So.... where is the moral argument?
Ultimately, @tylercowen makes vague grabs at "individual rights."
What right? The right to impose well-demonstrated lethal externalities on anyone you want? The right to inflict possibly permanent lung damage on your neighbor?
Ultimately, @tylercowen makes vague grabs at "individual rights."
What right? The right to impose well-demonstrated lethal externalities on anyone you want? The right to inflict possibly permanent lung damage on your neighbor?
Yes, we are infringing on a lot of liberties (for somewhere between 2 and 30 days).... but we are not doing so for kicks and giggles! We are doing so because failure to do so infringes on all of our first right: the right to *life*.
The state's first function is to guarantee the basic security of society from mortal threats to its existence. Foreign invaders, domestic murderers, etc. I am just as dead if I'm killed by COVID as by a gun.
It is not correct to call something "libertarianism" if it does not even support state coercion to prevent actual lethal threats. This is extremely basic "night watchman state" stuff. The critique being leveled is bashfully stated because the critique is anarchism.
Once you argue that the state lacks the legitimate moral right to quarantine people with a lethal disease (provided the quarantine is humane!), you've argued the state lacks the legitimate moral right to do anything.
Unless you place MASSIVE weight in society's social welfare function on *intentional* harms vs. unintentional. Which I wouldn't weight that at ZERO but it'd need to be a stupendously large weight.
Look, I like @tylercowen a lot. He's brilliant and has been formative to my thinking on numerous occasions.
But if this is the best the anti-quarantine crowd has got then the road to "CQ opposition is anti-vaxxing by another name" is gonna be very short indeed.
But if this is the best the anti-quarantine crowd has got then the road to "CQ opposition is anti-vaxxing by another name" is gonna be very short indeed.
