The Corona Virus

[tazbot]

Just saying, this is worse than traffic, where it has manifested itself, and it's still ongoing and expanding. The pandemic has now reached more peripheral areas like Russia and South America, then I assume it will blow up in Africa, although there's the advantage of a largely young population there. And we don't know what is going on in China.

What's going on in China is life is going back to (a new) normal. People are working, children are starting to go to school (depending on your age, each province has their own timeline for this), rush hour is back, restaurant and shops are open, some entertainment venues are open (no cinemas or swimming pools though, and most sports venues don't dare open yet), traveling within the country is allowed to almost everyone. The controls (temperature checks, leaving your contact info when entering certain places, etc.) are still in place, but not as rigorous as before. Some people are easing off on the masks but most still wear them even when not required to.

I work with people like this. Some of them are very smart. They're also convinced that this is just a hoax, and everyone wearing a mask or who isn't working is a sheep and a coward. When it hits our office, I guarantee you every one of the ones like that who gets it and lives will use that as proof that it's not that bad.

As far as I can tell, it really isn't that bad for most people. From looking at death rates of countries with wider testing, the death rate isn't all that high. What *is* bad is how contagious this is, and that someone without symptoms can infect others. It's impossible for you to not be aware that you have a flu, and most people next to you can probably figure out your sick too, so everyone takes precautions in this case. If there're no symptoms, you'd have to take precautions even when feeling healthy (a big argument for wearing a mask), which is much harder to do and easier to forget about.

[Ibrinar]

As far as I can tell, it really isn't that bad for most people.

That is the point I think. You probably (depending on age and other risk factors) won't die from it personally. That would be true even if it was much more devastating and killed say 10% of infected. (Well I guess even with 90% the survivors would by definition be the only ones talking about it, but the mountains of dead would probably suppress that.) So if someone argued from their own survival that it is not that bad they either don't understand math very well or are being intellectually dishonest. (I suppose not deadly but getting really really ill would probably also put it into the serious category for most. )

[Tyndmyr]

There's probably a difference between the effect of unemployment during an economic boom, versus the effect of unemployment during an overall recession. This paper for example seems to show across-the-board decreases in mortality in Europe in response to the 2007 recession (unfortunately, paywalled so I can't dig into the details): https://onlinelibrary.wiley.com/doi/....1002/hec.3495

The link I provided was regarding mass layoffs as part of a declining economy. It seems exactly applicable.

A lot of where studies get conflicting results in this area seems to be based on stratifying the data to see specific effects. In the paper I just linked, they see the strongest pro-cyclical effects with accidents

Yup. You can crunch data many ways, and some of them can be not terribly informative.

In the paper you linked, it was restricted to the age group of older men.

There's not a lot of women in the PA oil industry. That's the demographic that works there. However, that was only the first study cited by the source I linked. The next one was a 2009 study that was US workers in general, so you can see the effect exists in general, not merely in specific circumstances.

E.g. there may be things that increase your personal risk if you're unemployed but the people around you keep working, which are strongly outweighed by risks generated as byproducts of higher rates of overall economic activity - increases in traffic, pollution, spread of communicable diseases, etc. They also find that in the second paper I like - the pro-cyclical effect is a collective one, not an individual one. It's good for everyone, death-wise at least, when everyone doesn't work, but you wouldn't increase your life expectancy by quitting your job.

Wealth and expected lifespan are strongly correlated on absolutely every level from the individual to the country. This isn't even a controversial thing. Many of the studies focus on putting specific numbers on specific risk groups because the overall effect is so well known.

To put a number to this, it looks like roughly a 1% increase in unemployment tends to co-occur with a 0.5% decrease in mortality rates, both in Europe, and in older US studies I've found such as: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6028024/

In that study for example, the rate of auto accidents drops by 2.6% with a 1% increase in unemployment. It also shows that these effects are strongly age-stratified. A 1% increase in unemployment gave a 2% reduction in mortality rate in the 20-24 year old range, a 0.4% reduction for females overall, and a 0.2% reduction for males overall.

If you read deeper into it, you'll see that this was the *only* instance of where their predictions pan out, and they admit that the effect of more drivers on the road, primarily drunk drivers, cannot explain their overall hypothesis. More money => More ability to buy booze and drive, sure. That's a bad effect, but it's hardly the only one, and while sure, drunk driving is bad, it's only one cause of death among many. Suicide rates rising in response to layoffs is fairly well documented, for instance.

If we look at death rates over time, there's one obvious trend:



They are falling overall. Even, say, WW2 doesn't change that trend, and the war/postwar period of prosperity coincides with the smoothest, strongest gains on the chart. WW2 itself wasn't a helpful factor, since it adds a lot of excess deaths, so what explanation do you offer for this dropoff?

The US had a quite low prison-adjusted* incarceration rate in this time period.

Now, it's also true that the rate fell during the Great Depression, but the rate at which it fell did not appear to be unusually rapid for the long term trend-line. You've just got a massive spike in death rates earlier due to WW1/spanish flu. Save for that outlying spike, the great depression appears to see mortality decreasing at perhaps slightly lower than average rate. There isn't much of an argument that the great depression is actually *good* health wise, though it does seem likely that there are non linear effects due to unemployment**.

All data gets a little shakier when compared against CV unemployment for a few reasons. Mostly, we have literally nothing on this scale to compare against. The volume and speed is unprecedented, even in the great depression.



That was merely the first week. We're at north of 33mil so far, and this change was far, far more abrupt than any other ever has been. This is *probably* a bad thing in the same way that the great depression was generally bad, but we really don't know how bad. In part, because of non-linear effects, and in part because we don't know the duration. Being off work for a month and then being rehired may mitigate many negative effects in comparison to being off work for a much longer time.

Basically, we're not discussing if this can get worse than the great depression. It already cleared that bar in April***, and things are a lot worse now. We're discussing if we can patch it before we start to see all the consequences.

*Some measures consider those imprisoned to be unemployed, while other measures do not include them. I think it's probably most accurate to include them for comparing against death rate, since those imprisoned are part of the latter as well.
**While suicide was up during the depression, for instance, it would be difficult to argue that it is purely linear. A 1% unemployment bump may not be exactly a tenth as bad as a 10% unemployment bump.
*** Since methodology matters, this is U-6, including the post-April corrections to the unemployment rate because of inability to process everything rapidly enough, which hit 27.6%. The worst rate during the great depression was 24.9%

On masks :
the main advantage of a policy of "wear masks" is NOT that your mask will protect you from catching the virus.

It is that the mask will protect you from SPREADING the virus (from yourself to others).

You're not protected by your own mask, but by the masks others are wearing.

It does seem to me that there is a large consensus on the fact that wearing a mask make you less likely to infect others (the WHO recommends wearing a mask if you have symptoms, for example).

So I cannot really understand WHY it isn't obvious that a policy of "let's have everyone wear masks" would lower the spreading of the disease (it might be too costly or cause a shortage for medical staff, but it's not the same thing as being ineffective).

