Lockdown - Is there Death Rate Modelling available yet?

[G.D. Gentleman]

For those doing so much more research and sharing in this forum, thank you.

I was watching a movie last night again, the Big Short. In the movie Brad Pitt's character makes a troubling statement:

“Every 1 percent unemployment goes up, 40,000 people die. Did you know that?”


Curious if there was any reality to this or just hollywood attempt to drive home a point, I did a bit of Google searching and found this article from just a few days ago.

https://nypost.com/2020/04/20/explaining-the-link-between-unemployment-deaths-amid-coronavirus/


According to this article, the real academic study number is 37,000 deaths per percent just in the U.S. based on the studies performed in the 1980's.


The math then shows (for example) a 15% increase in unemployment in the U.S. would result in 555,000 deaths based on this modelling from years ago.
*Qualifier - this math is from 50 years ago and we should all be aware of this.

This perspective however does demonstrates a lot of the battle we see in the threads here - the lockdown is saving deaths from COVID-19 and I fully agree with the steps taken - but what is the data and modelling showing for deaths we are going to experience elsewhere from the actions we have taken? There is always a trade off in these difficult situations and I wanted to bring what little I have found so far to a thread for discussion.


Has anyone found more current/recent modelling related to the actions the world is taking for fight against COVID-19 and can share links in this thread perhaps? Please share if you do.

Take care everyone.

 

[Smallcock]

Thanks for sharing this. Very interesting and important stats.

 

[Malibuk]

This is a very interesting question, how to quantify the fatality rate of the shutdown?

I don`t doubt this 37k per 1% figure, but I don`t think it is just simple math if the US unemployment rate shoots up to 20% soon.

This is an orchestrated shutdown and expected collapsing of the economy, which presumably will bounce back much better and faster than a 20% rate in the Great Depression.
Or in 2008, without a massive concerted effort to prevent an economic collapse, there would have been another depression with a similar %, and this is no exaggeration.

I guess the question is how much of this unemployment is not coming back any time soon, if ever?

For sure there will be more suicides, domestic violence, armed robberies, and other poverty related maladies, but the biggest factor is likely the current efficiency of the social security net.

For the hundreds of millions with no social safety net and who could be on the brink of starvation this year, the shutdown fatality rate will of course be absolutely devastating.

This is a very tough question and will depend on the location and extensive research.

One thing that is easy to know, it is not a question that the trash-Sweden crowd will want to ponder.

When I read the title of this thread I thought, fuck, here we go with the totally useless case-fatality death rate bullshit again!!!

 

[G.D. Gentleman]

This is a very interesting question, how to quantify the fatality rate of the shutdown?

I don`t doubt this 37k per 1% figure, but I don`t think it is just simple math if the US unemployment rate shoots up to 20% soon.

This is an orchestrated shutdown and expected collapsing of the economy, which presumably will bounce back much better and faster than a 20% rate in the Great Depression.
Or in 2008, without a massive concerted effort to prevent an economic collapse, there would have been another depression with a similar %, and this is no exaggeration.

I guess the question is how much of this unemployment is not coming back any time soon, if ever?

For sure there will be more suicides, domestic violence, armed robberies, and other poverty related maladies, but the biggest factor is likely the current efficiency of the social security net.

For the hundreds of millions with no social safety net and who could be on the brink of starvation this year, the shutdown fatality rate will of course be absolutely devastating.

This is a very tough question and will depend on the location and extensive research.

One thing that is easy to know, it is not a question that the trash-Sweden crowd will want to ponder.

When I read the title of this thread I thought, fuck, here we go with the totally useless case-fatality death rate bullshit again!!!

Some great thoughts Malibuk, thank you. I realized when I started this thread it might be seen as 'just another death rate' thread and had to re-write the title twice in hopes to not confuse it with the many other threads looking at COVID-19 directly related death calculations.


I come from the I.T. Industry and while I do not run the 'Big Data' processes for my clients, my role has been building out the infrastructure - Compute and Petabytes at times of storage capacity - for them to operate their Big Data processing and modelling - hence my interest.

