The Corona Virus

[veti]

Production is being affected. In Maryland, food processing plans are being shut down, because the workers were infected, and infected people packing food is a terrible idea. Lack of packing places means that they simply can't process as much meat. So Tysons, unable to sell or process their meat, is simply killing and burying chickens instead. They don't have the facilities to house arbitrary numbers of chickens. Simply keeping them around isn't something they have the ability to do. So, in a very real sense, the production is simply lost.

I've seen stories like that from all over. In Belgium, the government is appealing to people to eat more chips, because no one has space to store all those potatoes. Elsewhere it's lettuce and tomatoes, because the restaurants are closed. The meat industry is even worse off for the same reason - it's not so simple to switch from selling huge slabs of meat for restaurants, to the handy little packs you pick up at the supermarket.

I'm in an "essential" business, namely selling electricity. Now, if I were drawing up a list of essential business, I would obviously include generating electricity, transmitting and distributing the stuff. But selling it? - which is to say, sending out bills and settling with generators and network operators? What exactly is essential about that?

Well, if people don't pay for their electricity, then all those other people who generate and distribute it would soon run out of money. And then they'd stop doing it, because they've got their own bills to pay.

But if that's enough to make something essential, then why isn't every business "essential"? In an economy, everything is connected to everything else. They all keep the money in circulation, without which - people won't be able to pay their bills, and eventually even the essential ones won't be able to keep going.

I have a strong feeling there's a problem with this line of reasoning, because it seems to put email scammers and estate agents and pimps on the same level as people who do decent work. But it bothers me all the same, because I can't see the answer.

[mjasghar]

With reference to the exponential growth vs linear growth
There’s a natural limit to exponential growth with regards to population density. In most cases around the world we are seeing big urban areas getting hit and the lockdown keeping it on those areas. Then we see people breaking lockdown and other areas start the epidemic cycle.
That linear growth you are talking about wasn’t some good sign as it was an almost vertical line - if it had continued growing exponentially you would have had a gradient of infinity.
Consider these stats on infection rates: the USA is currently at Switzerland and Italy levels and that’s with its main infection site (nyc) stabilising the growth of cases. Which suggests that with the surge in places that were are the start of the curve but want to be part of a national easing for whatever reason (not allowed to discuss here) the USA will end up with a higher rate. Where will all those who cried out about it being under control go then? I suspect they will move the goal posts and talk about big cities in those states and conveniently forget their previous arguments about population density
https://www.statista.com/chart/amp/21176/covid-19-infection-density-in-countries-most-total-cases/

[Rockphed]

Could you link the data you are using?

The top data set on this page. Helpfully they include the numbers below the graph, so you can plot them on a log vertical axis and see the exponential growth more easily. On a log plot it is also easier to see the corner in the middle of March.

Now they might have stopped making it because they figured out that there were significant problems with the data (less people caught the disease on tuesday and wednesday every week). They might also have stopped making it because they realized that even as a historic record it wasn't great (since every number increased during the 2 or so weeks that I followed that chart).

[Liquor Box]

This is my point though. You were saying that a 4 week total isolation would potentially destroy the economy. I'm saying that there are places that have effectively already paid a higher economic price than that, and it hasn't destroyed their economies.

Paid a higher economic price than what? Sorry, I don't understand what you are saying here. Are you saying that there are countries that have had a four week lockdown, and their economies have come out relatively unscathed? If s, I think examples may be required.

In Japan (where I am), there hasn't been much in the way of official interventions (almost everything has been of an 'advisory' nature), but from March there was something like a 20-30% drop in public activity because of people deciding on their own, and many companies started to roll out work-from-home policies from around that time (mine has been on 'WFH if you can from March, WFH mandatory from April' for example). That 20-30% drop in public activity is now approaching a target drop of 80% (measured via public transport activity).

San Francisco also had stay-at-home orders starting from March, and South Korea, Hong Kong, etc have been implementing their own interventions from then or even earlier. The Italy lockdown was also at the start of March. So, worldwide there have been aggressive interventions roughly for the last 2 months.

Ah, so what you are saying is that an immediate lockdown would have been better than the escalating course countries have taken.

I disagree for three reasons.

First, I don't think would have been realistic to shut down immediately, because you have to establish that the illness is serious enough to justify such a shut down first. Coronavirus was declared a pandemic until mid March, and even then it was not the first pandemic of the century (and it has not yet had a higher death toll than the 2010 N1H1 pandemic). Looking at matters objectively back in February it was not clear that Covid was worse than other outbreaks in recent history, so if we were going to shut down at that point, it would mean we'd be shutting down every ten years or so to avoid illnesses that might turn out to develop into something more widespread. Sure, we can look back now and say that Covd 19 was worse than most of the others, but we didn't have the benefit of hindsight in February. So, in my view shutting down before mid to late march (except in the worst hit countries) was not a realistic response.

Second your assumption (based on some theory) is that a lockdown of 4 to 6 weeks would be sufficient to eradicate the virus. My home country did move quickly after it was declared a pandemic to enter lockdown. We were largely restricted to our homes except for certain activities from 23 March (six and a half weeks ago) - at that point we had only 100 infected (with the first detection having been only two weeks prior) and no deaths. Despite that quick reaction, and very low starting point, we are still restricted to our homes with no indication of when that restriction might end. So the evidence from my own country (assuming the officials know what they are doing) 6 weeks is not sufficient to ensure the pandemic is finished locally.

Thirdly, and I think this is our main source of disagreement, a lockdown does much more harm than a lighter touch response. Even if the lighter touch response is far more drawn out (which is not guaranteed). There are a number of reasons for this. First, a lockdown effects the entire population (unless exceptions are made), but far less than the entire population will be prevented from working by Covid (some people will never get it, others are apparently asymptomatic). Second negative effects from a lighter touch are more spread out over time, which is less of a shock to the economy. Third, to the extent there is available workforce, temporary hires can cover the absence of covid victims due to sickness.
There are some good examples of this comparing similar countries (sweden vs Denmark or NZ vs Australia), but I hesitate to use them because I am sure there are counter-examples out there, and relying on isolate examples can be misleading.

That's beside the additional problem of being out of step with other countries in terms of response. You point out that this could be avoided by cooperating. I agree it could theoretically, but I don't believe the sort of cooperation you envisage is practically possible.

10% of GDP is less than the tax level in pretty much any country whose tax structure I'm familiar with, so this is a solvable problem via re-allocation of existing resources. It's not a matter of 'could it be done?', it's a matter of 'would we choose to do it?'. And actually, this is about the right order of magnitude for the stimulus package the US passed recently - that was $2T, which is about 10% of US GDP.

As per my previous post, the harm is not only 10% of GDP. Our treasury forecast that a 6 week lock down (which we've exceeded already) would mean a 26% reduction in GDP this year. This is because there are flow on effects).

It's not like taking 5 weeks (10%) unpaid leave from work and then returning where you left off. It's more like resigning suddenly, taking five weeks, and then looking to restart your career again while mitigating the damage that your sudden resignation caused.

Take the GFC as an example. That is estimated to have cost USA $4.6 trillion (or $70,000 per american), and that was without any shut down. Economic issues that were far far smaller than a 2 or 4 or 6 or 8 week lockdown snowballed, and economic stimulus packages only slowed the decline rather than arrested it. Here you are proposing a massive snowball to start rolling down the hill, and that is likely to lead to economic harm of a scale greater than in living memory.

[sihnfahl]

The top data set on this page. Helpfully they include the numbers below the graph, so you can plot them on a log vertical axis and see the exponential growth more easily. On a log plot it is also easier to see the corner in the middle of March.

With an up-to-date graph here.

Expand 'About The Data'.

[Liquor Box]

Except that at least a significant chunk of the production of goods is still happening. Food is still being produced at similar numbers. Houses are still being built, and housing is available. The economy is why farmers are plowing under their fields or selling beef at a loss while grocery stores and food banks are empty. Why people are sleeping in parking lots in Vegas while every hotel is vacant. We have the things we desire or need because of people doing the work, not "the economy", and anything that prioritizes the economy over the people doing the work during this crisis is missing the forest for the trees.

I'm not sure what you are suggesting, but if its lockdown that definitely does mean things not being produced. A lockdown that prevent people from going to work stops them from building houses.

I suggest that all the things you refer to (grocery stores and hotels being empty) is because the economy is unhealthy because of Covid-19 (either restrictions or fear of infection). It's damage to the economy, rather than the economy itself that is causing those problems.

Prioritising the economy is prioritising people - it keeps people in jobs, in homes and fed. Whether that's a higher priority than keeping them in health depends on the degree of damage caused to their financial and physical wellbeing by the range of available response options, and that is well up for debate (in fact that largely is what is being debated in this thread).

[Rockphed]

With an up-to-date graph here.

Expand 'About The Data'.

