On Lockdown

This area is for the discussion of what's new, what's on your mind, and general photographic topics. A place to meet, make comments on this site, and get the latest community news.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Lou Jost wrote:Ray, we do know that the measures helped. There is plenty of data about that.
Really? I'd like to see some. I think the only data is that the sum total of what has been done so far is "better" than the worst case models, but I would challenge you by stating there is absolutely zero scientific evidence that the measures helped.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

ray_parkhurst wrote:That 0.2267% is probably also suspect (as is anything from the NYT)
Well, you might have followed my reference, and done a couple of simple calculations. The reported number of deaths is 19,141. It's hard to fudge that by very much. Wikipedia lists the population of NYC as 8.4 million. Do the division: 19,141 divided by 8.4 million = 0.00228 = 0.228% . I assume that difference in the third (or is it fifth?) decimal place is due to NYT using a slightly different number for population.

Note that this number is simply an observed death rate for the entire population, infected and uninfected combined. Whatever the infection rate is, divide by that to get the infection fatality rate. Working from the latest estimate that New York City has an infection rate of 21.2% (based on antibody testing of 3000 people who were out-and-about but not working), you'd get an infection fatality rate of about 1.26%.

--Rik

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

ray_parkhurst wrote:I'd like to see some.
Much of the data on efficacy of mitigation measures is historical, dating back to the 1918 flu pandemic. As always, Bayes' Theorem shows with mathematical certainty that none of that data can possibly have any effect on somebody who 100% believes otherwise.

--Rik

kaleun96
Posts: 270
Joined: Sat Oct 20, 2018 3:47 pm
Location: Stockholm, Sweden

Post by kaleun96 »

ray_parkhurst wrote:
Lou Jost wrote:We will also have empirical data on that, from Brazil and Sweden.
I'm not sure how much mitigation there was in either place. Weren't both places essentially business as usual? In any event, a decent model would be able to calculate the historical results in both places, though impact due to variations in the secondary factors, and quality of the data, is still in question.
I don't think Sweden makes a good control case study. The three main factors for why I think a no-lockdown approach hasn't blown up in their face as much as you might expect:
1. Something like 50% of Swedes live alone.
2. While there was no lockdown, many Swedish companies, including the one I work for, sent everyone home in early-mid March, before a lot of countries had initiated lockdown. I've been working from home since March 12 if I recall correctly. I think they estimated in Stockholm 80% of companies had sent their employees to WFH.
3. Swedes generally keep their distance from others anyway and are pretty health conscious and hygienic.

Restaurants and bars etc had also blocked off every second seat/table and added floor markers to separate queues.

While I'm not sure Sweden's approach was the correct one, I think it's difficult to consider them a control due to the reasons above.
- Cam

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

rjlittlefield wrote:
ray_parkhurst wrote:That 0.2267% is probably also suspect (as is anything from the NYT)
Well, you might have followed my reference, and done a couple of simple calculations. The reported number of deaths is 19,141. It's hard to fudge that by very much. Wikipedia lists the population of NYC as 8.4 million. Do the division: 19,141 divided by 8.4 million = 0.00228 = 0.228% . I assume that difference in the third (or is it fifth?) decimal place is due to NYT using a slightly different number for population.

Note that this number is simply an observed death rate for the entire population, infected and uninfected combined. Whatever the infection rate is, divide by that to get the infection fatality rate. Working from the latest estimate that New York City has an infection rate of 21.2% (based on antibody testing of 3000 people who were out-and-about but not working), you'd get an infection fatality rate of about 1.26%.
Yes, the calculation is a simple one, but simple is not always correct. It all comes down to the raw data. The data from China may have greatly under-estimated the number of deaths. And regarding NY and possibly other parts of the US and in the EU, I've read that there may be a significant over-counting of deaths, attributing many unrelated deaths to COVID-19, though by all accounts the overall death rate has indeed risen, presumably due to the virus. It all comes down to the accuracy of the data, and the completeness of the model.
rjlittlefield wrote:
ray_parkhurst wrote:I'd like to see some.
Much of the data on efficacy of mitigation measures is historical, dating back to the 1918 flu pandemic. As always, Bayes' Theorem shows with mathematical certainty that none of that data can possibly have any effect on somebody who 100% believes otherwise.
You misunderstand me Rik. I believe mitigation efforts are likely useful, and I am aware of at least some the historical records. Masks and social distancing almost certainly lower the transmission rate, though eventually I expect most everyone will indeed be exposed to the virus and will either develop COVID-19 or not. I was (am) a strong believer in the use of social distancing to slow this process down so that the healthcare systems would not be overwhelmed, but will the damage done to the world by the draconian shutdown (and burgeoning surveillance apparatus) be worse?

Lou Jost
Posts: 5942
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Ray, the evidence is pretty clear that severe lockdowns caused a sharp downturn in new cases. The timing of the downturns seems closely related to the start dates of the lockdowns.

The death rates due to Covid-19 in most countries, including my own and probably including the US, are UNDER-reported, not over-reported. In most countries, including Ecuador, the reported deaths are those confirmed by tests. The scarcity of tests and the fact that not all victims die in hospitals result in under-reporting. Analysis of "excess deaths" in relation to historic death rates supports the conclusion of under-reporting in most parts of the world, though this method includes deaths due to people dying of other causes because the hospitals are full of virus patients.