Exactly. I support local guidance on the correct course of action because...if you have a mask shortage, and want to concentrate them among the highest risk folks(health care workers, etc), sure...that makes sense. But that doesn't mean masks are harmful.

It's fairly well understood that masking sick people mask is helpful, and it's common medical practice. In this case, if we want to mask *all* the sick people, we effectively have to mask everyone. It's the result of being contagious and asymptomatic.

Everyone wearing face protection and an N95(or better) would be ideal, but measures falling well short of this will probably still help some.

"That time period" was well after WW1 and the flu pandemic, which (thanks partly to some ruthless social distancing mandates) had burned out by the end of 1920.

Yes, there was a spike in suicides when the great crash struck, but total numbers were never more than about 1.5% of all deaths. The small increase was way more than offset by the reduction in deaths from road traffic accidents alone, which went sharply down during the depression.

But, interesting thing about suicide: rates had already been rising quite rapidly through the roaring 20s. They also rose sharply during the 1980s, and they've been climbing steadily since 2000. The correlation with economic prosperity or otherwise doesn't seem to be as simple as you might expect.

It's complicated, to be sure, and there are many other factors. However, on a broad scale, economic prosperity and health are strongly linked.

We're also no longer quite the same society we were in the great depression. My grandfather lived through it and used to talk about it a lot. He said it wasn't all that bad in the country, because people just stopped buying things and were mostly self sufficient, growing their own food. In the cities, it was different. Well, we're a lot more urbanized now, and I would expect that far fewer people are self sufficient, growing their own food and such. Mostly we rely on factory farms and that whole production chain. Personal budgets may be less flexible today. So, all of that complicates the math when we're comparing against an event nearly a hundred years ago. It's the most similar in pure economic magnitude, but it's so long ago that it may not accurately capture the impact.

[Joran]

That is the point I think. You probably (depending on age and other risk factors) won't die from it personally. That would be true even if it was much more devastating and killed say 10% of infected. (Well I guess even with 90% the survivors would by definition be the only ones talking about it, but the mountains of dead would probably suppress that.) So if someone argued from their own survival that it is not that bad they either don't understand math very well or are being intellectually dishonest. (I suppose not deadly but getting really really ill would probably also put it into the serious category for most. )

The best way I saw it put was "You're probably not going to die, but you'll know someone who did."

So far, I only know people tangentially (relative of my daughter's teacher, friend's grandmother), but we're not out of the woods yet.

[NichG]

The link I provided was regarding mass layoffs as part of a declining economy. It seems exactly applicable.

Yup. You can crunch data many ways, and some of them can be not terribly informative.

There's not a lot of women in the PA oil industry. That's the demographic that works there. However, that was only the first study cited by the source I linked. The next one was a 2009 study that was US workers in general, so you can see the effect exists in general, not merely in specific circumstances.

Wealth and expected lifespan are strongly correlated on absolutely every level from the individual to the country. This isn't even a controversial thing. Many of the studies focus on putting specific numbers on specific risk groups because the overall effect is so well known.

Your examples all seem to be US-based, and the thing I linked did say that the pro-cyclic effect has vanished in the US since 1995. Previously you had said you were interested if there was data for any countries other than the US where mortality and unemployment weren't correlated and there was - in Europe in the 2007 recession.

So I don't think US-specific counter-examples are necessarily relevant given the context that spun off this side-thread, which iirc was a comment from me about intending to move to a country where I'd heard that it was pretty hard to die of poverty.

I wouldn't be that surprised if the US reverses the trend since many things like access to health care are employer-provided in the US. But I don't think the US is a good example of the bounds of the possible here - it's a particular context where a certain direction of structuring things has been chosen, but that structure is actually pretty rare outside of the US. Even in Japan for example, which is pretty similar to the US in the way it does a lot of things, there's a mandatory state-provided coverage that you get if you're unemployed and which you have to buy into if your employer doesn't provide coverage of their own.

If you read deeper into it, you'll see that this was the *only* instance of where their predictions pan out, and they admit that the effect of more drivers on the road, primarily drunk drivers, cannot explain their overall hypothesis. More money => More ability to buy booze and drive, sure. That's a bad effect, but it's hardly the only one, and while sure, drunk driving is bad, it's only one cause of death among many. Suicide rates rising in response to layoffs is fairly well documented, for instance.

As has been mentioned earlier in the thread, suicides are a relatively small contribution to total death rates, so increases there can be swamped by decreases elsewhere. In terms of causes of death, suicides are 20 times lower than cardiovascular disease, 10 times lower than cancer, 7 times lower than combined respiratory diseases, 3 times lower than dementia, and about half the level of traffic accidents.

So a 10% reduction in respiratory diseases and traffic accidents for example can counterbalance almost a 100% increase in suicides.

If we look at death rates over time, there's one obvious trend:



They are falling overall. Even, say, WW2 doesn't change that trend, and the war/postwar period of prosperity coincides with the smoothest, strongest gains on the chart. WW2 itself wasn't a helpful factor, since it adds a lot of excess deaths, so what explanation do you offer for this dropoff?

Doesn't seem particularly relevant to this discussion, unless you're arguing that in general unemployment has fallen over the course of human history or some other long time-scale correlation.
From what I can tell, (in the US) it's fluctuated between 4% and 10% with occasional excursions (https://en.wikipedia.org/wiki/Unempl..._1892-2016.png). So 'more people are working and so are in a good economic position' can't explain the overall trend-line.

At best, the conclusion one would take from that dropoff is 'employment status has very weak effects on mortality compared to other factors that don't have to do with employment status at all'. Which also sort of denies your argument about mass unemployment causing lots of deaths. So I don't think this graph helps your argument at all.

[Aedilred]

Because wearing a mask means admitting that this is a serious problem. It's why the same people are arguing for reopening AND against mask-wearing AND against more aggressive testing and tracing.

This is a bit of a generalisation. I am broadly in favour of reopening, the sooner the better. I am also wholly in favour of more aggressive testing. I am in favour of tracing, provided we deal adequately with privacy/surveillance concerns (about which I will say no more here because of politics). I'm not really in favour of masks (see below) but if it's actually going to have a positive effect and it's the (temporary) price we have to pay to restart some semblance of real life then ok.

So at least in my case - and I don't think I'm alone - it's not nearly as straightforward as you suggest.

I can't pin down my objection to masks precisely. Part of it is personal and selfish: it would be uncomfortable and inconvenient and annoying and it's already hot enough at the moment without having something stuck to my face. There may be some psychological issues too, because I think I draw a mental association between people wearing masks in day-to-day life and dystopia - and I don't want to live in any more of a dystopia than I already do.

More generally I'm also not convinced that it'll make that much of a difference, and I am sceptical of "pandemic theatre" in the same way I am of "security theatre" or for that matter "green theatre": i.e. that certain measures that impose inconvenience on people in the name of a greater cause are sometimes in fact principally cosmetic and intended largely if not entirely to reassure people that the issue is being taken seriously and addressed - while doing nothing to address the actual problem and even at times hindering the overall cause.

