I read some claims that Covid-19 fatalities are declining. I want to — again — remind readers about the hazards of interpreting (1) administrative data, and (2) data revisions.
First, there are official tabulations of fatalities due to Covid-19. We should worry about suppression of data in, for instance, Florida, but let’s take the CDC data (which compiles the data provided by authorities) at face value. It’s not clear that all fatalities attributable to Covid-19 are caught by the tracking system. Then we might use “excess fatalities” as a check on the administrative tabulation. Figure 1 below shows how the CDC data on Covid-19 deaths matches the deviation from the fatalities we expect (“excess fatalities”).
Figure 1: Weekly fatalities due to Covid-19 as reported to CDC for weeks ending on indicated dates (black), excess fatalities calculated as actual minus expected (teal), fatalities as tabulated by Our World in Data (dark red). Note excess fatalities differ from CDC series which are bounded below at zero. Light green shading denotes CDC data that are likely to be revised. Source: CDC 9/23/2020 vintage, OurWorldinData accessed 9/24/2020 and author’s calculations.
Note that excess fatalities far exceed the officially designated Covid-19 fatalities for most of the sample. (See discussion in latest Economist on excess fatalities around the world.) Only recently have excess deaths fallen below designated Covid-19 fatalities.
In fact, both of the CDC series – Covid-19 Fatalities and Excess Fatalities – drop off dramatically in recent weeks. If you didn’t read the notes attached to the CDC spreadsheet, you’d conclude that we’ve won! But inspection of the spreadsheet reveals notes that indicate that the most recent data is incomplete. In fact, as I show in this post, about the four most recent weeks worth of data are going to be substantially revised. I shade this period in green in the above graph. A hint that this is a substantial problem is provided by comparing the trajectory of the unofficial tally compiled by the Our World in Data, which indicates a much smaller decline.
The current vintage’s August 8th observations is moved up from the previous vintage so that excess fatalities are rising through the 8th. CDC designated Covid-19 deaths for the week ending August 8 rose 149 going from previous to latest vintage; excess deaths rose 486 from the previous vintage to the current. For more comparisons, see Figure 2.
Figure 2: Excess fatalities, 9/23 vintage (blue), 9/16 vintage (tan), 9/9 vintage (green), 9/2 vintage (red), 8/25 vintage (brown). Note excess fatalities differ from CDC series which are bounded below at zero. Source: CDC , various vintages, and author’s calculations.
That means it is possible that “excess fatalities” — our proxy measure for Covid-19 fatalities — is still increasing through August 8th (although officially designated Covid-19 fatalities are probably slightly declining, as the Our World in Data series not subject to really large revisions).
Menzie, by know you have probably clued into the fact I consider myself slightly above average in intelligence, but from time to time I am willing to concede I am a lazy son-of-a b****. This is one of those 1,000,001 times when you get asked a question you wished people/students would look up for themselves. Are there other models that similar to IHME’s that “lean on” excess deaths numbers of prior years to make their projections or numbers?? If so can you list the 2–3 models you personally view as the best?? I know you would never depend on a single model or declare one model “the best”, but it seems you like IHME, and I was curious if you favor any others?? Is Gu’s model similar in that same aspect as IHME on factoring in excess deaths in their graph line?? Again, others that specifically factor in excess deaths that you like???
* first sentence in my above comment should read “by now”
Moses Herzog: I cite IHME because it had the easiest interface, was well known, and I know one of the two main developers. The model has changed a lot over time, so hard to know the attributes. Also hard to do forecasting comparison since each of the models has different conditions (incorporate opening up, or condition on current social distancing, etc.) The Economist cited Gu’s model as the most accurate several months ago, so that’s why I use his forecasts. Both of them typically end up in the middle of the distribution of forecasts of deaths.
Thank you Menzie. Not the reply I was expecting but still educational.
I think there is about an 80% chance the FDA will approve a Covid 19 vaccine before November’s presidential elections. This article is timely and useful. By the way, I am generally pro-vaccine for anyone wondering my agenda:
https://fivethirtyeight.com/features/how-to-know-when-you-can-trust-a-covid-19-vaccine/
Not a chance. There is no vaccine.