I can't comment on the balance of lives versus dollars.

Lou Jost
Posts: 5942
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

While I'm not sure Sweden's approach was the correct one, I think it's difficult to consider them a control due to the reasons above.
Point taken, it is good to have a firsthand source in Sweden! I hope it works out for your country.

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

kaleun96 wrote: I don't think Sweden makes a good control case study. The three main factors for why I think a no-lockdown approach hasn't blown up in their face as much as you might expect:
1. Something like 50% of Swedes live alone.
2. While there was no lockdown, many Swedish companies, including the one I work for, sent everyone home in early-mid March, before a lot of countries had initiated lockdown. I've been working from home since March 12 if I recall correctly. I think they estimated in Stockholm 80% of companies had sent their employees to WFH.
3. Swedes generally keep their distance from others anyway and are pretty health conscious and hygienic.

Restaurants and bars etc had also blocked off every second seat/table and added floor markers to separate queues.

While I'm not sure Sweden's approach was the correct one, I think it's difficult to consider them a control due to the reasons above.
It sounds like some mitigation happened, in a smart way, and that the culture is more attuned to warding-off this sort of problem. I wish the US had been so lucky.

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Lou Jost wrote:Ray, the evidence is pretty clear that severe lockdowns caused a sharp downturn in new cases. The timing of the downturns seems closely related to the start dates of the lockdowns.

The death rates due to Covid-19 in most countries, including my own and probably including the US, are UNDER-reported, not over-reported. In most countries, including Ecuador, the reported deaths are those confirmed by tests. The scarcity of tests and the fact that not all victims die in hospitals result in under-reporting. Analysis of "excess deaths" in relation to historic death rates supports the conclusion of under-reporting in most parts of the world, though this method includes deaths due to people dying of other causes because the hospitals are full of virus patients.

I can't comment on the balance of lives versus dollars.
Downturns? None of the data I have seen shows downturns, only exponential increases.

As I said to Rik above, I do believe the social distancing helped, but there is no data to show one way or the other, let alone any scientific evidence.

I am not so sure about under-reporting in the US. Doctors here have been instructed (according to multiple accounts) to assign COVID-19 as cause of death even in marginal or questionable circumstances. The only under-counting I've heard about is due to overwhelmed nursing homes and such not reporting for whatever reason.

Lou Jost
Posts: 5942
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Downturns? None of the data I have seen shows downturns, only exponential increases.
We are apparently living in alternate universes. Nearly every country that did lockdowns has seen sharp downturns.

See:
https://aatishb.com/covidtrends/

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Lou Jost wrote:
Downturns? None of the data I have seen shows downturns, only exponential increases.
We are apparently living in alternate universes. Nearly every country that did lockdowns has seen sharp downturns.

See:
https://aatishb.com/covidtrends/
Apparently we're just not talking the same language, Lou, specifically in the definition of the term "downturn" and its timing. Are you saying that in the chart you showed, that the cases increased exponentially until the countries went into "shutdown" of various extremes, and the sharp downturns (reduction in the exponential increases) are due to these shutdowns? I can certainly tell you that in the US, this is not the case. I've been on shutdown since mid-March, when there were very few cases, and since then the cases have increased exponentially, only very recently going linear (even slightly less than).

Or are you saying that because the countries did shutdowns, that they are now seeing downturns long after the shutdowns were put in play? Any pandemic will eventually show downturns once it has "run its course".

What the shutdowns (due to social distancing and thus reduced transmission) more likely did was to keep some of the folks from getting infected early, the so-called "flattening the curve" effect, but there is no evidence that I can see of any sort of "downturn" associated with the shutdowns.

Lou Jost
Posts: 5942
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

In your earlier post, you said:
Downturns? None of the data I have seen shows downturns, only exponential increases.
I think the chart I linked to shows very sharp, very sudden downturns except in countries that did not do strict lockdowns (Brazil for example). Your statement in that earlier post contradicts the facts.
Are you saying that in the chart you showed, that the cases increased exponentially until the countries went into "shutdown" of various extremes, and the sharp downturns (reduction in the exponential increases) are due to these shutdowns?
Yes.
Or are you saying that because the countries did shutdowns, that they are now seeing downturns long after the shutdowns were put in play?
Yes to that too. A lockdown can't stop the spread instantly. As you mentioned earlier, there will initially be some in-family spread and particularly some internal spread in nursing homes, and no lockdown is 100%.
Any pandemic will eventually show downturns once it has "run its course".


None of the countries that experienced a downturn were anywhere near reaching "herd immunity", and there have been resurgences in some countries that had come out of lockdown. All the evidence shows that the disease has not "run its course" and strongly suggest that the lockdowns are the causal factor in lowering the infection rate.

The US is unusual due to the patchwork nature of the lockdowns and the bizarrely denialist national leadership.