[Joran]

I can't pin down my objection to masks precisely. Part of it is personal and selfish: it would be uncomfortable and inconvenient and annoying and it's already hot enough at the moment without having something stuck to my face. There may be some psychological issues too, because I think I draw a mental association between people wearing masks in day-to-day life and dystopia - and I don't want to live in any more of a dystopia than I already do.

I understand and sympathize. I have a homemade mask and quite a few disposable surgical paper masks and I wear them in stores. I notice that I grind my teeth whenever I'm wearing the masks and it's not comfortable. Longest I've done is about half an hour in a Costco. If I'm indoors, I'm wearing a mask though, no question.

I also do not wear a mask outdoors since I'm avoiding crowds, taking a walk, and generally the closest I'll be is a momentary brush past someone.

One thing someone suggested was a bandana for outside, occasional use. I could move it up and down as needed, like coming across a stagecoach I need to rob, and should be more comfortable than a more fitted mask.

More generally I'm also not convinced that it'll make that much of a difference, and I am sceptical of "pandemic theatre" in the same way I am of "security theatre" or for that matter "green theatre": i.e. that certain measures that impose inconvenience on people in the name of a greater cause are sometimes in fact principally cosmetic and intended largely if not entirely to reassure people that the issue is being taken seriously and addressed - while doing nothing to address the actual problem and even at times hindering the overall cause.

The science seems pretty good that masking makes a difference, given the nature of transmission (droplets), especially if everybody's doing it. It is true that nothing short of a properly fitted N95 mask + faceshield is going to 100% protect you, but even a cloth mask seems to make a difference in protecting you and it makes a big difference is protecting everyone else if you're an asymptomatic carrier.

https://www.preprints.org/manuscript/202004.0203/v1
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7191274/

The "Pandemic theater" is people wiping down packages and their mail. Contaminated surfaces seems like a less important avenue of transmission.

[Tyndmyr]

Your examples all seem to be US-based, and the thing I linked did say that the pro-cyclic effect has vanished in the US since 1995. Previously you had said you were interested if there was data for any countries other than the US where mortality and unemployment weren't correlated and there was - in Europe in the 2007 recession.

So I don't think US-specific counter-examples are necessarily relevant given the context that spun off this side-thread, which iirc was a comment from me about intending to move to a country where I'd heard that it was pretty hard to die of poverty.

The specific strength of wealth being tied to good outcomes no doubt varies on a country by country basis, but wealth is tied to outcomes on a worldwide level, not merely at the level of any specific nation. Even comparing country against country, richer countries almost invariably have more access to medical care and other niceties that provide a higher quality of life and a longer lifespan.

Pro-cyclic effects are probably the exception, rather than the norm.

Doesn't seem particularly relevant to this discussion, unless you're arguing that in general unemployment has fallen over the course of human history or some other long time-scale correlation.
From what I can tell, (in the US) it's fluctuated between 4% and 10% with occasional excursions (https://en.wikipedia.org/wiki/Unempl..._1892-2016.png). So 'more people are working and so are in a good economic position' can't explain the overall trend-line.

Economic prosperity in general has improved over that timeline. Unemployment is part of that, but not the entire story. Overall productivity gains also help, but as it seems deeply unlikely that Coronavirus is making us more productive, there seems to be little value in digging into that.

Unemployment is often used as a quick gauge of economic health. You can use other gauges, such as the stock market, but unemployment is probably better in most cases, since the stock market is affected by many things, and rapid changes appear to be more easily reversed. In addition, stock ownership tends to be a bit concentrated among the wealthy, while nearly everyone has jobs, or is affected by not having one. Job loss isn't the only negative outcome of economic troubles, but it is a fairly good indicator of negative outcomes while being a significant one itself.

At best, the conclusion one would take from that dropoff is 'employment status has very weak effects on mortality compared to other factors that don't have to do with employment status at all'. Which also sort of denies your argument about mass unemployment causing lots of deaths. So I don't think this graph helps your argument at all.

There are always other factors. Increased medical knowledge is generally helpful overall. If you discounted effects because stronger effects existed, you'd throw out nearly all of science. Why bother correcting cigarette smoking if obesity kills more rapidly? Because it still kills. You don't ignore every danger but the strongest one, even if it is necessary to do some balancing of risk factors for risk management in many cases.

After all, even if Coronavirus isn't as bad as initially feared and we choose to re-open, we should still try to mitigate as much risk as we can, for the same reason. It may not be the greatest threat, but it is still a threat.

If we want to look at measurements other than unemployment, we can look at other metrics of financial health instead, but as they are also universally negative, I'm not sure how much difference it makes. We can look at what percentage of stores are still functional enough to pay basic bills, such as rent, and right now, most of them are not(81% paid rent in march, only 47% in april. Even 81% is lower than last year by a good bit. Numbers aren't in for May yet, but they are probably worse). As a fairly obvious result of this, property owners are missing mortgages. Mall of America is two mortgage payments behind, for example. Homeowner mortgage delinquency rates have also nearly doubled. This is pretty bleak. Any financial measure we care to look at is awful, and a scale of awful that is far, far beyond usual recession metrics. Any way you look at it, wealth is getting burned pretty fast, and that's going to hurt a lot of people somehow. Sure, we can quibble over how who gets hurt and how much, but some significant amount of people are going to lose quite a lot.

The short summary is
1. Less wealth means a shorter average lifespan.
2. We're losing a ton of wealth.
3. Unless fixed, this is going to have a brutal effect on average lifespan.

If we want to look at the mechanisms of how, studies exist for that. You'll note that it often happens on a societal level, not an individual one. A richer society is more able to care for the poor among it. Even if you personally are not wealthy, living in a wealthy area is generally helpful, though obviously individual wealth is also nice. There is no ceiling or floor to this effect. This strongly indicates that no amount of redistribution can entirely mitigate the health risk, and it may not make any significant difference to average outcomes at all, as we'd largely be trading away the health of one population for that of another.

If you would prefer to examine mortality worldwide, a decent visualization is available at Our World in Data, and if you compare that against wealth, it would be very difficult to deny that wealthy countries enjoy longer lifespans than poorer ones.

More generally I'm also not convinced that it'll make that much of a difference, and I am sceptical of "pandemic theatre" in the same way I am of "security theatre" or for that matter "green theatre": i.e. that certain measures that impose inconvenience on people in the name of a greater cause are sometimes in fact principally cosmetic and intended largely if not entirely to reassure people that the issue is being taken seriously and addressed - while doing nothing to address the actual problem and even at times hindering the overall cause.

I have seen some folks "wearing" a mask in such a way as to technically comply with requirements, but in ways that do nearly nothing. Such as having it pulled down below their nose so it's "more comfortable". At that point, I agree that it's basically theater. If you're not bothering to do it at least vaguely correctly, there is probably little value to it. I think it's worth doing if you put in the effort to do it right, but performative measures do seem awfully pointless.