As Big Data and Modelling has become more and more part of our lives the last 3-5 years, I can only believe there are companies/governments/etc running these models to plan for what we might see in the years to come.

Hopefully over the next few months Modelling will be shared for all of us to learn from and understand. This thread hopefully will collect some opinions of course (lol) and look forward to any other links to reports/findings/etc in this theme.

Take care everyone.

 

[doggystyle99]

This is a very interesting question, how to quantify the fatality rate of the shutdown?

I don`t doubt this 37k per 1% figure, but I don`t think it is just simple math if the US unemployment rate shoots up to 20% soon.

This is an orchestrated shutdown and expected collapsing of the economy, which presumably will bounce back much better and faster than a 20% rate in the Great Depression.
Or in 2008, without a massive concerted effort to prevent an economic collapse, there would have been another depression with a similar %, and this is no exaggeration.

I guess the question is how much of this unemployment is not coming back any time soon, if ever?

For sure there will be more suicides, domestic violence, armed robberies, and other poverty related maladies, but the biggest factor is likely the current efficiency of the social security net.

For the hundreds of millions with no social safety net and who could be on the brink of starvation this year, the shutdown fatality rate will of course be absolutely devastating.

This is a very tough question and will depend on the location and extensive research.

One thing that is easy to know, it is not a question that the trash-Sweden crowd will want to ponder.

When I read the title of this thread I thought, fuck, here we go with the totally useless case-fatality death rate bullshit again!!!

This is the exact issue with the people who keep questioning the fatality rate of the Coronavirus and those who keep promoting Sweden as a success.
In the same post you question the fatality rate of the Coronavirus which is based on current data you also make an extreme assumption based on your feelings and opinions on a future fatality rate which no one can except you can forsee.

Like I've said before these people who keep questioning the fatality rate of the Coronavirus and promoting the idea of opening up businesses based on the fact that Sweden kept businesses open are posting these non factual statements in an attempt to open businesses back up (which is a point started by the right wing nutbars).

 

[doggystyle99]

For those doing so much more research and sharing in this forum, thank you.

I was watching a movie last night again, the Big Short. In the movie Brad Pitt's character makes a troubling statement:

“Every 1 percent unemployment goes up, 40,000 people die. Did you know that?”


Curious if there was any reality to this or just hollywood attempt to drive home a point, I did a bit of Google searching and found this article from just a few days ago.

https://nypost.com/2020/04/20/explaining-the-link-between-unemployment-deaths-amid-coronavirus/


According to this article, the real academic study number is 37,000 deaths per percent just in the U.S. based on the studies performed in the 1980's.


The math then shows (for example) a 15% increase in unemployment in the U.S. would result in 555,000 deaths based on this modelling from years ago.
*Qualifier - this math is from 50 years ago and we should all be aware of this.

This perspective however does demonstrates a lot of the battle we see in the threads here - the lockdown is saving deaths from COVID-19 and I fully agree with the steps taken - but what is the data and modelling showing for deaths we are going to experience elsewhere from the actions we have taken? There is always a trade off in these difficult situations and I wanted to bring what little I have found so far to a thread for discussion.


Has anyone found more current/recent modelling related to the actions the world is taking for fight against COVID-19 and can share links in this thread perhaps? Please share if you do.

Take care everyone.

GD
There is an important part of the article that I found interesting

"According to one study [the one by Bluestone et al.] a 1 percent increase in the unemployment rate will be associated with 37,000 deaths [including 20,000 heart attacks], 920 suicides, 650 homicides, 4,000 state mental hospital admissions and 3,300 state prison admissions.”

The second book published in 2005 is quoting the same study which is from 1981.
Without multiple studies to prove or dissaprove that number how can a single study be used to quote the 37 000 deaths per 1% unemployment to write an article by the author.

As I said in another post there are people specially on the right wing nut bar spectrum who are grasping at any argument or any loose fact they can get their hands on to argue about opening the economy up for business. Hence where the "Sweden is a success" started.