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Lou Jost wrote:In your earlier post, you said:
Downturns? None of the data I have seen shows downturns, only exponential increases.
I think the chart I linked to shows very sharp, very sudden downturns except in countries that did not do strict lockdowns (Brazil for example). Your statement in that earlier post contradicts the facts.
Are you saying that in the chart you showed, that the cases increased exponentially until the countries went into "shutdown" of various extremes, and the sharp downturns (reduction in the exponential increases) are due to these shutdowns?
Yes.
Or are you saying that because the countries did shutdowns, that they are now seeing downturns long after the shutdowns were put in play?
Yes to that too. A lockdown can't stop the spread instantly. As you mentioned earlier, there will initially be some in-family spread and particularly some internal spread in nursing homes, and no lockdown is 100%.
Any pandemic will eventually show downturns once it has "run its course".


None of the countries that experienced a downturn were anywhere near reaching "herd immunity", and there have been resurgences in some countries that had come out of lockdown. All the evidence shows that the disease has not "run its course" and strongly suggest that the lockdowns are the causal factor in lowering the infection rate.
This last statement is the only thing I agree with. My statement was incorrect from timing perspective, since the downturn is just the beginning of the virus "running its course", not the end.

Let's talk about timing. The downturns in the data are not due to whether a country went into lockdown, but due to the virus starting to run out of hosts to infect. This would happen whether or not a shutdown was initiated. The "flattening of the curve" caused by the shutdowns and social distancing, by definition, should result in more gradual downturns after the virus finds fewer and fewer hosts, not sharper, so the sharp and sudden downturns must be due to some other factor. What I interpreted you as saying (and you confirmed) was that the shutdowns were associated in time with downturns, and this is clearly incorrect. I think what you're trying to say is that countries with shutdowns may have an earlier downturn than those without, though this is also entirely inconsistent with the "flatten the curve" concept, so maybe you're trying to say something different?

Lou Jost
Posts: 5942
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

The downturns in the data are not due to whether a country went into lockdown, but due to the virus starting to run out of hosts to infect.
I don't think this is true. Take China. Most of the country was never exposed to the disease. That is true in most countries. Do you have some examples or candidate countries where you think herd immunity is actually stopping the disease? Certainly that is not true in the US.
The "flattening of the curve" caused by the shutdowns and social distancing, by definition, should result in more gradual downturns after the virus finds fewer and fewer hosts, not sharper, so the sharp and sudden downturns must be due to some other factor.
The graph I linked to shows the sudden stopping of exponential growth. That was to address your earlier claim that all the data showed continuing exponential growth. we may differ on what "sudden" and "gradual" means. But I would claim that in most cases there was a sudden change from exponential growth, and this does not seem to be due to lack of hosts.
What I interpreted you as saying (and you confirmed) was that the shutdowns were associated in time with downturns, and this is clearly incorrect.
Your interpetation of what I said is correct. But maybe you thought I meant "simultaneous with" rather than "associated with". Of course the drop won't be instantaneous.
I think what you're trying to say is that countries with shutdowns may have an earlier downturn than those without
I would go further. I don't think there is any country where there has been a sharp downturn without mitigation measures. Herd immunity levels have not been reached in any country, as far as I know. Maybe I am wrong about that. If you have an example, please cite it. But even if there are such cases, the evidence is clear that mitigation saves many lives. Earlier and more strict mitigation measures in the US would certainly have saved many lives, and would probably have been shorter, so that even under a purely economic value system, it would have been preferable to the current course.

Edit: I should add that the "lockdown" in the US is a joke. It's a patchwork of different policies from one place to another, with few controls on people moving from place to place. That is unlikely to work well, though it may "flatten the curve".

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Lou Jost wrote:
The downturns in the data are not due to whether a country went into lockdown, but due to the virus starting to run out of hosts to infect.
I don't think this is true. Take China. Most of the country was never exposed to the disease. That is true in most countries. Do you have some examples or candidate countries where you think herd immunity is actually stopping the disease? Certainly that is not true in the US.

RP>> I said nothing about herd immunity. It does not take herd immunity for infections to peak, then fall.
The "flattening of the curve" caused by the shutdowns and social distancing, by definition, should result in more gradual downturns after the virus finds fewer and fewer hosts, not sharper, so the sharp and sudden downturns must be due to some other factor.
The graph I linked to shows the sudden stopping of exponential growth. That was to address your earlier claim that all the data showed continuing exponential growth. we may differ on what "sudden" and "gradual" means. But I would claim that in most cases there was a sudden change from exponential growth, and this does not seem to be due to lack of hosts.

RP>> it is also uncorrelated with timing of shutdowns.
What I interpreted you as saying (and you confirmed) was that the shutdowns were associated in time with downturns, and this is clearly incorrect.
Your interpetation of what I said is correct. But maybe you thought I meant "simultaneous with" rather than "associated with". Of course the drop won't be instantaneous.

RP>> All the models I have seen show the shutdowns as causing a delay and reduction in the peak. This was ostensibly their purpose.
I think what you're trying to say is that countries with shutdowns may have an earlier downturn than those without
I would go further....

RP>> I assume this means "yes", yet models say the exact opposite to be expected. That doesn't mean they are right though!

Locked Previous topicNext topic