As an aside, in the US it appears you can fairly easily get a test at CVS's now. Friend got exposed to an unmasked person who tested positive, thus the research, but I guess it's good that tests are fairly available? Hopefully useful if you or someone you know needs a test.

[Aedilred]

I have seen some folks "wearing" a mask in such a way as to technically comply with requirements, but in ways that do nearly nothing. Such as having it pulled down below their nose so it's "more comfortable". At that point, I agree that it's basically theater. If you're not bothering to do it at least vaguely correctly, there is probably little value to it. I think it's worth doing if you put in the effort to do it right, but performative measures do seem awfully pointless.

As an aside, in the US it appears you can fairly easily get a test at CVS's now. Friend got exposed to an unmasked person who tested positive, thus the research, but I guess it's good that tests are fairly available? Hopefully useful if you or someone you know needs a test.

I'm not in the US, alas. Well, I say "alas" but all things considered I'm happier being here.

The testing situation here is... fraught. And somewhat political. What I really need at this point is a reliable antibody test, because if (as I suspect) I have had it and now have immunity to it (and I recognise that immunity in this instance does not necessarily follow automatically from infection) it would not only provide peace of mind but mean I can be more actively useful to my parents without fear of infecting them: they have (due to a special vulnerability on the part of my mum) been completely housebound for two months and will remain so until at least the end of June. And then as soon as the hopsital permits, my dad needs fairly significant surgery which will necessitate another couple of months of housebound-ness.

[NichG]

The specific strength of wealth being tied to good outcomes no doubt varies on a country by country basis, but wealth is tied to outcomes on a worldwide level, not merely at the level of any specific nation. Even comparing country against country, richer countries almost invariably have more access to medical care and other niceties that provide a higher quality of life and a longer lifespan.

Pro-cyclic effects are probably the exception, rather than the norm.

Economic prosperity in general has improved over that timeline. Unemployment is part of that, but not the entire story. Overall productivity gains also help, but as it seems deeply unlikely that Coronavirus is making us more productive, there seems to be little value in digging into that.

Um, this seems to be conflating all sorts of effects at different timescales. The claim isn't that Covid lockdowns will increase GDP. It's that even if over 100-year timescales both life expectancy and per-capita GDP have gone up (as have a lot of other things), the effects of unemployment surges have a stronger and more persistent positive effect on reducing sources of death, than they have negative effect in decreasing life expectancy via any wealth-mediated things.

If I were working as a cashier, I could probably increase my wealth a bit by going to work in a coal mine instead, but the positive effect - whatever it is - of that extra wealth on my life expectancy would likely be outweighed by the larger negative health impacts of working in a coal mine.

Since the data says there have been pro-cyclic effects in various countries in response to downwards fluctuations in economic activity, that's saying that whatever the long-term effects on any kind of wealth-related expansion in life-expectancy caused by those dips were on the net out-weighed by the risk reductions that resulted.

If the drop in economic activity from Covid results in a lower mortality rate overall, it's not going to be because it made us wealthier. It will be because whatever effect there is from wealth and whatever effects there are from e.g. dangers of participating in economic activity, hazards produced by economic activity, etc all combine together to create an aggregate effect which on the sum goes one way rather than another.

[Tyndmyr]

I'm not in the US, alas. Well, I say "alas" but all things considered I'm happier being here.

The testing situation here is... fraught. And somewhat political. What I really need at this point is a reliable antibody test, because if (as I suspect) I have had it and now have immunity to it (and I recognise that immunity in this instance does not necessarily follow automatically from infection) it would not only provide peace of mind but mean I can be more actively useful to my parents without fear of infecting them: they have (due to a special vulnerability on the part of my mum) been completely housebound for two months and will remain so until at least the end of June. And then as soon as the hopsital permits, my dad needs fairly significant surgery which will necessitate another couple of months of housebound-ness.

Sorry to hear that. If it's of any help, around me, pediatricians have been doing the antibody test(probably because kids are so frequently non-symptomatic). I have no idea if it's similar there, but maybe something to check out? Definitely good to know for sure if you can.

[Liquor Box]

Primarily reductions in tuberculosis , flu, and pneumonia as well as traffic deaths. Suicide increased but was a minor fraction in total. Doesn't look like starvation ranked: https://www.pnas.org/content/106/41/17290

I see the graph from Tyndmyr indicates a pretty steady reduction in mortality rates over the last century.

It suggests that the downward trend is such that even quite significant death events do not register on it. So the great depression may have caused a very large number of deaths, but that number was insufficient to interrupt the general downward trend. I note that the mortality rate does not even seem to pause for WW2 (which caused 75 million deaths - more than we'd expect to occur if every person on the planet contracted Covid19).

Internally to a country (a currency, really) money is an accounting device to record promises of future labor in exchange for resources and services delivered at the moment. So it's lubrication, but the goods and services being exchanged don't equate to it in any kind of absolute sense.

If e.g. a currency collapses then it devalues things you did in the past relative to things you will do in the future. So at some point the guy who does shed replacement might rather be paid in chickens than dollars. That's bad news for the country (because no faith in its currency means no international leverage, and taxation stops being an effective way to fund infrastructure and public services) but I don't think it'd stop the shed from getting replaced. It might stop the farmer from getting gas for their tractor or stop the shipping company from transporting the food though.

You appear to be talking about high inflation here. That is not the only effect of a depression - if it is an effect at all (sometimes inflation goes down because central banks seek to lower interest rates).

Hm... Yeah, this needs the payout for lying about your preferences to be lower than telling the truth. So I suppose that however you vote in the market it has to cost and pay out according to the same structure regardless of your vote. Anyhow, it starts to look more like a government in that direction.

I think we agree that it is governments who should b making decisions about covid. But I think we disagree about whether a heavy handed approach or a light touch approach presents itself as obvious.

About masks, there isn't a one-size-fits all solution. Wuhan's subway had a daily ridership of 3.6 million passengers per day in 2019, that's over 2/3 of New Zealand's population passing through the same crowded metal tubes every day. So I'd say the odds of you getting infected or infecting someone in Wuhan are much higher than almost anywhere in NZ (assuming all other conditions are the same, which they aren't, I know), and if a mask can help reduce that, it's probably a good idea to wear one (assuming you're doing it right, that's another issue altogether). But it's hard to prove how effective the masks are (or even if they are effective at all, I admit I could be wrong), since getting enough data to get conclusive results is probably not worth the effort.

I absolutely agree that different masks are different and every country's circumstances are a bit different, such that whether wearing masks is sensible may differ from country to country.

The only slight objection I have to your post is where you say the question of whether you are wearing the mark right is another issue altogether. I don't agree. Sure, if you are recommending masks, you should make every effort to educate people about how they work. But, despite those efforts there will be some people who just put it on, and don't follow the procedure. And (to the extent that improperly worn masks can increase risks) you have to factor that point in when deciding whether to recommend wearing them at all.