I don't see an equivalent dataset on that page. I see "cases new reported by day", but not "when cases were contracted by day".

[sihnfahl]

I don't see an equivalent dataset on that page. I see "cases new reported by day", but not "when cases were contracted by day".

The 'New Cases by Day' replaces 'Date by Illness Onset'.

They can't pin down the illness onset due to reporting issues; sometimes the best they could do is estimate it as being 3 days prior, which they believed too unreliable due to how COVID-19 progresses. So they are using the date of reporting or test verification.

[mjasghar]

Prioritising the economy is prioritising people - it keeps people in jobs, in homes and fed. Whether that's a higher priority than keeping them in health depends on the degree of damage caused to their financial and physical wellbeing by the range of available response options, and that is well up for debate (in fact that largely is what is being debated in this thread).

So what’s your valuation on an individual life vs % gdp loss? And do remember you don’t get any weighting for yourself or your family and friends
And, furthermore, if as seems clear that relaxing restrictions causes an upsurge in cases as seen in Texas as I showed above, what will you do if people decide not to turn up to work for fear of infecting themselves and then their loved ones? How will you avoid the gdp loss then? Perhaps have the army round them up and force them to work at gunpoint with whips etc?

[Tyndmyr]

I've seen stories like that from all over. In Belgium, the government is appealing to people to eat more chips, because no one has space to store all those potatoes. Elsewhere it's lettuce and tomatoes, because the restaurants are closed. The meat industry is even worse off for the same reason - it's not so simple to switch from selling huge slabs of meat for restaurants, to the handy little packs you pick up at the supermarket.

I'm in an "essential" business, namely selling electricity. Now, if I were drawing up a list of essential business, I would obviously include generating electricity, transmitting and distributing the stuff. But selling it? - which is to say, sending out bills and settling with generators and network operators? What exactly is essential about that?

Well, if people don't pay for their electricity, then all those other people who generate and distribute it would soon run out of money. And then they'd stop doing it, because they've got their own bills to pay.

But if that's enough to make something essential, then why isn't every business "essential"? In an economy, everything is connected to everything else. They all keep the money in circulation, without which - people won't be able to pay their bills, and eventually even the essential ones won't be able to keep going.

I have a strong feeling there's a problem with this line of reasoning, because it seems to put email scammers and estate agents and pimps on the same level as people who do decent work. But it bothers me all the same, because I can't see the answer.

That does seem to be a fundamental issue here. Economics cannot be entirely ignored, even in a pinch like this, but it does remove any sort of coherent lines. If one person in sales is essential, why not another? The most coherent line of reasoning seems to be targeting entertainment as un-essential, as well as some service industries. But overall, the shutdown only crimped things so much. Perhaps ten percent, as mentioned earlier, which is consistent with google's reports of reduced vehicle travel, etc.

Which puts it in the awkward place of being pretty awful for the industries targeted(I would not wish to own a movie theater right now), but overall, maybe not reducing exposure all that much. If perhaps 90% of people are still going to work as normal(which is also roughly consistent with jobless claims thus far), the effect on transmission does not seem as if it would be immense.

A 10% reduction in exposure is certainly something, but it's not amazing. And people sent home, while they may have reduced exposure, probably do not have zero exposure, as they still need to go buy groceries and such. I'm not sure what the total exposure reduction is from the shutdown, but it seems as if it'd have to be a single digit reduction.

The above is, of course, considering purely the shutdown as a reaction. Other reactions may be more effective. Widespread mask wearing probably has a significantly larger effect, even with the caveats involved.

You could probably get a larger effect by shutting down more industries, but that would require revisiting what is actually essential, and would probably have significant tradeoffs. If you shut down every billing department for the 6+ weeks of shutdown we've had so far, a nasty snarl might develop. Companies would probably run out of money and be unable to continue, and once re-opened, it would take quite some time to sort out the mess.

So what’s your valuation on an individual life vs % gdp loss? And do remember you don’t get any weighting for yourself or your family and friends
And, furthermore, if as seems clear that relaxing restrictions causes an upsurge in cases as seen in Texas as I showed above, what will you do if people decide not to turn up to work for fear of infecting themselves and then their loved ones? How will you avoid the gdp loss then? Perhaps have the army round them up and force them to work at gunpoint with whips etc?

I don't think anybody is suggesting the latter. People are going to make decisions to limit their personal risk, and that's a good thing. If your job seems to be unsafe, and you choose to go elsewhere, that's a fine moral decision. I don't think any one person has the answer for all circumstances, and you probably know your own life best.

Rather than looking at it as buying and selling lives, it might be more helpful to look at both the economic impacts and the direct loss of lives as negative effects of the disease. Neither is entirely avoidable, but we want to mitigate as much as possible.

If a valuation is absolutely necessary, I would suggest looking at whatever economic loss causes an equivalent amount of human suffering/loss of life. That seems, at minimum, to be similar. Taken together with other preventative losses of life(increased rates of family abuse, reduced ability to respond to depression/suicidality, etc), you generally want to take the action that results in the least harm.

[Friv]

Which puts it in the awkward place of being pretty awful for the industries targeted(I would not wish to own a movie theater right now), but overall, maybe not reducing exposure all that much. If perhaps 90% of people are still going to work as normal(which is also roughly consistent with jobless claims thus far), the effect on transmission does not seem as if it would be immense.

A 10% reduction in exposure is certainly something, but it's not amazing. And people sent home, while they may have reduced exposure, probably do not have zero exposure, as they still need to go buy groceries and such. I'm not sure what the total exposure reduction is from the shutdown, but it seems as if it'd have to be a single digit reduction.

Keep in mind that there is a world of difference between "90% of people going to work" and "90% of people going to work as normal."

Where I am, the number of people allowed in grocery stores has been dramatically reduced; people are shopping far less, and doing larger shops. So that reduces exposure. Office workers are working from home as much as possible, and mostly doing online work. That reduces exposure. Restaurants are doing takeout only, and minimizing contact with customers, which reduces exposure both for workers and for customers.

There are a lot of ways to reduce exposure without entirely closing economies, and we've done a lot of them.

[Tyndmyr]

Keep in mind that there is a world of difference between "90% of people going to work" and "90% of people going to work as normal."

Where I am, the number of people allowed in grocery stores has been dramatically reduced; people are shopping far less, and doing larger shops. So that reduces exposure. Office workers are working from home as much as possible, and mostly doing online work. That reduces exposure. Restaurants are doing takeout only, and minimizing contact with customers, which reduces exposure both for workers and for customers.

There are a lot of ways to reduce exposure without entirely closing economies, and we've done a lot of them.

Correct, those would be the other measures I mentioned. Most of those have a relatively low impact on economies relative to laying people off or closing entirely.

[monomer]

I don't think anybody is suggesting the latter. People are going to make decisions to limit their personal risk, and that's a good thing. If your job seems to be unsafe, and you choose to go elsewhere, that's a fine moral decision. I don't think any one person has the answer for all circumstances, and you probably know your own life best.

The problem is that the latter is almost exactly what is happening. In the States where they have lifted the restrictions, if your employer asks you to return to work and you refuse, you not only lose your job, but you lose your access to unemployment insurance, and possibly more critically you lose your health insurance. You can't just choose to go elsewhere either, as good luck getting a new job at this time.

So now you have to make a decision to either risk your health and that of your family, or lose all sources of income and access to healthcare for the foreseeable future. If you are a generally healthy young adult living alone, the choice is a no-brainer, but if you or someone you are living with are older or in poor health, then you are left with two really bad options.

[NichG]

But if that's enough to make something essential, then why isn't every business "essential"? In an economy, everything is connected to everything else. They all keep the money in circulation, without which - people won't be able to pay their bills, and eventually even the essential ones won't be able to keep going.

I have a strong feeling there's a problem with this line of reasoning, because it seems to put email scammers and estate agents and pimps on the same level as people who do decent work. But it bothers me all the same, because I can't see the answer.

I think one problem with this line of reasoning is assuming that the things you're comparing are 'everything as normal' and 'everything the same, but one piece turned off'. Whereas the comparison should be 'everything as normal' vs 'turn one piece off, then rearrange everything else to pick up the slack if possible' and then seeing if such a re-arrangement is possible.

Lets take the email scammer. They need to eat, and they need money to eat, so under the first criterion it appears that email scams are essential. But now lets say that it was really valuable to get that scammer to stop scamming, could we in principle cause that to happen without killing them? Sure, we just give them some other job they can do which pays the same; or we just say 'fine, we'll just give you money, you aren't expected to do anything in return'. The economy has flex in it, in the sense that you've got a bunch of different resources but money enables them to be fungible. So if you don't produce oil, you can buy it from someone who does instead, and so on. And to some extent, they might be able to fill this need themselves without someone coming in and doing it for them - if they can code, they could do contract work; or they could do Mechanical Turk packets; or they could ask their family for money; or any number of other things. So that job is filling an essential role for them, but the job itself isn't essential.