@ Rage
I think most regular readers here know, just because I’m predicting FDA approves a vaccine before the November election, doesn’t mean I am endorsing that vaccine or I am saying it is efficacious. To me, it’s a very serious thing, I have said at least twice before on this blog that I believe the true/real vaccine will appear sometime after July 2021. Not that anyone ever listens to me, but I will be telling all my friends and family to avoid ANY Covid-19 vaccines that are approved near the presidential election. PERIOD, no provisos.
The flu shot is a different deal, and people will have to make their own judgements on whether to get a flu shot when it means self-exposure to other things. I lean towards recommending getting the flu shot, but I think it’s a judgement call for each individual. I know someone who can get the syringe etc and give me the shot at home. If you’re older (50+ age) and are lucky enough to know an RN or other medical professional, that is how I would get the flu shot—have them get the dilly bopper equipment and give the flu vaccine to you in your own home.
Moses,
This is from another thread, but it looks that you are about to be right about something that was discussed earlier. Apparently Claudia Sahm will be leaving the Center for Equitable Growth in Washington where she has been a Senior Research Economist for about a year or so and will be moving to Bloomberg next month full time as a columnist.
The timing seems strange. I would think the Center for Equitable Growth job would be better. Maybe the Bloomberg job bestows more status. I honestly don’t know. I think when someone is always switching jobs it leaves some questions in the air. She may be a terrific person, but I think it’s better to hold onto jobs for at least 5 years, especially if it’s lateral move to the new job. She’ll have more freedom with the Bloomberg job to put some wave making opinions out there, and maybe on a personal level that’s what she needs right now.
Lots more public exposure at Bloomberg. The Center is known to readers here, but not to the broader educated public. Good for future income prospects. Also probably good for the all-important perception of “balance”.
Moses and macoduck,
I do not know nor have ever met Claudia Sahm, so I do not know why she is doing what she is doing. It could be as simple as the Center for Equitable Growth having a funding shortfall. I do know its director, Heather Bo9ushey, a capable progressive economist, but I am not going to ask her.
As it is it is inaccurate to say she is changing jobs a lot. She was at the Board of Governors for nearly 20 years and had gotten to a serious managrerial position. The mystery is why she left that, which I have not seen her explain. She has complained about there being people there who just were always giving her a hard time, and that may be it. I remind that Chairman Powell has acknowledged specifically in response to her July 29 post that the Fed does have diversity issues and that they are trying to deal with them.
@ Barkley Junior
I believe it was only 12 years Sahm worked at the Fed, not 20. So you’re only 8 years off there “mathematical economist”. Is that one of your “small errors” or where do you rank being 8 years off Junior??
Sure, Moses, you are right on this one. She worked there for 12 years, 07-19. I was thinking she started in 02, but I was wrong on that. I said she worked there “nearly 20 years,” not a full 20 as you claimed. I do not know her, and I brought this up to say you were right that she was working at Bloomberg, although it is that she will be doing so soon.
BTW, Moses, while it was “only” 12 years she worked at the Fed rather than “nearly 20,” it remains the case that I am right and you are wrong regarding your claim that she is “always switching jobs.” She had one for a good 12 years and then took another that is now ending after about a year. That is not “always switching jobs.” Sorry.
Based on the published clinical trial submissions, the Pfizer vaccine is the only one that ‘might’ be approved prior to the election. However, they recently announced that they are increasing the number of patients in the trial which might mean they cannot complete the trial ahead of the October 22 FDA Vaccines Advisory Committee meeting. I don’t think they will have enough data on adverse reactions which is troubling. It’s also unclear about the impact of FDA’s recent announcement on possible new guidelines for COVID-19 vaccine manufacturers will mean. The Pfizer vaccine will be hobbled at roll out as it requires cold storage at -70 Celsius which means ultra-cold freezers and/or lots of dry ice. CDC said yesterday that most of America won’t be vaccinated until the middle of 2021.