Literally any advice *can* be bad if sufficient numbers of people follow it sufficiently badly. Do you have any evidence that, in this case, they will?

People improperly using masks can be bad because it can increase the risk of infection. Washing hands is not bad because even if done improperly it will not increase the risk of infection.

Sometimes you still have to act on incomplete data.

There is at least data showing that mask use is highly effective when done correctly. Which, of course, is why medical professionals insist on using masks.

Everything against is supposition. If you're prioritizing supposition over data, you're on shaky ground. Sure, the data is not complete, and we don't have any hard data on how much masks would help, but that's not the same as proof they don't.

No, the data is mixed on whether generally available masks used correctly is at all effectively (let alone highly effective). I have already provided a link on this point.

So it is no more supposition that improperly used masks increase infection chance than it is supposition that a properly worn mask decrease infection rates. It is no more supposition that the false sense of security from mask wearing increases infection chance than that a properly worn mask decreases infection rates.

And the water droplet carrying the coronavirus are significantly larger than the virus itself. The point of wearing masks is to reduce risks from this transmission vector.

And the water droplet may contaminate the improperly worn mask increasing infection risk.

Some masks are also a great deal more effective than you list above. There's a pretty wide range depending on specific type of mask.

And the best masks available to the public do not offer protection in the same way as the PPE equipment used by doctors, and may offer none at all. Part of the reason for that is that the virus can be contracted through the eyes, so wearing masks alone (even if they were themselves infallible) may not offer protection.

If they are, in fact, usually ahead of the CDC when dispensing advice that holds up as correct, then yeah, that makes them more useful.

It isn't about authority, it's about utility. Saying the WHO doesn't recommend one way or another doesn't mean much if the WHO is always slow to suggestion action. It isn't proof against it as a valid measure, it could merely be another slow reaction.

Great. My home country was ahead of CDC in closing borders, so by your logic (it having been earlier to make a recommendation demonstrated to be correct with the benefit of hindsight) is therefore the stronger authority, and thus by your logic its recommendations not to use masks should be preferred to the CDC recommendation to use them.

To be clear, that is not my view. My view is that each organisation (WHO, CDC, and most other country's relevant authorities) each have a great deal of expertise. And the fact their views on recommending mask use vary demonstrates that there's non consensus and suggests that different conclusions can sensibly be drawn from the available (incomplete) facts and data.

Irrelevant. Large organizations have suggested mask wearing, so comparing the credibility of large organizations to that of individual doctors is a bit of a red herring. It isn't an obscure

Yes, large organisation have recmmendedmask use. Large organisations have recommended against mask use. Hence my original point that there's no consensus.

Many people have masks that are of high quality that they can use. I myself had masks with very good filters(and a bunch of spare filters) because it's typical safety gear when working with concrete. I'm probably not the only person who has shopped for a project at Home Depot and has some stuff sitting around.

Wearing eye protection is also helpful, and wearing both is probably better than either alone, but wearing only one is still superior to wearing none. Even if exposed, you will typically reduce the viral load you are exposed to. Exposure is not strictly binary, and reduced exposure still improves outcomes. In any case, this doesn't argue against wearing a mask. It argues for wearing a mask AND eye protection.

Maybe you are right. Maybe not. When the experts look at it they differ on whether your conclusions are correct.

[Rockphed]

You appear to be talking about high inflation here. That is not the only effect of a depression - if it is an effect at all (sometimes inflation goes down because central banks seek to lower interest rates).

My memory from freshman economics is that, in general, interest rates and inflation are negatively correlated. So when a central bank is trying to reduce inflation (because high inflation is associated with economic shrinking), they raise interest rates. When they want to let the economy expand a little faster, they reduce interest rates. Whether they are ever right in doing either is one of those questions that some people could probably discuss without politics, but I have no faith in my ability to do so.

[NichG]

I see the graph from Tyndmyr indicates a pretty steady reduction in mortality rates over the last century.

It suggests that the downward trend is such that even quite significant death events do not register on it. So the great depression may have caused a very large number of deaths, but that number was insufficient to interrupt the general downward trend. I note that the mortality rate does not even seem to pause for WW2 (which caused 75 million deaths - more than we'd expect to occur if every person on the planet contracted Covid19).

That's the point about conflating effects at different timescales. Analyzing causality generally means making an intervention and seeing if the system changes relative to where it would have gone if you didn't intervene. Since we generally can't have two Earth histories to do this cleanly, instead you can try to look at naturally occurring fluctuations and responses to those fluctuations within a small time window. That's part of the logic behind using 'excess deaths' when trying to analyze un-reported Covid deaths for example - you can compare relative to previous years to get an idea of what Covid is doing as an intervention.

If we did excess death analysis compared to the death rate in 1940, then this overall downward trend would mess it up, but on a year-to-year basis the trend can be subtracted.

You appear to be talking about high inflation here. That is not the only effect of a depression - if it is an effect at all (sometimes inflation goes down because central banks seek to lower interest rates).

I think we agree that it is governments who should b making decisions about covid. But I think we disagree about whether a heavy handed approach or a light touch approach presents itself as obvious.

Not how I meant that 'this is looking like a government'. What I meant was, in trying to design a market that extract's people's latent value they assign to the outcome of having Covid eradicated, the issue is that people can refuse to play under the expectation that others will pay and they'll still benefit from the public good. Normally you'd fix that with an iterated game - a reputation system or something like that - but this is a one-off so we can't easily use that mechanism. Governments can enforce that kind of extended consequence for defecting because they can restrict access to public goods if a participant refuses to support them - e.g. they can arrest you for not paying taxes. So the point there wasn't 'governments should be the ones making decisions', it was 'maybe you can't extract this value with a voluntary participation one-off market that has no ability to project consequences for non-participation'.

[Brother Oni]

I am also wholly in favour of more aggressive testing.

I'm going to go out on a limb and say you haven't experienced the standard nasal swab. They essentially take a cotton bud, insert it about 4 inches up your nose and twist. They then repeat this for the other nostril.
Demonstration by a NHS doctor.

It is deeply, deeply uncomfortable and I would foresee issues with compliance (especially repeat compliance), if more aggressive testing with the current sampling method was implemented.

If the serology (antibody) tests became more accurate then it would be more acceptable, but until then, that's the most accurate method we have.

I think a couple of people here have had the swab - how many of them would be prepared to be sampled on a regular fortnightly basis?

[Strigon]

corona virus it's fake

Oh.
Well, there you have it, folks; we can all come out now!
Discussion's over; the virus is fake.

[Rockphed]

That's part of the logic behind using 'excess deaths' when trying to analyze un-reported Covid deaths for example - you can compare relative to previous years to get an idea of what Covid is doing as an intervention.