Essential jobs are hard stops in that fungibility - things which you can't just buy off by moving numbers around on a spreadsheet or making agreements with each-other to act a certain way rather than a different way. If we take your example of billing administration, it would be possible (e.g. if a government wanted to do so) to say 'okay, rather than sending individual contracts and bills to all of your customers, appoint one guy who talks with us, and we'll just pay for everyone this month and raise taxes next year'. Or for that matter, a rich individual could do the same, or a corporation, or a bank - at least to some extent. So those jobs could be reduced down to almost nothing (or in the presence of strong trust, even reduced all the way down to nothing) I think. But if the cost for strategies to implement that reduction (e.g. the government picking up the tab for everyone's electric bills for a month) is too high on the balance, then it may be effectively 'essential' anyhow since no one would be willing to do the thing that could make the job go aware without breaking things.

But if no one was allowed to produce food then no matter how much money one might throw at the problem or how much we'd try to re-arrange the system to pick up that slack, we'd still starve.

Paid a higher economic price than what? Sorry, I don't understand what you are saying here. Are you saying that there are countries that have had a four week lockdown, and their economies have come out relatively unscathed? If s, I think examples may be required.

I'm saying that places have already paid for 6 week and 8 week lockdowns. Whether their economics are unscathed, scathed, destroyed, etc, it is what it is. So a response to a 4 week lockdown plan that projects a greater impact than those existing 8 week plans can't be 'but a 4 week lockdown would destroy the economy!' since either the economy is already being destroyed, or in fact no, a 4 week lockdown doesn't actually destroy an economy.

It's a point of logic. If you spend $8 to get 1kg of flour and I say 'but, you could have gotten 2kg of the same flour for $4 here' it doesn't make sense to say 'but $4 is too expensive to pay for flour!' as an objection to shopping at the other place. You can say 'I don't think we should buy flour at all' or 'well I already bought flour and so I don't need more right now' or even 'I know a place where you can get 2kg for $1!', but saying '$4 is expensive!' is a non-sequitur when the argument is that in fact the second place is cheaper.

Ah, so what you are saying is that an immediate lockdown would have been better than the escalating course countries have taken.

I disagree for three reasons.

First, I don't think would have been realistic to shut down immediately, because you have to establish that the illness is serious enough to justify such a shut down first. Coronavirus was declared a pandemic until mid March, and even then it was not the first pandemic of the century (and it has not yet had a higher death toll than the 2010 N1H1 pandemic). Looking at matters objectively back in February it was not clear that Covid was worse than other outbreaks in recent history, so if we were going to shut down at that point, it would mean we'd be shutting down every ten years or so to avoid illnesses that might turn out to develop into something more widespread. Sure, we can look back now and say that Covd 19 was worse than most of the others, but we didn't have the benefit of hindsight in February. So, in my view shutting down before mid to late march (except in the worst hit countries) was not a realistic response.

Second your assumption (based on some theory) is that a lockdown of 4 to 6 weeks would be sufficient to eradicate the virus. My home country did move quickly after it was declared a pandemic to enter lockdown. We were largely restricted to our homes except for certain activities from 23 March (six and a half weeks ago) - at that point we had only 100 infected (with the first detection having been only two weeks prior) and no deaths. Despite that quick reaction, and very low starting point, we are still restricted to our homes with no indication of when that restriction might end. So the evidence from my own country (assuming the officials know what they are doing) 6 weeks is not sufficient to ensure the pandemic is finished locally.

Thirdly, and I think this is our main source of disagreement, a lockdown does much more harm than a lighter touch response. Even if the lighter touch response is far more drawn out (which is not guaranteed). There are a number of reasons for this. First, a lockdown effects the entire population (unless exceptions are made), but far less than the entire population will be prevented from working by Covid (some people will never get it, others are apparently asymptomatic). Second negative effects from a lighter touch are more spread out over time, which is less of a shock to the economy. Third, to the extent there is available workforce, temporary hires can cover the absence of covid victims due to sickness.
There are some good examples of this comparing similar countries (sweden vs Denmark or NZ vs Australia), but I hesitate to use them because I am sure there are counter-examples out there, and relying on isolate examples can be misleading.

That's beside the additional problem of being out of step with other countries in terms of response. You point out that this could be avoided by cooperating. I agree it could theoretically, but I don't believe the sort of cooperation you envisage is practically possible.

Starting with point 2, since I think points 1, 3, and 3.5 have a shared response.

Lock-downs have had a huge variation in how they've been implemented, and some lock-down strategies seem to be much more or less effective than others. In Italy for example, you could see the moment the lock-down hit the Covid curve (at the usual 2 week delay) - it was a sharp detune from exponential. However, the peak of active cases and overall shape of the decay seems to be quite different than, say, Iran, where peak happened shortly after the de-tune from exponential and active cases dropped by over 50% in a month. China's decay curve was very striking, with a reduction by a factor of 5 in active cases a month after peak. And the US lockdowns seem to have at best slowed the growth to a steep linear rather than exponential. So details of implementation matter. I'm not aware of any place actively taking half of its essential workers and telling them to stay home for a month, work a month in shifts like that. Many places with lock-downs still permit people to go out to buy groceries under some conditions, and may use different strategies as to how to reduce contacts in those cases (the Czech Republic for example has specific shopping hours for different age groups, and no more than one person per household can go out at once; and cases in isolation in South Korea have groceries dropped off).

That's in part what gave me this thought that the degree of synchronicity of measures might be one such detail that could have a strong effect on the effectiveness of lockdowns or other Covid responses. The theoretical limit is pretty hard to argue with I think - if you literally make sure that no one comes into contact with anyone else, it should be impossible for Covid to spread. The only mechanistic counter-argument I can think of to that would be if there's some fat tail in the distribution of how long people remain infectious, such that people who have recovered from Covid can still be carriers months or years later. I don't think there's any data supporting that though - everything I've seen suggests that peak infectiousness is early.

To the 1st, 3rd, and 3.5th points:

So with all of this, the point is to do better next time because we can't change the past. We can't change the current state of things but we can make observations about what we could have done, that we either decided not to do or were unable to do because of the situation. That way, we can honestly evaluate how much things are worth to us and decide on the trade-offs. I agree that cooperation of this scale wasn't realistic in the world of January 2020. But, if we conclude that cooperation would have decreased the economic impacts of Covid-19 by 90% so that this would just be a bump (even if it means occasionally jumping at shadows that we could have locally ignored), then we can ask 'would it make sense to build some kind of infrastructure now such that cooperation at that scale is in fact feasible the next time there's a global emergency of this form?'.

Of if that's too much in a murky uncertain future to talk about, pandemics tend to have a multi-bump structure - things appear to die down, people open up or relax, and then there's still a naive population that can be rapidly infected by the bug. So even with Covid, there's a high chance we're going to be repeating this exercise. In a few months (at least in places that are managing to get active cases down to the hundreds level) there will be some time to think about what we're going to do then, stockpile things in advance of that second bounce, etc.

And in terms of cost, imagine that we have 2 or 3 such bumps before there's a broadly available vaccine (e.g. 16 months down the road). Or worse, that we don't actually develop long-lasting immunity to Covid and so we see one of these bumps every year once a year forever on. Paying a one-time steep cost to eradicate Covid has a pay-off of whatever those risks would cost us (modulo different subjective valuations of risks versus immediate returns, where people are on the utility curve, etc). And then, to move towards the cooperation strategy, what would have to follow is a frank discussion of those relative valuations, combined with negotiations to subsidize actions that are locally against the interest of that locality but have more global valuation tied up in them than the local excess cost - and deal with all of the usual stuff that complicates such things (negotiation in bad faith, using the negotiations and the ability to force them to fail to leverage payments in excess of the break-even values, etc).

That part of things wouldn't be easy. But, as a researcher, it's very useful to know if you should be working on the theory of how infections spread, or if you should be working on the theory of how to converge to an efficient equilibrium in that sort of subsidy negotiation where every participant can in principle hold the entire market hostage. Going back to the point that started this line of discussion, if the major difficulties of responding to something like Covid are human factors rather than immunological factors, it's important to realize that as early as possible and start planning in those directions rather than spending a lot of time on SEIRS models and structured quarantine strategies.

As per my previous post, the harm is not only 10% of GDP. Our treasury forecast that a 6 week lock down (which we've exceeded already) would mean a 26% reduction in GDP this year. This is because there are flow on effects).

It's not like taking 5 weeks (10%) unpaid leave from work and then returning where you left off. It's more like resigning suddenly, taking five weeks, and then looking to restart your career again while mitigating the damage that your sudden resignation caused.

Well as you point out, you've already done a lockdown at these prices. So even if we quibble over the details of how much exactly the cost is, it's not salient to the argument since the argument is that spending what (or less than what) you were already willing to spend but in a different way (coordinated) would produce significantly higher return per cost.