The big thing for me is which vaccine will be best for seniors and it is unclear whether a non-adjuvant preparation such as the Pfizer & Moderna vaccines will meet my criteria.
What is not clear to me is whether drug time will go along with a premature announcement of the sort Moses has in mind. I agree that Trump will do anything to make such an announcement, but I kinda think Fauci may be playing nice sonas to be in the saddle in case of that sort of treachery. Cross Trump when it counts most.
J&J has a single dose vaccine in trials. Pfizer’s is a two-dose vaccine, I think. So Pfizer will require twice as many syringes. Is the syringe shortage sorted out?
The criteria is both safe and effective. If they really lucked out on a large “natural rate” of infection in their trial population AND a hugely effective set of injections – then they may be able to prove 50% efficacy before the election (highly unlikely but not impossible). But there is no way they could prove safety. Because this “drug” messes with the immune system, the side-effects could easily occur several months after the last injection. Unless they have included cardiac MRI in their protocol the most serious potential side-effect of damage from cardiac inflammation could take a lot longer. If you don’t look you won’t see.
All data sources are somewhat suspect because they attempt to aggregate information that uses differing criteria for attributing deaths to COVID-19. Still, they tend to be directionally correct. The CDC is probably the most conservative source of information because it attempts to count deaths based on death certificates and the cause of death provided in each. This requires time for confirmation which is why it takes 6-8 weeks to get a level close to reliable from the CDC. https://www.dropbox.com/s/gikhixkstxdxwhy/Covid-19%20Deaths%2C%20Cases%2C%20and%20Hospitalizations%20-%209-22-20.pdf?dl=0
Other sources automatically include “probable” cause of death attributed to CDC. In Michigan, for example, “probable” COVID-19 cases are 9.6% of the total, while “probable” deaths are 4.5% of the total. There’s a bit of “eyeballing” in the data.
The CDC does provide its own estimates of excess deaths including a seasonality adjustment. The last few weeks will be subject to upward revision as death certificates are processed. What is clear is that the early spike in Spring was about twice as high as the recent spike peaking near the end of July.
From the CDC’s other data, it is clear that the elderly and particularly those in nursing homes account for a significant portion of those excess deaths and can be directly linked to state policies that forced nursing homes to accept infected patients which put non-infected residents and staff at risk. Furthermore, states such as New York which had specific policies forcing infected patients into nursing homes have “gamed” the statistics by not including those who became infected in nursing homes, but died in hospitals as nursing home related deaths. https://www.chicagotribune.com/coronavirus/ct-coronavirus-six-months-of-death-illinois-nursing-homes-20200924-blximtfyujbeta57hwhmxism3e-story.html
Of course, nursing home residents were the expendables. They were going to die soon anyway.
Still blaming the victims in NY with your childish lying. Please watch when Rand Paul tried this stupid stunt and the blow back he got from Dr. Fauci.
Bruce Hall must not have read his own Chicago Tribune link. First of all this pathetic liar claimed it said something it did not say. But how did this idiot miss the title?
‘Nobody knew how to handle this situation’: How COVID-19 decimated Illinois nursing homes, exposed government flaws and left families in frustrating limbo
The states did not know because Trump lied to them. But put this stupid politicalization of this aside and look at the graphs. The places that initially got hit hard saw dramatic decreases in daily deaths. Why? Because they ignored the advice of people like Bruce Hall and listened to people like Dr. Fauci.
Look – we have asked Brucie boy to actually READ his own damn links before he just cuts and paste the lies given to him by Kelly Anne Conway. But I guess that would be asking too much of this little boy.
Bruce Hall ends his latest Trumpian rant by calling old folks “expandables”. Yes – that is Bruce Hall for you. BTW – he sort of forget to tell us about this paragraph in Chicago Trib link:
“Six months into the pandemic, the new virus has infected more than 28,000 Illinois long-term care residents and killed more than 4,000. It’s also fueled debate over the Illinois Department of Public Health’s oversight of a mostly for-profit industry. That includes how the agency cut back inspections, at times breaking state law, as the virus raced through facilities.’