One question is how many excess deaths are suicides caused by the lockdown? I have seen reports of spikes in suicide in some places. Also, I have had a tooth ache for the last 3 months. I was going to go see a dentist about it when the lockdown happened. I have a friend on facebook whose daughter needed a surgery, but was denied because of the wuhan scare. How many people have died because they didn't receive the right care due to telemedicine? If we hadn't locked down, we could state how many deaths we missed by looking at excess deaths, but the lockdown changes things to the point that we don't have a control. And without the threat of a disease I don't think anyone could have enforced this lockdown, so we are unlikely to ever get a control.

[NotASpiderSwarm]

I can't find the analysis right now, but I saw one that showed excess deaths in places that locked down and had few or no Covid cases were basically flat compared to previous years, while excess deaths in places that had locked down and had major Covid outbreaks had skyrocketed.

The thing is, you're going to see more domestic abuse deaths and fewer muggings. More people not going to the doctor for chest pains that turn out to be a heart attack, but fewer cases of the flu. Car wreck data is wild right now, far fewer people are driving, so there's more accidents at 90+ mph and fewer at 30 mph. Excess deaths will take all of that in, so it's definitely not a clear "10,000 more people died this year than 2019, those are all Covid". But it's a simple, clean, number that isn't going to be wrong due to increased testing or what MEs are told to put on death certificates, so for the basic analysis that most people care about, it's probably the closest to accurate we're going to get.

[Tyndmyr]

I see the graph from Tyndmyr indicates a pretty steady reduction in mortality rates over the last century.

It suggests that the downward trend is such that even quite significant death events do not register on it. So the great depression may have caused a very large number of deaths, but that number was insufficient to interrupt the general downward trend. I note that the mortality rate does not even seem to pause for WW2 (which caused 75 million deaths - more than we'd expect to occur if every person on the planet contracted Covid19).

Essentially, yeah. Obviously both WW2 and the Great Depression are undesirable things. And, if we could have somehow avoided them without creating other problems, that'd be the way to go. Presumably, humanity has not perfectly optimized progress at all points in its history, and we could have done better in some regards. In fact, you could argue that most eras have *some* sort of problem going on.

I'd imagine that Covid will also not interrupt this steady progress, nor will the reaction to it. However, like the problems listed above, they're certainly still to be minimized.

I think we agree that it is governments who should b making decisions about covid. But I think we disagree about whether a heavy handed approach or a light touch approach presents itself as obvious.

I don't particularly care why a person does a given action. If they wear a mask because the gov told them to or because of a health care advisory or some other reason, it's still just as useful no matter who is doing the telling. Mostly, I want the information available to be as accurate as possible, but beyond that...the rest may not matter much. As noted, people start and stop things like social isolation largely independently from what they are told to do, based on their perception of risk.

People improperly using masks can be bad because it can increase the risk of infection. Washing hands is not bad because even if done improperly it will not increase the risk of infection.

The latter is untrue. Overly frequent washing or using too hot of water for handwashing can cause skin damage and create avenues for infection. These are not particularly commonly found, and advising handwashing is still reasonable, but there is surprisingly little difference here.

No, the data is mixed on whether generally available masks used correctly is at all effectively (let alone highly effective). I have already provided a link on this point.

You are, again, putting "generally available mask" instead of what I said. You're basically arguing against the lowest possible masks, rather than taking the entire range into account. I acknowledge that a homemade mask from tshirt materials is probably not great, but you can go buy a KN-95 off Amazon, and a face shield to go with it if you are so inclined. They are generally available.

And the water droplet may contaminate the improperly worn mask increasing infection risk.

For the described use case of a mask on a sick person to prevent them from sickening others...the wearer is already sick. It's not a threat in that fashion.

And the best masks available to the public do not offer protection in the same way as the PPE equipment used by doctors, and may offer none at all. Part of the reason for that is that the virus can be contracted through the eyes, so wearing masks alone (even if they were themselves infallible) may not offer protection.

This has already been covered previously.

Great. My home country was ahead of CDC in closing borders, so by your logic (it having been earlier to make a recommendation demonstrated to be correct with the benefit of hindsight) is therefore the stronger authority, and thus by your logic its recommendations not to use masks should be preferred to the CDC recommendation to use them.

New Zealand, yes? Sorry, you were shortly after the US in closing borders to China, not before.

And stronger is not quite correct. More useful is more accurate. In a disease scenario, earlier information has value, so faster authorities are more useful than slower ones, provided they are both correct.

To be clear, that is not my view. My view is that each organisation (WHO, CDC, and most other country's relevant authorities) each have a great deal of expertise. And the fact their views on recommending mask use vary demonstrates that there's non consensus and suggests that different conclusions can sensibly be drawn from the available (incomplete) facts and data.

Certainly some disagreement still remains. But it's likely that of "wear masks" and the opposite, one will eventually be found to be correct, and folks will generally consolidate around that viewpoint. Of the two viewpoints, momentum appears to be towards "wear masks if available."

One question is how many excess deaths are suicides caused by the lockdown? I have seen reports of spikes in suicide in some places. Also, I have had a tooth ache for the last 3 months. I was going to go see a dentist about it when the lockdown happened. I have a friend on facebook whose daughter needed a surgery, but was denied because of the wuhan scare. How many people have died because they didn't receive the right care due to telemedicine? If we hadn't locked down, we could state how many deaths we missed by looking at excess deaths, but the lockdown changes things to the point that we don't have a control. And without the threat of a disease I don't think anyone could have enforced this lockdown, so we are unlikely to ever get a control.

We can spitball some estimates, but it's certainly some. As noted, it's very difficult to do this precisely, as we don't have a spare world to test against. We can compare against last year, but every year has many different things going on, and on average, you see some yearly improvement anyways. We can make estimates for all of these factors, but error and bias does creep in. Even the current numbers for covid deaths are subject to challenge in some cases, and existing information is not always ideal.

The exact tradeoff point will probably vary by location. If you're in an area with very high covid infection/death rates, more precautions will be warranted than if in a mostly uninfected area. The sane choice for NYC is probably a far higher disease mitigation priority than it would be for any given rural setting.

[NotASpiderSwarm]

The exact tradeoff point will probably vary by location. If you're in an area with very high covid infection/death rates, more precautions will be warranted than if in a mostly uninfected area. The sane choice for NYC is probably a far higher disease mitigation priority than it would be for any given rural setting.

No, if you're in an area with minimal COVID exposure, the goal is to keep that exposure minimal. If you're some isolated town in rural TX or Kansas, then very strict procedures are the way to keep your death count at 0. The fact that it hasn't hit you yet is an opportunity to keep it out, not an excuse to ignore the possibility.