[Liquor Box]

So whatÂ’s your valuation on an individual life vs % gdp loss? And do remember you donÂ’t get any weighting for yourself or your family and friends
And, furthermore, if as seems clear that relaxing restrictions causes an upsurge in cases as seen in Texas as I showed above, what will you do if people decide not to turn up to work for fear of infecting themselves and then their loved ones? How will you avoid the gdp loss then? Perhaps have the army round them up and force them to work at gunpoint with whips etc?

I don't know what you main by not getting any weighting for myself and famly and friends. One of the lens I do look at it through is what I would rather for myself. I think that is sensible lens for each of us to look at it through.

So which of these risks would I rather run (or would you rather run):

• a 30% chance of contracting the virus meaning a 0.2% chance of dying as well as a 5% chance of losing my job (and therefore my home) due to economic impact (if there's a light touch reaction); or
• a minimal chance of contracting the virus and dying as well as a 30% chance of losing my job (and therefore my home) due to the economic impact (if there's a heavy handed reaction).

Obviously the above figures, particularly the economic ones, are largely guesswork, and relies on all sorts of assumptions. But I do think looking at it through that sort of lens (each of us from our own perspective) is valid.

As to your other questions:

• I don't have a figure for a valuing a small chance of losing my life vs a high chance of losing my job - I'd make a gut call based on what the relative chances of those things happening are.
• I also don't have a figure about whether I'd rather other people to face a big risk to their jobs or a small risk ot their lives.
• The laws regarding people choosing to not come to work because they perceive a health risk will differ from place to place. I doubt they'd extend to the army rounding people up, but they may risk being dismissed and replaced. I suggest discussing whether this is right or wrong, or otherwise taking this aspect any further might approach politics so we should not do so.

NichG, I see you've replied to my earlier post. I'm sorry I can't respond right now because I'm sure it will require a close read, but I will do so.

[veti]

First, I don't think would have been realistic to shut down immediately, because you have to establish that the illness is serious enough to justify such a shut down first. Coronavirus was declared a pandemic until mid March, and even then it was not the first pandemic of the century (and it has not yet had a higher death toll than the 2010 N1H1 pandemic).

The WHO sprung the p-word on 11 March, just weeks after announcing they would no longer try to apply a technical definition to the word. They had been calling it an 'international health emergency' for some weeks by that time. On 7 March, they "reminded all countries and communities that the spread of this virus can be significantly slowed or even reversed through the implementation of robust containment and control activities."

H1N1 only officially (confirmed by lab tests) killed about 19,000 people, but credible estimates put the true figure at about 284,000. The number of confirmed COVID-19 deaths is much higher, and it is likely that the reported figures are (a) being suppressed in some places for political reasons, (b) under-reported in other places because if you die alone at home, it can be weeks before anyone notices, and (c) lagging behind reality in many places because paperwork. So with the official COVID-19 toll standing at 265,000, and deaths accruing at several thousand per day, it's a bold claim that the actual toll is "still" lower than H1N1.

By mid-February COVID-19 was known to be spreading outside China, and it had already killed over a thousand people. The Chinese were officially estimating the mortality rate at between 2 and 3% - which is at least one order of magnitude higher than H1N1, and more than enough to ring alarm bells.

I agree that shutting down the world back in February would have been premature. Shutting down affected countries, however, would not. The tipping point when I started to take the whole thing seriously was late February, when cases started to appear in Australia - I knew that once it got that far, there was no chance it wouldn't also reach NZ. So "looking at matters objectively", I think it was clear by the beginning of March that this was worse than any other pandemic of recent times.

Second your assumption (based on some theory) is that a lockdown of 4 to 6 weeks would be sufficient to eradicate the virus. My home country did move quickly after it was declared a pandemic to enter lockdown. We were largely restricted to our homes except for certain activities from 23 March (six and a half weeks ago) - at that point we had only 100 infected (with the first detection having been only two weeks prior) and no deaths. Despite that quick reaction, and very low starting point, we are still restricted to our homes with no indication of when that restriction might end.

Incorrect - we're not 'restricted to our homes'. Under level 3, we're allowed to come and go as we like, including making longer journeys if we think they're necessary. And there are "indications" that the restrictions will lift to "level 2" next week.

Thirdly, and I think this is our main source of disagreement, a lockdown does much more harm than a lighter touch response. Even if the lighter touch response is far more drawn out (which is not guaranteed).

Well, there I would agree with you. With the benefit of hindsight, it would have been better to close the border at the beginning of March and avoid the lockdown phase entirely - even though it would mean keeping the border closed for well over a year. But by mid-March, it was too late for that. The longer we left it, the more heavy-handed response was needed to contain it.

Our treasury forecast that a 6 week lock down (which we've exceeded already) would mean a 26% reduction in GDP this year. This is because there are flow on effects).

A "lighter touch" response would also have economic effects. Closing the borders destroys tourism, which is almost 6% of the economy by itself, with all the knock-on effects from that. (Of course that's happened anyway, so that's presumably included in the treasury's estimate.) Then there's the effect from lockdowns and recessions in our trading partners, which could easily amount to another 10 to 20%, depending what assumptions you want to make about them. The saying "when America sneezes, the world catches a cold" has never seemed more literal.

Between "deciding when to press the panic button" and "which button to press", there's a whole panoply of judgment calls that ultimately can only be made by the politicians who have been given that responsibility. And one thing we can rely on with absolute certainty is that whatever they do, their opponents will excoriate them for it. That's what political opponents do. But the real question is: do we believe the other lot, faced with the same choices, would have done better?

Not "could". Obviously they "could" have done better. But would they, really? On balance, which lot would you rather trust? Because you've got to trust someone. There is no opting out.

That's not a question that can be debated here, because politics. But I think it's a question worth thinking about.

Amended to add: I had taken your "26%" figure on trust, but on checking - I can't find any trace of support for it. Where did you get it from? The treasury scenarios published here describe an impact of barely half of that, from an eight-week lockdown (four weeks at level 4 followed by 4 weeks at level 3, then ten months at levels 1-2).

The most coherent line of reasoning seems to be targeting entertainment as un-essential, as well as some service industries. But overall, the shutdown only crimped things so much. Perhaps ten percent, as mentioned earlier, which is consistent with google's reports of reduced vehicle travel, etc.

Which puts it in the awkward place of being pretty awful for the industries targeted(I would not wish to own a movie theater right now), but overall, maybe not reducing exposure all that much. If perhaps 90% of people are still going to work as normal(which is also roughly consistent with jobless claims thus far), the effect on transmission does not seem as if it would be immense.

But it's not linear. If you look at some traffic models, you'll quickly notice that a 10% increase in traffic results in a much greater than 10% increase in congestion. If we all abandon the 10% of our daily activities that involve being within 2 metres of someone not in our "bubble", then the hit to the economy would be about 10%, but the effect on virus transmission would be significantly greater. Some people need to get closer to their customers than this (health workers, hairdressers, prostitutes...) and those people's business would be severely affected, but I think they'd still be less than 10%.

I don't think anybody is suggesting the latter. People are going to make decisions to limit their personal risk, and that's a good thing. If your job seems to be unsafe, and you choose to go elsewhere, that's a fine moral decision. I don't think any one person has the answer for all circumstances, and you probably know your own life best.

But people absolutely suck at assessing personal risk. "Unintentional injury" is the fourth most popular cause of death in the USA - it accounts for a full 5% of all deaths. Approximately every one of those people thought they were managing their own risk, and they got it wrong. And that's with physical injuries, which are the easiest kind of hazard to see and avoid. Don't even get me started on smokers.

Sure you know your own life best, but when it comes to making decisions about risk, people are deeply, deeply irrational. In areas like public health (which is by definition a public good), it makes sense to have rational steps mandated by a central authority.

[Liquor Box]

I'm saying that places have already paid for 6 week and 8 week lockdowns. Whether their economics are unscathed, scathed, destroyed, etc, it is what it is. So a response to a 4 week lockdown plan that projects a greater impact than those existing 8 week plans can't be 'but a 4 week lockdown would destroy the economy!' since either the economy is already being destroyed, or in fact no, a 4 week lockdown doesn't actually destroy an economy.

Well yes, places have already paid the economic (and health) prices of whatever response strategy they have implemented to date. But I don't see that helps your position.

Either we look backward at what each country should have done - and in that case the lighter the approach the less damage would have been done to the economy. Or we look forward to what shouldbe done now, in which case what has happened to date is a sunk cost, and is not relevant to the best strategy going forward - still the lighter the approach the less harm will be done to the economy (although harm already done will have to be added to that.