Privatization and less regulation and inspections is Trumpian policy. But do not admit that such policies led to more deaths. Did Bruce Hall not really read his own link or is just too dishonest to admit what it said?
Bruce Hall states such as New York which had specific policies forcing infected patients into nursing homes
Yet another lie. You keep repeating this lie that was first made by Trump’s public affairs assistant at the HHS. You also kept repeating the same misinformation about Gov. Whitmer’s policy. You need to find better sources. Cuomo’s directive required nursing homes to ACCEPT positive patients, but did it did not FORCE patients into nursing homes. See the difference? Furthermore, nursing homes were not required to accept positive patients if they could not care for them. Nursing homes always had the option of refusing to accept positive patients if those nursing homes were not able to provide the required level of care. Here is Cuomo’s directive:
https://dmna.ny.gov/covid19/docs/all/DOH_COVID19%20_NHAdmissionsReadmissions_%20032520.pdf
The impartial Kaiser Health News has rated your claim about New York forcing infected patients into nursing homes as “Mostly False.”
In a tweet, the HHS assistant secretary for public affairs said that New York Gov. Andrew Cuomo “forced” nursing homes across the state to admit COVID-positive patients and that this policy fueled the spread of COVID-19 that led to thousands of deaths in the nursing home population.
Although nursing homes felt pressure to accept COVID-positive patients, they were not actually forced to do so. State regulations require nursing homes to accept patients only if they can care for them, and they could have refused them on those grounds.
In addition, it’s unclear the extent to which the governor’s policy was responsible for nursing home COVID-19 deaths. Infection control is a long-standing problem in nursing homes, predating the pandemic, and a report showed peak numbers of nursing home deaths came prior to the peak influx of patients as a result of Cuomo’s advisory. While the introduction of COVID-19 positive patients into nursing homes no doubt had an effect on infection spread, Caputo’s statement suggests it was solely responsible. That’s not what the evidence shows
We rate this Mostly False.
https://khn.org/news/is-cuomo-directive-to-blame-for-nursing-home-covid-deaths-as-us-official-claims/
If you’re going to post here, then you really need to perform some level of due diligence. This is getting ridiculous. You and CoRev keep wasting everyone’s time with your poor research and analytical skills.
As to the Chicago Tribune link…did you actually read it? It doesn’t sound like it.
There is a reason Brucie boy repeated this lie just now. Gov. Cuomo sees how Trump is interjecting his political crap into the FDA review of the efficacy and safety of any vaccine and has decided to have a group of scientists in NY to do their own independent and honest review of the science.
Of course Trump is enraged and decided it was all Cuomo’s fault for 11,000 deaths.
And we have known Bruce Hall is Kelly Anne Conway’s little puppet. Time to attack Cuomo with these lies so why are we surprised that Bruce does his little job?
“You and CoRev keep wasting everyone’s time with your poor research and analytical skills.”
Our host is telling CoRev that if he keeps lying here, he will be banned. Not to defend CoRev as he really has become tiresome with his incessant disinformation but he is not nearly as bad at this at Kelly Anne Conway’s little pet puppet.
I noticed bruce does not comment on the poor performance of the FEDERAL prison system.
Or Georgia, or Texas, or Florida, or Arizona. And now he could be commenting on the exploding case numbers in the Dakotas. But those places are run by Republicans who have no excuse. They could have learned from the mistakes New York and other places made. But, as the death cultists they are, they chose not to. So it goes.
Thank you for some excellent links to debunk the right wing lies about nursing homes in NY. It should also be added that Cuomo at the time had no good choices because of the pathetic lack of plans and support from the Federal Government and Trump administration. People in NYC were dying because hospitals were running out of beds and caregivers. Filling up hospital rooms and hallways with old folks who could just at well be isolated in their (nursing) homes would have cost a lot more lives. All patients (old, young, nursing homes, apartments…..) were treated to the same “you are not sick enough to take up a hospital bed” policy. It’s called triage and it is a part of any emergency. If the Trump administration had been competent this would not have happened – they would have build emergency half-way centers to house those who needed a lower level of care. But after a couple of idiotic failed responses, Trump/Kushner decided that it was not their job to step in and help during a national health emergency.