 

[jcpro]

How are you suppose to get the correct death rate modelling when we have no idea about the infection rates?

 

[Valcazar]

It's a complicated relationship and seems to be very heterogeneous in how it affects populations. (At a population level, in general, recessions improve health. But that is too simple and ignores real effects.)

https://www.nature.com/articles/d41586-019-00210-0

The classic Rhum study in 2000 found that 1% drop in employment was 11,000 lives saved.

There are all kinds of strange patterns here. Even weather.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6368579/

 

[shack]

“Every 1 percent unemployment goes up, 40,000 people die. Did you know that?”

So that means with 100% unemployment only 4M people would die. For a country of 350M people that's just over 1% of the population.

 

[Malibuk]

How are you suppose to get the correct death rate modelling when we have no idea about the infection rates?

It is impossible.

The only thing known now are case-fatality rates, number of dead divided by number of known cases.
This is totally useless for 2 reasons, first of all because they are not even remotely accurate.

Secondly, they are not useful for real-time comparisons.
The reason for this is that they are not comparing countries that have tested the same percentage of people, and using the same sampling methodology.

To illustrate this, let`s compare Ontario to Ontario.
The current case-fatality rate in Ontario is 763 dead divided by 13,519 cases = 5.6%.

If Ontario had tested half as many people now, and presumably found half as many cases, the case-fatality rate would be 763 divided by 6795 = 11.3%.

Regarding sampling methodology, Ontario if focusing the testing on those most likely to be sick.
They are going to test 100% of the people in old-age and long-term homes, where infection rates are astronomical and where over 70% of deaths have come from.

If they did a broad diverse random sampling, of course they would find way fewer cases.
Again, the denominator would be way smaller so the case-fatality rate would be way higher.
Expanding on the previous example, the case-fatality rate would be way higher than 11.3%.

 

[Malibuk]

How are you suppose to get the correct death rate modelling when we have no idea about the infection rates?

This German anti-body test found a death rate of 0.37%.

 

[Darts]

"37,000 deaths [including 20,000 heart attacks], 920 suicides, 650 homicides, 4,000 state mental hospital admissions and 3,300 state prison admissions.”"

Could be from a variety of causes.

 

[G.D. Gentleman]

This is the exact issue with the people who keep questioning the fatality rate of the Coronavirus and those who keep promoting Sweden as a success.
In the same post you question the fatality rate of the Coronavirus which is based on current data you also make an extreme assumption based on your feelings and opinions on a future fatality rate which no one can except you can forsee.

Like I've said before these people who keep questioning the fatality rate of the Coronavirus and promoting the idea of opening up businesses based on the fact that Sweden kept businesses open are posting these non factual statements in an attempt to open businesses back up (which is a point started by the right wing nutbars).

Hello DS99,

As mentioned in my original post - I fully agree with what Canada is doing. This thread is in no way an attempt to defend/justify Sweden.

What I am calling on from my fellow members is to share here any modelling insight they find related to what majority of countries are in the midst of now - locking down their economies.

There has been modelling in our society for decades now - traffic flow, car accidents - amount of liquor sold in an area and alcohol related deaths - just to name a few old school models that come to mind. Modern day modelling of course comes from social media - Facebook and Google ad campaigns for example.


My hope is to find some current modelling for what our collective actions is having on our country in the short term and long term, to add to the overall conversation. You have shown to be quite resourceful in finding information to share and hope you might stumble onto some such modelling to share with all of us.

 

[G.D. Gentleman]

GD
There is an important part of the article that I found interesting

"According to one study [the one by Bluestone et al.] a 1 percent increase in the unemployment rate will be associated with 37,000 deaths [including 20,000 heart attacks], 920 suicides, 650 homicides, 4,000 state mental hospital admissions and 3,300 state prison admissions.”

The second book published in 2005 is quoting the same study which is from 1981.
Without multiple studies to prove or dissaprove that number how can a single study be used to quote the 37 000 deaths per 1% unemployment to write an article by the author.