It's a point of logic. If you spend $8 to get 1kg of flour and I say 'but, you could have gotten 2kg of the same flour for $4 here' it doesn't make sense to say 'but $4 is too expensive to pay for flour!' as an objection to shopping at the other place. You can say 'I don't think we should buy flour at all' or 'well I already bought flour and so I don't need more right now' or even 'I know a place where you can get 2kg for $1!', but saying '$4 is expensive!' is a non-sequitur when the argument is that in fact the second place is cheaper.

I'm not sure I follow your analogy and how it applies here. Is the $ the economic damage and the flour the disease? If so, every country we are discussing needs more flour (has more porblems with the disease) and each country should be seeking the cheapest (lightest touch) price for it.....

Starting with point 2, since I think points 1, 3, and 3.5 have a shared response.

Lock-downs have had a huge variation in how they've been implemented, and some lock-down strategies seem to be much more or less effective than others. In Italy for example, you could see the moment the lock-down hit the Covid curve (at the usual 2 week delay) - it was a sharp detune from exponential. However, the peak of active cases and overall shape of the decay seems to be quite different than, say, Iran, where peak happened shortly after the de-tune from exponential and active cases dropped by over 50% in a month. China's decay curve was very striking, with a reduction by a factor of 5 in active cases a month after peak. And the US lockdowns seem to have at best slowed the growth to a steep linear rather than exponential. So details of implementation matter. I'm not aware of any place actively taking half of its essential workers and telling them to stay home for a month, work a month in shifts like that. Many places with lock-downs still permit people to go out to buy groceries under some conditions, and may use different strategies as to how to reduce contacts in those cases (the Czech Republic for example has specific shopping hours for different age groups, and no more than one person per household can go out at once; and cases in isolation in South Korea have groceries dropped off).

That's in part what gave me this thought that the degree of synchronicity of measures might be one such detail that could have a strong effect on the effectiveness of lockdowns or other Covid responses. The theoretical limit is pretty hard to argue with I think - if you literally make sure that no one comes into contact with anyone else, it should be impossible for Covid to spread. The only mechanistic counter-argument I can think of to that would be if there's some fat tail in the distribution of how long people remain infectious, such that people who have recovered from Covid can still be carriers months or years later. I don't think there's any data supporting that though - everything I've seen suggests that peak infectiousness is early.

A lot of the above is true, but I don't see how it runs counter to my point that it is not necessarily a correct assumption that 4-6 weeks will be sufficient. You mentioned that the curve flattened in Italy after two weeks, but a flattening of the curve is not the same thing as the illness having been irradicated.

There are lots of reasons why lock downs may not be effective - an example is that some people will not follow the rules (even when there are attempts to enforce them. That's been the experience here. In Australia they actually fenced off the beach and had aguard, but still surfers climbed the fence and went surfing. You are of course in Japan, which has a rule following culture, so people may follow the rules more closely there.

To the 1st, 3rd, and 3.5th points:

So with all of this, the point is to do better next time because we can't change the past. We can't change the current state of things but we can make observations about what we could have done, that we either decided not to do or were unable to do because of the situation. That way, we can honestly evaluate how much things are worth to us and decide on the trade-offs. I agree that cooperation of this scale wasn't realistic in the world of January 2020. But, if we conclude that cooperation would have decreased the economic impacts of Covid-19 by 90% so that this would just be a bump (even if it means occasionally jumping at shadows that we could have locally ignored), then we can ask 'would it make sense to build some kind of infrastructure now such that cooperation at that scale is in fact feasible the next time there's a global emergency of this form?'.

I hadn't known we were talking about the future - I'd assumed we were talking about what we should have done, or what we should do now. But everything I said applies equally. If the same thing occurs in 2050 I still don't think we could have reacted any sooner relative to the disease.

Jumping at false shadows (to the extent that the jump is to lockdown) brings huge economic costs each time. We are arguing whether a lockdown is worth the cost when we have an actual illness. If we were comparing the cost to a much less major epidemic it would be even further balanced against a lockdown. I am not yet certain that locking down this time was to great of a cost, but I am certain that locking down every ten years or so so we react more quickly to a true pandemic every 50 years or so is not worth it.

Of if that's too much in a murky uncertain future to talk about, pandemics tend to have a multi-bump structure - things appear to die down, people open up or relax, and then there's still a naive population that can be rapidly infected by the bug. So even with Covid, there's a high chance we're going to be repeating this exercise. In a few months (at least in places that are managing to get active cases down to the hundreds level) there will be some time to think about what we're going to do then, stockpile things in advance of that second bounce, etc.

And in terms of cost, imagine that we have 2 or 3 such bumps before there's a broadly available vaccine (e.g. 16 months down the road). Or worse, that we don't actually develop long-lasting immunity to Covid and so we see one of these bumps every year once a year forever on. Paying a one-time steep cost to eradicate Covid has a pay-off of whatever those risks would cost us (modulo different subjective valuations of risks versus immediate returns, where people are on the utility curve, etc). And then, to move towards the cooperation strategy, what would have to follow is a frank discussion of those relative valuations, combined with negotiations to subsidize actions that are locally against the interest of that locality but have more global valuation tied up in them than the local excess cost - and deal with all of the usual stuff that complicates such things (negotiation in bad faith, using the negotiations and the ability to force them to fail to leverage payments in excess of the break-even values, etc).

I agree that 'bumps' are realistic, but I think they are realistic even if there's a cooperative world wide lockdown for a couple of months. to me the possibility of these bumps weighs it even more in favour of just weathering them when they happen, rather than shutting the world down - possibly several times.

That part of things wouldn't be easy. But, as a researcher, it's very useful to know if you should be working on the theory of how infections spread, or if you should be working on the theory of how to converge to an efficient equilibrium in that sort of subsidy negotiation where every participant can in principle hold the entire market hostage. Going back to the point that started this line of discussion, if the major difficulties of responding to something like Covid are human factors rather than immunological factors, it's important to realize that as early as possible and start planning in those directions rather than spending a lot of time on SEIRS models and structured quarantine strategies.

I agree we should be working on the theory and try to improve thing. i also agree that cooperation is obviously the ideal strategy, but I think it's unrealistic (why would countries with very low infection rates want to cooperate with a heavy handed strategy?).

But assuming cooperation was practical, I think our main disagreement is what countries are cooperating to do. You are assuming (or perhaps explicitly saying) that the cooperation would be to simultaneously lock down. i am contending that that is not obvious. It may be a better outcome for the world to cooperate in not locking down, but instead closing borders, cancelling large events, and isolating/tracing sick people.

Well as you point out, you've already done a lockdown at these prices. So even if we quibble over the details of how much exactly the cost is, it's not salient to the argument since the argument is that spending what (or less than what) you were already willing to spend but in a different way (coordinated) would produce significantly higher return per cost.

Yes, so the questions for us might be:

• A retrospective was it worth it? i don't know the answer to that yet, but I think we'dd have a better perspective on that in a few years with the benefit of hindsight.
• A decision on what we should do now - should we come out of lockdown now, or should we continue for another week, another two , another three? That a decision our government will make next week, and every week thereafter for as long as we stay in lockdown, so it's the decision with most current practical implications.
• As you were getting at above, a decision on what we do next time.

The other thing we can do, is use the impact of our lockdown (both economic and health) as data to estimate what the impact would have been had other countries taken this option - which might inform the same points above (discussion of whether they made the correct decisions, discussion of what they should do now, or discussion of what they should do next time).

The WHO sprung the p-word on 11 March, just weeks after announcing they would no longer try to apply a technical definition to the word. They had been calling it an 'international health emergency' for some weeks by that time. On 7 March, they "reminded all countries and communities that the spread of this virus can be significantly slowed or even reversed through the implementation of robust containment and control activities."

Yes. And given those dates, along with the number here, I don't think it would have been sensible for us to lockdown any earlier than we did.

H1N1 only officially (confirmed by lab tests) killed about 19,000 people, but credible estimates put the true figure at about 284,000. The number of confirmed COVID-19 deaths is much higher, and it is likely that the reported figures are (a) being suppressed in some places for political reasons, (b) under-reported in other places because if you die alone at home, it can be weeks before anyone notices, and (c) lagging behind reality in many places because paperwork. So with the official COVID-19 toll standing at 265,000, and deaths accruing at several thousand per day, it's a bold claim that the actual toll is "still" lower than H1N1.

By mid-February COVID-19 was known to be spreading outside China, and it had already killed over a thousand people. The Chinese were officially estimating the mortality rate at between 2 and 3% - which is at least one order of magnitude higher than H1N1, and more than enough to ring alarm bells.

Wikipedia says the death toll from H1N1 was "151,700-575,400", the midpoint being about 360,000. It says that Covid 19 has killed 258,000. Even if those numbers are not accurate the point stands unless the death from H1N1 were disproportionately smaller than those from Covid 19 (at what ever point you are making a decision how to respond). The point being that whatever response we say ought to have been taken for Covid 19 also ought to have been taken for H1N1 (and other virus's before that) because at the time they appeared just as deadly.

https://en.wikipedia.org/wiki/List_of_epidemics

I agree that shutting down the world back in February would have been premature. Shutting down affected countries, however, would not. The tipping point when I started to take the whole thing seriously was late February, when cases started to appear in Australia - I knew that once it got that far, there was no chance it wouldn't also reach NZ. So "looking at matters objectively", I think it was clear by the beginning of March that this was worse than any other pandemic of recent times.