“t should also be added that Cuomo at the time had no good choices because of the pathetic lack of plans and support from the Federal Government and Trump administration. People in NYC were dying because hospitals were running out of beds and caregivers. Filling up hospital rooms and hallways with old folks who could just at well be isolated in their (nursing) homes would have cost a lot more lives. All patients (old, young, nursing homes, apartments…..) were treated to the same “you are not sick enough to take up a hospital bed” policy. It’s called triage and it is a part of any emergency. If the Trump administration had been competent this would not have happened – they would have build emergency half-way centers to house those who needed a lower level of care. But after a couple of idiotic failed responses, Trump/Kushner decided that it was not their job to step in and help during a national health emergency.”
Precisely right but let’s make this easy for mental midgets like CoRev. It is called a lack of resources. Oh, oh, CoRev had a rebuttal. The CDC told folks on March 13 not to let visitors in these nursing homes. Like that would have changed a damn thing. Even Trumpian clown Bruce Hall is not saying it was an open visitor policy that led to these deaths.
Memo to CoRev – stay out of an issue where you have no clue. Of course, that means he would never comment again as he is clueless about everything.
Liberal logic in action: “Bruce Hall states such as New York which had specific policies forcing infected patients into nursing homes”
2slugs response “Cuomo’s directive required nursing homes to ACCEPT positive patients, but did it did not FORCE patients into nursing homes. See the difference? Furthermore, nursing homes were not required to accept positive patients if they could not care for them. …”
Synonyms for required: “compulsory, forced, imperative, incumbent, involuntary, mandatory, necessary, nonelective, obligatory, peremptory”
https://www.merriam-webster.com/thesaurus/required
In 2slugs own quote: “In addition, it’s unclear the extent to which the governor’s policy was responsible for nursing home COVID-19 deaths. Infection control is a long-standing problem in nursing homes, predating the pandemic, and a report showed peak numbers of nursing home deaths came prior to the peak influx of patients as a result of Cuomo’s advisory. While the introduction of COVID-19 positive patients into nursing homes no doubt had an effect on infection spread, Caputo’s statement suggests it was solely responsible. ”
Peak numbers achieved/known before the policy, and yet Cuomo issued it?
This parsing and logic failure is getting ridiculous.
Oh Gee. You should have been a tax attorney the way you twist words around. Bruce has been lying about this issue for a long time and he has been called on this issue over and over. But do pretend you know everything from this single exchange of comments. Trust me troll – you do not.
BTW – I live in NYC so I have followed this issue for a while. We could have done better had we been provided with timely information and better resources. But the Federal government failed us on both.
Also note if either you or Bruce decided to take the train here and started blaming us for our own plight, you would not make it out of Penn Station. And our police force would have better things to do with their time than defend two pathetic clowns like you.
CoRev How dumb are you? Bruce Hall was claiming that patients were forced into nursing homes, as in “forced against their will.” Cuomo’s directive was a conditional requirement for nursing homes, not patients. If a patient wanted to go to a nursing home, and if a nursing home was able to take that patient, then the nursing home could not deny that patient admittance. As the factchecker noted, as a practical matter nursing homes could refuse acceptance if they didn’t think they could handle the patient. You completely misunderstood what I wrote.
This parsing and logic failure is getting ridiculous.
Your reading comprehension is ridiculous.
Its too bad trump did nothing to protect the vulnerable in the nursing homes. As he has said, it wasnt in his job description.
Remember – Trump said he is not responsible.
Baffled claim: “Its too bad trump did nothing to protect the vulnerable in the nursing homes. As he has said, it wasnt in his job description.”