[Forum Explorer]

One question is how many excess deaths are suicides caused by the lockdown? I have seen reports of spikes in suicide in some places. Also, I have had a tooth ache for the last 3 months. I was going to go see a dentist about it when the lockdown happened. I have a friend on facebook whose daughter needed a surgery, but was denied because of the wuhan scare. How many people have died because they didn't receive the right care due to telemedicine? If we hadn't locked down, we could state how many deaths we missed by looking at excess deaths, but the lockdown changes things to the point that we don't have a control. And without the threat of a disease I don't think anyone could have enforced this lockdown, so we are unlikely to ever get a control.

It's complicated, like really complicated. I mean, we can point to more people dying or suffering from domestic abuse but there's also a decrease in muggings and gang violence. Or even just things like less car accidents. It's really tempting to call it a wash but that's just because it's too complex to give an accurate answer. At least in the current conditions.

Medical treatments is also very hard. Now you aren't (likely) going to die from a tooth, no matter how bad it gets, so you are low priority. But for more critical cases, well I know that at least some of those operations are still being done. At least in Canada, since it might vary from region to region. But the big thing is that they are able to do those operations in Canada because Covid-19 didn't overwhelm our medical system. Which may very well be because we went into lockdown. In places where the medical system was overwhelmed (like Italy), people were denied life saving treatments simply because there wasn't a doctor available to perform the treatment. And in that case, did the person die because of Covid-19? Because odds are that it was reported otherwise.

But it's very much a triage and use your best judgement situation. If you are deemed a critical case, they want to delay your operation. Which can lead to stuff like a cancer patient who was already on a waiting list for some time being delayed further, and not really knowing when they'll get an opportunity to receive treatment. The doctors made the call that she'll likely survive until they are ready for her, but they could be wrong.

[Bartmanhomer]

Please tell me that the scientists made a vaccine to get rid of the virus yet.

[Liquor Box]

That's the point about conflating effects at different timescales. Analyzing causality generally means making an intervention and seeing if the system changes relative to where it would have gone if you didn't intervene. Since we generally can't have two Earth histories to do this cleanly, instead you can try to look at naturally occurring fluctuations and responses to those fluctuations within a small time window. That's part of the logic behind using 'excess deaths' when trying to analyze un-reported Covid deaths for example - you can compare relative to previous years to get an idea of what Covid is doing as an intervention.

If we did excess death analysis compared to the death rate in 1940, then this overall downward trend would mess it up, but on a year-to-year basis the trend can be subtracted.

Sure, but if we try to relate it back to the discussion - this was raised in the context of whether economic hardship in the depression led to significant additional deaths. For all the reasons above, comparing the depression years to the preceding or following years isn't useful in either proving or disproving this.

Not how I meant that 'this is looking like a government'. What I meant was, in trying to design a market that extract's people's latent value they assign to the outcome of having Covid eradicated, the issue is that people can refuse to play under the expectation that others will pay and they'll still benefit from the public good. Normally you'd fix that with an iterated game - a reputation system or something like that - but this is a one-off so we can't easily use that mechanism. Governments can enforce that kind of extended consequence for defecting because they can restrict access to public goods if a participant refuses to support them - e.g. they can arrest you for not paying taxes. So the point there wasn't 'governments should be the ones making decisions', it was 'maybe you can't extract this value with a voluntary participation one-off market that has no ability to project consequences for non-participation'.

I think I follow you. But so we are on the same page, what we are talking about here is purely theoretical isn't it? You agree that these models you are proposing are not something that could have practically been used in a covid-19 response?

[NichG]

Sure, but if we try to relate it back to the discussion - this was raised in the context of whether economic hardship in the depression led to significant additional deaths. For all the reasons above, comparing the depression years to the preceding or following years isn't useful in either proving or disproving this.

Um, no. The entire point of what I just posted is that you can do this comparison, by restricting the timescale of the analysis to a narrow window in which the overall trend of the century is small and can be subtracted out. The fluctuation timescale and 'general march of progress' timescales are different, can be separated, and therefore you can do the analysis just fine.

The fluctuations in year-to-year changes around the trend-line in the depression part of the curve Tyndmyr posted are larger than the average movement over that interval.

I think I follow you. But so we are on the same page, what we are talking about here is purely theoretical isn't it? You agree that these models you are proposing are not something that could have practically been used in a covid-19 response?

I'm not sure why this point keeps coming up. The whole reason to generate these models, work out their properties, etc, is that so in the future when we have a disaster, pandemic, whatever, we can just go and do these things. Rather than always saying 'well its all well and good to theorize that we could do better, but we had to make the best decision we could in a limited time-span'. The time to figure this stuff out was years ago, but we didn't (or if we did, we didn't end up implementing the infrastructure needed to make these things a reality). Someone could have done so - probably even did so, but I can't talk to that. But, if so, it wasn't successfully instituted as preparations at that time.

And if the primary cause of problems is to do with the organization of human behavior and not the dynamics of infectious disease, then that suggests a certain research focus or direction that may actually not have been emphasized as part of the broad strategy of pandemic preparation. Which would make it a new idea, and by virtue of that still worthy of being developed and, if practicable, implemented. The question of 'how to design a market such that people's latent valuation of socially shared outcomes is fully engaged' for example seems important whether we're talking about pandemics or getting people to pay for public infrastructure alike - and from talking to people who follow such things, its evidently a fairly active frontier of research (the relevant search phrase seems to be 'quadratic funding' or 'quadratic voting', though that's focused more on designing market-based decision algorithms to be more equitable than to solve Prisoner's Dilemma).

So, given that we can't change the past, the time to figure this stuff out is right now. If not for Covid, than for any future pandemics that might occur, or even other types of tradgedies of the commons.

And as for Covid, if it behaves like Spanish Flu then we'll have a second wave that's worse than the first. When we enter that period, we could do more of the same as we just did, or we could do something better. If it turns out we don't do better in the future, then it absolutely is the fault of our present selves, and 'we did the best we could' won't be a valid excuse.

[factotum]

Please tell me that the scientists made a vaccine to get rid of the virus yet.

They haven't, and aren't likely to for a while. A vaccine is currently being tested by a Cambridge lab, and there are no doubt similar efforts going on around the world, but it'll be months before any such thing is ready for roll-out--you have to be sure that firstly, the vaccine actually works to prevent Covid-19, and secondly, that it doesn't cause worse issues down the line. That requires extensive testing.

[Brother Oni]

They haven't, and aren't likely to for a while. A vaccine is currently being tested by a Cambridge lab, and there are no doubt similar efforts going on around the world, but it'll be months before any such thing is ready for roll-out--you have to be sure that firstly, the vaccine actually works to prevent Covid-19, and secondly, that it doesn't cause worse issues down the line. That requires extensive testing.

To clarify factotum's point, technically they have made a vaccine, they just haven't proven it's safe, efficacious and fit for purpose (in that order) yet, so it's not been released for general use.

[Lvl 2 Expert]

To clarify factotum's point, technically they have made a vaccine, they just haven't proven it's safe, efficacious and fit for purpose (in that order) yet, so it's not been released for general use.

So they haven't actually made a vaccine, they've made a thing that could turn out to be a vaccine, but for now they're not actually sure what it is yet.