Sure, that might have been your opinion. There were undoubtedly others who reached the same conclusion weeks earlier. There were (I think including in the threads of this board) others arguing much later that it was not worse than any recent pandemic.

Incorrect - we're not 'restricted to our homes'. Under level 3, we're allowed to come and go as we like, including making longer journeys if we think they're necessary. And there are "indications" that the restrictions will lift to "level 2" next week.

I assume we are both talking about the same country - New Zealand? If so, then no not "incorrect". I said "we are largely restricted to our homes", the way it is put officially is "People instructed to stay home in their bubble other than for essential personal movement...". I'm pretty comfortable that my statement was a fair reflection of the restrictions imposed.

I don't think going into the detail of our specific lockdown requirements was helpful to the conversation we were having, but we are clearly still locked down - as noted we are not to leave our homes except for particular 'essential' purposes, the shops remain closed and most people who can't work from home aren't working.

https://covid19.govt.nz/alert-system...-alert-system/

Additionally, I don't think a relaxation next week has been signaled. We have been told that a decision wont be made before next week and it will then be 48 hours before it comes into effect. That means next Wednesday is the earliest the lcokdown will be lifted, not that it will happen that day. When we went from level 4 to level 3, we waited an extra week after the date that had been specified as 'not before'. My guess is that it's about 50/50 whether we go down a level next week.

Well, there I would agree with you. With the benefit of hindsight, it would have been better to close the border at the beginning of March and avoid the lockdown phase entirely - even though it would mean keeping the border closed for well over a year. But by mid-March, it was too late for that. The longer we left it, the more heavy-handed response was needed to contain it.

With the benefit of hindsight, I agree. that we should have closed the borders earlier. However, I am very clear that that's the benefit of hindsight, and I don't really think the wrong decisions were made at the time they were made.

It's questionable whether the heavy handed response we did take was warranted when it occurred. Possibly a less restrictive lockdown (like in Australia) or a much lighter touch (like Sweden) would have been the better approach. I'm really not sure.

As you are someone who has come to this discussion a little later, it may be worth me reiterating my overriding point. I am not saying that lockdowns are a bad idea, or that we should not have done it, or that anyone else should not do it. All I am saying is that there are competing considerations some of which favour a heavy handed approach and some of which favour a lighter touch approach. It is by no means a 'no brainer' that lockdowns are the appropriate response as some people seem to assume, it is a tough decision because lockdowns have huge consequences just like deciding not to lockdown does.

A "lighter touch" response would also have economic effects. Closing the borders destroys tourism, which is almost 6% of the economy by itself, with all the knock-on effects from that. (Of course that's happened anyway, so that's presumably included in the treasury's estimate.) Then there's the effect from lockdowns and recessions in our trading partners, which could easily amount to another 10 to 20%, depending what assumptions you want to make about them. The saying "when America sneezes, the world catches a cold" has never seemed more literal.

Completely agree. We would be sufferring economically even if the virus had never made its way here. However, locking down makes the economic damage much higher.

Between "deciding when to press the panic button" and "which button to press", there's a whole panoply of judgment calls that ultimately can only be made by the politicians who have been given that responsibility. And one thing we can rely on with absolute certainty is that whatever they do, their opponents will excoriate them for it. That's what political opponents do. But the real question is: do we believe the other lot, faced with the same choices, would have done better?

Not "could". Obviously they "could" have done better. But would they, really? On balance, which lot would you rather trust? Because you've got to trust someone. There is no opting out.
That's not a question that can be debated here, because politics. But I think it's a question worth thinking about.

Completely agree. As noted above, I am not saying that we made the wrong decision. I am only saying that there are competing considerations. There was strong justification to do what we did, but there's also strong justification to take a very light touch , and strong justification for the range of options in between. It's a tough call. As noted above, I am just objecting to those who seem to assume that lockdown is the obvious answer that all countries should have made. It's simply not clear.

I also agree that each country's governemnt is obviously the appropriate person to be making decisio. Also agree that we should go any further with that because of the prohibition on discussing politics (I guess we don't have to because we agree).


In conclusion, I will say that as mcuh as I don't thin its obvious whether NZ's approach was best, or the approach of these countries who have taken a much lighter approach was best, I do think this may become obvious when we look back at this in the future with the benefit of hindsight.

[pendell]

Has anyone mentioned Folding At Home's crowdsource virus simulation? The idea is to simulate viral proteins in the hopes of finding spots drugs can bind to. If you can afford the power and can spare a few CPU cycles, it's one way to make headway against the virus. Heck, we might even form a Playground Team if there's enough interest.

Respectfully,

Brian P.

[Friv]

I don't know what you main by not getting any weighting for myself and famly and friends. One of the lens I do look at it through is what I would rather for myself. I think that is sensible lens for each of us to look at it through.

So which of these risks would I rather run (or would you rather run):

• a 30% chance of contracting the virus meaning a 0.2% chance of dying as well as a 5% chance of losing my job (and therefore my home) due to economic impact (if there's a light touch reaction); or
• a minimal chance of contracting the virus and dying as well as a 30% chance of losing my job (and therefore my home) due to the economic impact (if there's a heavy handed reaction).

First, don't forget that in your first analysis, you're going to want to add in the chance that you will be responsible for the death of a loved one after you get sick and then get them sick. Secondly, don't forget that even if you survive, there's a good chance of permanent lung or heart damage; the likelihood of that is still being assessed, but it's pretty bad. Third, consider that if you get sick, you loop back around to having a chance of losing your job and your home. Fourth, even if you go into your job, there's a good chance that other people aren't going to visit it and you'll lose it anyway and also get sick. (This is actually a big one - in most countries, if lockdown procedures end, businesses can't take money to stay closed, but if they don't get enough clients they'll go bankrupt.)

And of course fifth, that risk analysis is why governments are providing money to people laid off, so that they won't lose their houses and can continue to eat, and why they're calling on rent and mortgage forgiveness. Obviously, if people have to choose between sickness and starvation, they're going to go with the one that might not kill them. I'm not sure we can get further into that branch of things without this becoming a political discussion, though.

(Oh, and I guess sixth, the current worldwide mortality rate for confirmed COVID cases is 3.4% Even if we assume that 80% of cases aren't being reported, but all mortalities are, which is a pretty lax assumption, that would leave the mortality rate about five times what you're estimating.)

[Tyndmyr]

I'm saying that places have already paid for 6 week and 8 week lockdowns. Whether their economics are unscathed, scathed, destroyed, etc, it is what it is. So a response to a 4 week lockdown plan that projects a greater impact than those existing 8 week plans can't be 'but a 4 week lockdown would destroy the economy!' since either the economy is already being destroyed, or in fact no, a 4 week lockdown doesn't actually destroy an economy.

This doesn't follow. First off, economics isn't a simple binary of "destroyed/not destroyed", but instead, damage is cumulative. We use destroyed as an easy label to describe when something has gone very poorly. This does not mean things cannot get even worse. Things can pretty much always get worse.

Lock-downs have had a huge variation in how they've been implemented, and some lock-down strategies seem to be much more or less effective than others. In Italy for example, you could see the moment the lock-down hit the Covid curve (at the usual 2 week delay) - it was a sharp detune from exponential. However, the peak of active cases and overall shape of the decay seems to be quite different than, say, Iran, where peak happened shortly after the de-tune from exponential and active cases dropped by over 50% in a month. China's decay curve was very striking, with a reduction by a factor of 5 in active cases a month after peak.

As China's numbers are known to be pure garbage at this point, relying on the shape of their curve seems incredibly unwise. Iran vs Italy is also poor support. Again, it's cherry picking two countries, and countries that are not extremely similar. On the one hand, Italy had roughly twice the ratio of sick people, which is pretty bad, but on the other, Iran's infections/day number is currently rising, not decreasing. Neither seems like a good example of lock-downs working, or a model to follow overall.

And the US lockdowns seem to have at best slowed the growth to a steep linear rather than exponential. So details of implementation matter. I'm not aware of any place actively taking half of its essential workers and telling them to stay home for a month, work a month in shifts like that.

This is pretty common here in Maryland. A lot of stuff is on minimal manning. Probably not everything, because you can't split every job in half like that. Sometimes you need more than half the given positions, because they are essential. You can't actually cut food workers in half and expect the same amount of food.

Some places are desperately hiring, even. Grocery stores, for instance. People are leaning on them more heavily given the lack of restaurants, so their workload is actually increasing. They couldn't let half the people go home and keep up.