As I said in another post there are people specially on the right wing nut bar spectrum who are grasping at any argument or any loose fact they can get their hands on to argue about opening the economy up for business. Hence where the "Sweden is a success" started.

I agree in what little I have seen so far appears to be manipulated to fit the presenter's prerogative. Hence I gave an example, noted it was based on 50 year old calculations - simply to give a framework of what we could hope to find - and hope some reputable modern day Big Data source comes out with neutral and detailed modelling for the public, and policy makers - to have and learn from.

 

[G.D. Gentleman]

"37,000 deaths [including 20,000 heart attacks], 920 suicides, 650 homicides, 4,000 state mental hospital admissions and 3,300 state prison admissions.”"

Could be from a variety of causes.

Agreed - hence a modelling on Big Data here in 2020 would be much more complete and detailed.

 

[G.D. Gentleman]

So that means with 100% unemployment only 4M people would die. For a country of 350M people that's just over 1% of the population.

Good catch on the math there. Without reading the original source information myself, I can only speculate there are thresholds and the 37,000 per 1 percent went up to a XX percent level and then increased there above.

Again, my hope is to find more detailed and modern modelling from a reputable source to be shared here for all of us to learn from.

 

[G.D. Gentleman]

One more point I wanted to share based on my original thought for this thread:

Dr. Fauci and Trump announced a few weeks ago the modelling for the U.S. if no action was taken (social distancing, lockdown, etc) the estimated deaths from COVID-19 would be around 2.2 million Americans. That's a big, big number.

My hopes are those with the data and ability to do so are modelling, for every country (I reference the U.S. only due to having enough information to frame this conversation and thread) of what the lockdown impact will be, what areas will be hit hardest, demographics, etc - so we can prepare for what is to come.

For very simple math, if COVID-19 results in 70,000 American deaths - to compare that count to the possible 2.2 million originally estimated is incomplete and unfair. To perhaps add the 70,000 in this hypothetical outcome with say 280,000 increased deaths due to lockdown and economic depression would give a much more clear picture of the 350,000 lives lost but 1.85 million estimated lives saved based on the original modelling.

And yes, all I share here is just adding my small part to the many threads in motion here in attempt to better understand, together with all you my TERB friends, what is going on around us and what the next few years have in store for us.

Take care everyone.

 

[jcpro]

This German anti-body test found a death rate of 0.37%.

That would be more inline with the findings from the LA County study which tested for antibodies as well. In the end, it looks like, we will find that covid19 is nowhere near as deadly as predicted.

 

[Malibuk]

It's a complicated relationship and seems to be very heterogeneous in how it affects populations. (At a population level, in general, recessions improve health. But that is too simple and ignores real effects.)

https://www.nature.com/articles/d41586-019-00210-0

The classic Rhum study in 2000 found that 1% drop in employment was 11,000 lives saved.

There are all kinds of strange patterns here. Even weather.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6368579/

Some interesting info there Valcazar.

On the one hand, people who lose their jobs are much more likely to die than had they not lost their jobs.

But for the whole population, there is less death due to factors like less traffic, less pollution, and less money to blow on unhealthy vices.

This is relevant for the net effect on the population size, but I think this is just a bonus side effect, and it is still possible to independently consider the deaths that are the result of the job losses.
We can shut down and save lives in this manner without a virus.

This virus is killing people.
It wouldn`t make sense to say, yeah but now there is less traffic and pollution, less CO2, less drain on resources, more wealth to inherit, etc...

The virus has killed X number of people.
The shutdown has killed Y number of people.

The issue here is how do we quantify Y.

 

[G.D. Gentleman]

In the end, it looks like, we will find that covid19 is nowhere near as deadly as predicted.

Percentage wise, yes.

2,300+ Canadians and 200,000 people world wide have died already because someone coughed near them. It's not perhaps how many people have died so much as 'how' they died, compared to being an accident, cancer, etc. It's all about perspective I suppose.