[factotum]

So they haven't actually made a vaccine, they've made a thing that could turn out to be a vaccine, but for now they're not actually sure what it is yet.

There's actually an Indian pharmaceutical company which is already working on replicating the Cambridge vaccine on the assumption it'll turn out OK and to be in a position to have millions of them ready for September--presumably if the vaccine turns out to be a bust they'll just have to throw all that away and start again from scratch.

[Liquor Box]

Um, no. The entire point of what I just posted is that you can do this comparison, by restricting the timescale of the analysis to a narrow window in which the overall trend of the century is small and can be subtracted out. The fluctuation timescale and 'general march of progress' timescales are different, can be separated, and therefore you can do the analysis just fine.

The fluctuations in year-to-year changes around the trend-line in the depression part of the curve Tyndmyr posted are larger than the average movement over that interval.

Sorry, I'm not following this. Am I understanding you right that you are saying that if we look at only the few years prior and the few years after the general downward trend will be so slight that any vairation arising from a depression (or war, or pandemic) should show up?

I'm not sure why this point keeps coming up. The whole reason to generate these models, work out their properties, etc, is that so in the future when we have a disaster, pandemic, whatever, we can just go and do these things. Rather than always saying 'well its all well and good to theorize that we could do better, but we had to make the best decision we could in a limited time-span'. The time to figure this stuff out was years ago, but we didn't (or if we did, we didn't end up implementing the infrastructure needed to make these things a reality). Someone could have done so - probably even did so, but I can't talk to that. But, if so, it wasn't successfully instituted as preparations at that time.

It comes up because my opinion is that these sorts of models may (but not so far) be theoretically possible, but that are not something that I think would realistically work in practice. It also comes up because I was under the impression that the two of us were on the same page that we didn't think of these as realistic options.

For an example let's take your idea about creating a company where people would buy stock or sell short.

I addressed that on its merits as a theory - hence my objection that people would be incentivised to lie.

But that problem fades in comparison with the practical problems. A significant proportion of the population will not be able to understand how it works or its implications, even in an ideal case of this company being set up years in advance with education. Persons who have no money would struggle to buy shares or sell short. Some people may indeed have lots of money and would want to invest int he company one way or the other, but are unable to access it because it's unsufficiently liquid. Savvy traders will take advantage of the mechanism to invest one way or the other by buying or selling when the price is favourable. Some people will wait to the last minute to see if they need to invest, and not do so if the stock price is sufficient to support their preferred situation, disadvantaging early trend setters.

And if the primary cause of problems is to do with the organization of human behavior and not the dynamics of infectious disease, then that suggests a certain research focus or direction that may actually not have been emphasized as part of the broad strategy of pandemic preparation. Which would make it a new idea, and by virtue of that still worthy of being developed and, if practicable, implemented. The question of 'how to design a market such that people's latent valuation of socially shared outcomes is fully engaged' for example seems important whether we're talking about pandemics or getting people to pay for public infrastructure alike - and from talking to people who follow such things, its evidently a fairly active frontier of research (the relevant search phrase seems to be 'quadratic funding' or 'quadratic voting', though that's focused more on designing market-based decision algorithms to be more equitable than to solve Prisoner's Dilemma).

How is the bigger problem about the organisation of human behaviour? Do you mean because the primary problem is (or may be) the economic downturn rather than the health effects? Or is it your view that cooperation is best?

if it's cooperation, I don;t think that is the bigger problem. While I agree that co-operation may have been a more optimal solution than non-cooperation, I don't think the difference between the two is great enough that you can label it the primary cause of problems.

So, given that we can't change the past, the time to figure this stuff out is right now. If not for Covid, than for any future pandemics that might occur, or even other types of tradgedies of the commons.

And as for Covid, if it behaves like Spanish Flu then we'll have a second wave that's worse than the first. When we enter that period, we could do more of the same as we just did, or we could do something better. If it turns out we don't do better in the future, then it absolutely is the fault of our present selves, and 'we did the best we could' won't be a valid excuse.

Happy to talk about what we should do to prepare.

In terms of decision making, my view is that allowing governments to make the decision is better than spreading the decision across all persons. That is because this is a complicated question of balancing competing concerns and factors. When decision making is by a small group of intelligent people it is reasonable to expect that they will be across all relevant information, will understand it, and make a rational decision on that basis. Where it is individuals, I think it is reasonable to expect that many (perhaps most) people wont be across all information (or even have access to it), understand all implications, and therefore be able to make a rational decision. That is not to say that the smaller group will always make the better decision, only that they wioll have the better chance of making a good decision.

[NichG]

Sorry, I'm not following this. Am I understanding you right that you are saying that if we look at only the few years prior and the few years after the general downward trend will be so slight that any vairation arising from a depression (or war, or pandemic) should show up?

Yes, at zeroth order. Slightly better would be to subtract a moving average, so even if there is a general downward trend as long as that trend is relatively steady you'll subtract it out.

It comes up because my opinion is that these sorts of models may (but not so far) be theoretically possible, but that are not something that I think would realistically work in practice. It also comes up because I was under the impression that the two of us were on the same page that we didn't think of these as realistic options.

For an example let's take your idea about creating a company where people would buy stock or sell short.

I addressed that on its merits as a theory - hence my objection that people would be incentivised to lie.

But that problem fades in comparison with the practical problems. A significant proportion of the population will not be able to understand how it works or its implications, even in an ideal case of this company being set up years in advance with education. Persons who have no money would struggle to buy shares or sell short. Some people may indeed have lots of money and would want to invest int he company one way or the other, but are unable to access it because it's unsufficiently liquid. Savvy traders will take advantage of the mechanism to invest one way or the other by buying or selling when the price is favourable. Some people will wait to the last minute to see if they need to invest, and not do so if the stock price is sufficient to support their preferred situation, disadvantaging early trend setters.

Well, if you notice in my example, the market design was asymmetric so you don't need money to sell short - it's just one kind of vote. Similarly, buying shares doesn't get you returns, it gets you outcomes. So it's a bit different than something like the stock market - it's more of a cross between a voting system and an insurance pool.

I was also envisioning the participants in this kind of thing being entire countries, not private individuals. But from a design perspective you can use a generic term like 'agents' to mean that it could be either (or companies, blocs, etc).

How is the bigger problem about the organisation of human behaviour? Do you mean because the primary problem is (or may be) the economic downturn rather than the health effects? Or is it your view that cooperation is best?

if it's cooperation, I don;t think that is the bigger problem. While I agree that co-operation may have been a more optimal solution than non-cooperation, I don't think the difference between the two is great enough that you can label it the primary cause of problems.

If we agree that there's a Prisoner's Dilemma problem here, then even if we have things which could be better responses, the issue is getting people to actually do them - e.g. not defecting for a short-term gain. A lot of this discussion has taken the form of 'that's nice, but you could never get people to do it' - thus, a human problem and not a biology problem.