That's in part what gave me this thought that the degree of synchronicity of measures might be one such detail that could have a strong effect on the effectiveness of lockdowns or other Covid responses. The theoretical limit is pretty hard to argue with I think - if you literally make sure that no one comes into contact with anyone else, it should be impossible for Covid to spread. The only mechanistic counter-argument I can think of to that would be if there's some fat tail in the distribution of how long people remain infectious, such that people who have recovered from Covid can still be carriers months or years later. I don't think there's any data supporting that though - everything I've seen suggests that peak infectiousness is early.

There are also other methods of transmission that are not direct person to person. Infected person A touches object, uninfected person B touches it and becomes infected. This is usually a lot less dangerous than person to person contact, but even if we assumed an unusual example of nobody coming into direct contact with anyone else, spreading would not be impossible, though it would be a great deal less likely.

Well as you point out, you've already done a lockdown at these prices. So even if we quibble over the details of how much exactly the cost is, it's not salient to the argument since the argument is that spending what (or less than what) you were already willing to spend but in a different way (coordinated) would produce significantly higher return per cost.

The assumption that unified action is better than diverse action is just that, an assumption.

We are frequently working with incomplete data, and frankly do not know...quite a lot. Massively staying at home is essentially a giant experiment. If you guess absolutely perfectly correct, then yes, everybody doing your guess may be best. It is unlikely that working from wildly incomplete info, we guessed perfectly on the first try. For *any* particular strategy/country/etc. This is just the nature of working with unknowns.

Areas trying many different strategies at least provides us with more information about what works and what doesn't.

But it's not linear. If you look at some traffic models, you'll quickly notice that a 10% increase in traffic results in a much greater than 10% increase in congestion. If we all abandon the 10% of our daily activities that involve being within 2 metres of someone not in our "bubble", then the hit to the economy would be about 10%, but the effect on virus transmission would be significantly greater. Some people need to get closer to their customers than this (health workers, hairdressers, prostitutes...) and those people's business would be severely affected, but I think they'd still be less than 10%.

This assumes that the 10% cut is the 10% it is most optimal to cut. I am not sure this is the case. Generally speaking, large, heavily trafficked businesses remain open, because they are more likely to have, somewhere in their product mix, goods that qualify as essential. Smaller specialty shops are more likely to be closed entirely. Now, I'm not going to say that the small business with a handful of customers per day is entirely risk free, but it probably has a good deal less incidental contact than going to Walmart.

Even recreational activities like boating, which can be done entirely solo, or with people you live with anyways, are frequently locked down(including where I am). I suppose some tiny fraction of people might have giant party boats, but the vast majority of people do not. So, the targeting does not seem to be anything like what you assume.

But people absolutely suck at assessing personal risk. "Unintentional injury" is the fourth most popular cause of death in the USA - it accounts for a full 5% of all deaths. Approximately every one of those people thought they were managing their own risk, and they got it wrong. And that's with physical injuries, which are the easiest kind of hazard to see and avoid. Don't even get me started on smokers.

Those are largely automobile accidents, in which the person who dies may unfortunately not be the person who made the risky decision.

Additionally, the fact that a small minority of people die to accidents does not mean that they suck at assessing risk. The risk from not having a job may have indeed been higher than the risk of dying in a car accident on the way to work. Choosing to drive to work doesn't mean that person is stupid, merely unlucky. Even a lower risk is still a risk.

Yeah, 5% of people die to accidents, but 95% of people die to everything else. It's pretty reasonable to worry more about everything else. Most people are not stupid, they simply don't always have the luxury of pursuing ideal choices, so they're doing the best they can with what they have available.

[NichG]

Well yes, places have already paid the economic (and health) prices of whatever response strategy they have implemented to date. But I don't see that helps your position.

Either we look backward at what each country should have done - and in that case the lighter the approach the less damage would have been done to the economy. Or we look forward to what shouldbe done now, in which case what has happened to date is a sunk cost, and is not relevant to the best strategy going forward - still the lighter the approach the less harm will be done to the economy (although harm already done will have to be added to that.

I'm not sure I follow your analogy and how it applies here. Is the $ the economic damage and the flour the disease? If so, every country we are discussing needs more flour (has more porblems with the disease) and each country should be seeking the cheapest (lightest touch) price for it.....

The $ is the immediate economic damage, and the flour is 'lives saved' or 'mitigation of future economic risks'. Lets say both have some kind of valuation - you really think flour is important, or you really think $ is important, or some combination of the two (in general, this doesn't have to be linear where more $ or more flour is always equally good - but it should be monotonic).

If I can make an exchange for flour in as small increments as I want, and there are a bunch of sellers offering different price points, then we can distinguish between strategies (which sellers to purchase from) which are equally rational but correspond to different relative valuations, and strategies which are irrational - in the sense that no matter what your valuation is, you would be better served by picking some other purchase strategy. The strategies corresponding to different valuations form something called a Pareto front - a boundary of a region in the space of trade-offs - with strategies lying within that region rather than at it's boundary being strictly sub-optimal no matter how you value various resources. The axes of this space are how much flour you end up with and how much cash you end up with/spend, and if your value function is linear in those quantities then the Pareto front is a convex polygon.

So my argument is, if you give me a strategy where you buy X units of flour for Y $ and you tell me that that strategy reflects your values as to how important it is to have flour or how important it is to have $, then if I give you a deal to buy X units for Y/2 $, you should prefer it. There are rational reasons not to prefer it, but they involve bringing extra elements into the scenario - this is an iterated social game and the strategy you're seen to choose will impact future costs, you have a non-monotonic value function, there are other variables which you value but we haven't included, etc.

It does become a bit more complex if we consider that sellers have minimum and maximum quantities of flour they're willing to sell. If e.g. you can buy 10kg of flour for $2 or 1kg of flour for $1 or no flour, then there are value functions for which the $1 for 1kg deal can be better to take. But if you've already shown that you're willing to pay $3, then that does still exclude the $1 for 1kg deal as being a rational choice for your value function.

So my argument is, if you've already shown a willingness to pay 6 weeks of lockdown for a partial reduction in Covid, you should be willing to pay 4 weeks of lockdown for a total removal of Covid.

A lot of the above is true, but I don't see how it runs counter to my point that it is not necessarily a correct assumption that 4-6 weeks will be sufficient. You mentioned that the curve flattened in Italy after two weeks, but a flattening of the curve is not the same thing as the illness having been irradicated.

There are lots of reasons why lock downs may not be effective - an example is that some people will not follow the rules (even when there are attempts to enforce them. That's been the experience here. In Australia they actually fenced off the beach and had aguard, but still surfers climbed the fence and went surfing. You are of course in Japan, which has a rule following culture, so people may follow the rules more closely there.

In Japan there have basically been no rules at all at the government level other than the school closures. Just guidelines, and people don't particularly follow them more than other places as far as I can tell... Individual companies have their own policies, but again you hear things of people justifying going to the office every day just because they have to physically stamp paperwork with their personal seal.

But yes, if you're going to do this, 10% of people violating the lockdown could add another month, and if it's >40% then you'll still see growth even in lockdown (since R0 is ~2.5, if 40% violate lockdown then you'll have continued linear growth of cases even during lockdown and if slightly more than 40% violate lockdown you'll still see exponential growth).

I hadn't known we were talking about the future - I'd assumed we were talking about what we should have done, or what we should do now. But everything I said applies equally. If the same thing occurs in 2050 I still don't think we could have reacted any sooner relative to the disease.

I mean, unless you have a time machine... I'm not really thinking 'in 2050' here, I'm thinking 'in July or August this year when we're dealing with resurgences starting to hit countries that recently successfully passed below the 100 cases mark'. We did a set of interventions, it had an outcome, we analyze that outcome, use mechanistic understanding to evaluate counterfactuals and alternative strategies we could have used, use the time we have to prepare for the things we can predict are coming, and it's cheaper when we next have to do all of this again.

Because of the human factors, a lot of this is also about discussing what values people place on things ahead of time. Because doing all of that in the public eye amidst an emerging disaster has added complications I won't get into here.

I agree that 'bumps' are realistic, but I think they are realistic even if there's a cooperative world wide lockdown for a couple of months. to me the possibility of these bumps weighs it even more in favour of just weathering them when they happen, rather than shutting the world down - possibly several times.

I agree we should be working on the theory and try to improve thing. i also agree that cooperation is obviously the ideal strategy, but I think it's unrealistic (why would countries with very low infection rates want to cooperate with a heavy handed strategy?).

Well, the point is that a country with a very low infection rate is still destined to be a country with a high infection rate at some future time. So this is sort of like asking 'why would I pay $3 now to save $6 in three months time?'. Unless you have some circumstance where you really need that $3 now, and you won't need $6 in three months, it's generally a good deal.

But assuming cooperation was practical, I think our main disagreement is what countries are cooperating to do. You are assuming (or perhaps explicitly saying) that the cooperation would be to simultaneously lock down. i am contending that that is not obvious. It may be a better outcome for the world to cooperate in not locking down, but instead closing borders, cancelling large events, and isolating/tracing sick people.

Yes, so the questions for us might be:

• A retrospective was it worth it? i don't know the answer to that yet, but I think we'dd have a better perspective on that in a few years with the benefit of hindsight.
• A decision on what we should do now - should we come out of lockdown now, or should we continue for another week, another two , another three? That a decision our government will make next week, and every week thereafter for as long as we stay in lockdown, so it's the decision with most current practical implications.
• As you were getting at above, a decision on what we do next time.

The other thing we can do, is use the impact of our lockdown (both economic and health) as data to estimate what the impact would have been had other countries taken this option - which might inform the same points above (discussion of whether they made the correct decisions, discussion of what they should do now, or discussion of what they should do next time).

This issue though is that if you have one country that did a severe lockdown who has an important trading partner who did a mild one or no lockdown at all, then the overall course of the pandemic in the first country is going to reflect the policy of their partner and not their own. That's the 'cooperative' aspect of this game. If I play X and you play Y, then X was a bad move. If I play X and you play X, then X was a good move.

So I expect we'll start to see lots of examples of countries that had severe policies and in the end weren't spared a second bump - because their policies can't be evaluated individually in isolation.

[Aedilred]

So my argument is, if you've already shown a willingness to pay 6 weeks of lockdown for a partial reduction in Covid, you should be willing to pay 4 weeks of lockdown for a total removal of Covid.

Not necessarily.

It's not either/or. Yes, in isolation, a smaller cost for a better result is obviously better. But it's not six weeks versus four weeks, because that completely ignores the fact that the cost is cumulative. It's six weeks versus ten weeks. The idea that you've already tolerated six so you should tolerate four more is the sunk cost fallacy Liquor Box (with whom I almost entirely agree) was talking about.

Then there are the subjective factors. We do not know whether another four weeks will be sufficient to completely "remove" the virus. It almost certainly won't be. It would certainly suppress the virus, but it can clearly spread very quickly, so even a couple of surviving cases would be enough to let it loose again. The only way the virus can be "removed" is the population developing, whether naturally or artificially (i.e. by vaccine) a sufficient herd immunity that it can no longer spread, and that's a lot further off than four weeks.

It also presupposes that it's agreed that the six-week lockdown was worth it. That remains an open question. If it wasn't, then that's enough to render the whole argument a non-starter.

Lockdown sceptics have nevertheless been for the most part willing to give the lockdown the benefit of the doubt until now in the hope that it would actually work, but that is not the same as agreeing that it has been worth the cost. As Liquor Box has implied, we probably won't know whether it is worth it until all this over, and even then that will be subject to both some speculation as to data and individual opinion.

In short, whether one considers the lockdown to be worth it depends on a number of factors, most notably:

1. Whether and if so to what extent actually works for its intended purpose of containing the virus;
2. How much damage to the economy results from it;
3. Where one considers the balance of interest to lie between the damage done to the economy from a lockdown versus the likely death count from the virus;
4. One's overall appetite for risk.

We don't yet know the answers to (1) and (2), and both (3) and (4) are matters of opinion. So there can as yet be no conclusive answer.

The argument that the lockdown should have been put up earlier and should remain in place for longer however takes (1) for granted, assumes that (2) is within tolerance, and assumes both a fairly extreme position on (3) (which in turn raises the tolerance level for (2)) and a very cautious risk profile under (4).

It is a reasonable position to take, but other positions of equal validity exist.

[Rockphed]

It is a reasonable position to take, but other positions of equal validity exist.

I'm not sure where I sit on the issue of whether lockdowns were worth it, but I find the tendency to declare any other opinion invalid more annoying than either the pro or con arguments.

[Tyndmyr]

I'm not sure where I sit on the issue of whether lockdowns were worth it, but I find the tendency to declare any other opinion invalid more annoying than either the pro or con arguments.

Agreed. A *lot* of folks want to declare something a nice, easy answer, despite the lack of data and everything else. There's significant uncertainty still, and a lot of caution is warranted.

The strategies corresponding to different valuations form something called a Pareto front - a boundary of a region in the space of trade-offs - with strategies lying within that region rather than at it's boundary being strictly sub-optimal no matter how you value various resources.

There are sufficient unknown variables that the idea of being at Pareto optimum is an impossibility. It's not a realistic real world tradeoff, and that's a poor way to model this problem. It's of limited use outside of the hypothetical almost everywhere in real world economics, because we almost never have optimized everything to the point that only tradeoffs remain, and usually do not possess perfect information, so we cannot reasonably do so. When dealing with a previously unknown pandemic, the uncertainties are much greater than normal. We're mostly throwing stuff at the wall that seems like a good idea, and seeing what sticks.

Now, we are actually getting some early data that might be relevant to the shutdowns in particular: https://www.forbes.com/sites/lisette...U#6ef121ad1655

Cliff notes: In NYC, two thirds of the infections were among locked-down people. Folks not going to work, school, or travelling. This significantly calls into question how well lockdown is working in NYC at least. That said, it's location specific, and may have geographical limitations. The close proximity of NY apartments comes to mind as a possible factor to make it potentially less effective.

[Kaptin Keen]

I'm not sure where I sit on the issue of whether lockdowns were worth it, but I find the tendency to declare any other opinion invalid more annoying than either the pro or con arguments.

I feel it's been shown that the lockdowns were prudent - but overdone. Sweden being my main example of a less extreme reaction that hasn't led to disaster. I personally feel the swedes have been too laisse faire about the whole thing, but ... well, their unlikely to drop by to ask my opinion on the matter =)

Oh, also ... the lockdowns propably need a scale of severity based on some little known factors such as population density, average age and health in the population. Denmark has changed recommendations and warnings, so that the single group most at risk is the obese. 36% of everyone comitted have been overweight, more than any other subgroup. I mean, age is still a major factor, but that's not surprising. If you're old and overweight, you're apparently in more trouble than if you're old and diabetic, but well controlled.

Frankly: This is good. We lucked out BIG TIME. The first really big pandemic to hit us isn't particularly dangerous, and now we have time to learn and prepare for one that's actually dangerous. As world population crawls towards 10 billion, we will have more of these. And they won't all be 'just the flu'.

[Ibrinar]

Not necessarily.

It's not either/or. Yes, in isolation, a smaller cost for a better result is obviously better. But it's not six weeks versus four weeks, because that completely ignores the fact that the cost is cumulative. It's six weeks versus ten weeks. The idea that you've already tolerated six so you should tolerate four more is the sunk cost fallacy Liquor Box (with whom I almost entirely agree) was talking about.

NichG was discussing a theoretical perfect and coordinated lockdown in the beginning not in addition to the current ones. "Instead of" instead of in addition is correct in context.

[Vinyadan]

To lighten the mood: https://assets.amuniversal.com/1e23e...ea005056a9545d (it's today's Dilbert strip.)

[Sian]

https://news.cgtn.com/news/2020-05-0...k9y/index.html

Apparently there is some evidence suggesting that the first retroactively recognized case in Europe is from back at 16th of November (a day before the first registered case in China), in and surrounding the City of Colmar in Alsace, France ... which is a minor hotspot for Chinese tourism due a season of a Chinese culinary celebrity/reality show was filmed there

[Rockphed]

https://news.cgtn.com/news/2020-05-0...k9y/index.html

Apparently there is some evidence suggesting that the first retroactively recognized case in Europe is from back at 16th of November (a day before the first registered case in China), in and surrounding the City of Colmar in Alsace, France ... which is a minor hotspot for Chinese tourism due a season of a Chinese culinary celebrity/reality show was filmed there

I've seen people suggest that what we are dealing with now is the second wave of the pandemic. Any idea how correct that idea is?

[warty goblin]

https://www.forbes.com/sites/lisette...U#6ef121ad1655

Cliff notes: In NYC, two thirds of the infections were among locked-down people. Folks not going to work, school, or travelling. This significantly calls into question how well lockdown is working in NYC at least. That said, it's location specific, and may have geographical limitations. The close proximity of NY apartments comes to mind as a possible factor to make it potentially less effective.

I have a hard time getting from that number to any meaningful inference about the effectiveness of lockdowns. What that gives you is probability of being a stay-at-home person, given you are in the hospital, and P(at home | hospitalization) != P(hospitalization | at home), which tells you something about the effectiveness of the lockdown. But without knowing P(at home), you can't do the Bayes' Rule to flip the conditioning, so it doesn't really tell you anything about how well the lockdown constrains hospitalization, let alone transmission. Since essential vs. non-essential workers are reasonably likely to have different age and comorbidity distributions, therefore different hospitalization rates, it's very far from getting at infection rates.