We have these references: Posted at 7:21 PM, Mar 13, 2020 https://www.thedenverchannel.com/news/coronavirus/cdc-issues-new-guidance-for-nursing-homes-including-banning-visitors
and last updated 12:27 AM, Mar 14
Or the actual CDC guidance on, 2020 Updated Apr. 30, 2020 and again in June Updated June 25, 2020. You can look up the CDC guidance links. it’s easy. I’l even help with a search argument: “CDC nursing home guidance”
If you think that guidance was either timely or adequate, you are the dumbest troll ever. But nice try.
And this was while trump was publicly saying no worries, the virus will disappear soon. Its just the flu. Trump publicly denigrated anybody and anything that reacted to the virus. This public dispay overran anything the cdc could do ti sway public behavior. Trump knew of the virus and its issues in february, but did not act to protect the nation. He preferred to protect his reelection chances. And the nation suffered as a result.
September 24, 2020
Coronavirus
US
Cases ( 7,185,471)
Deaths ( 207,538)
India
Cases ( 5,816,103)
Deaths ( 92,317)
Mexico
Cases ( 710,049)
Deaths ( 74,949)
France
Cases ( 497,237)
Deaths ( 31,511)
UK
Cases ( 416,363)
Deaths ( 41,902)
Germany
Cases ( 281,345)
Deaths ( 9,519)
Canada
Cases ( 149,094)
Deaths ( 9,249)
China
Cases ( 85,314)
Deaths ( 4,634)
@ ltr
I’m so glad you’re on this blog ltr, you give me added motivation to post more about what Beijing does to its own people everyday:
https://www.nytimes.com/2020/09/24/world/asia/china-muslims-xinjiang-detention.html?action=click&auth=login-email&login=email&module=News&pgtype=Homepage
How is Huang Kunming doing lately ltr ??? If he hired you he needs to have his A$$ fired immediately.
September 24, 2020
Coronavirus (Deaths per million)
US ( 626)
UK ( 616)
Mexico ( 580)
France ( 483)
Canada ( 245)
Germany ( 114)
India ( 67)
China ( 3)
Notice the ratios of deaths to coronavirus cases are 10.1%, 6.3% and 10.6% for the United Kingdom, France and Mexico respectively.
https://news.cgtn.com/news/2020-09-25/Chinese-mainland-reports-8-new-COVID-19-cases-all-from-overseas-U459W8rGlq/index.html
September 25, 2020
Chinese mainland reports 8 new COVID-19 cases
The Chinese mainland registered 8 new COVID-19 cases on Thursday, all from overseas, the National Health Commission said on Friday.
One of the new cases was detected in Beijing and was first reported as an asymptomatic case on September 19.
This is the 40th consecutive day without domestic transmissions on the Chinese mainland. No deaths were reported, and nine patients were discharged from hospitals.
The total number of confirmed cases stands at 85,322 and the death toll at 4,634 on the Chinese mainland, while 398 asymptomatic patients are under medical observation.
Chinese mainland new imported cases
https://news.cgtn.com/news/2020-09-25/Chinese-mainland-reports-8-new-COVID-19-cases-all-from-overseas-U459W8rGlq/img/22ac882be8084a2983b84138f4b048fd/22ac882be8084a2983b84138f4b048fd.jpeg
Chinese mainland new asymptomatic cases
https://news.cgtn.com/news/2020-09-25/Chinese-mainland-reports-8-new-COVID-19-cases-all-from-overseas-U459W8rGlq/img/d6edb1fe5bf14452bf794b8da17673c9/d6edb1fe5bf14452bf794b8da17673c9.jpeg
@ ltr
I found this on Reddit, what do you think ltr ??
https://xjdp.aspi.org.au/map/?
It is called the “Xinjiang Data Project”: Mapping Xinjiang’s detention system with 380 sites of suspected re-education camps, detention centers and prisons that have been built/expanded since 2017.
I hope you and Huang Kunming enjoy it ltr. When you’re not busy eating exotic meat at wet markets you can take a holiday trip over to Xinjiang and let us know about the local tourist attractions.
There has been no coronavirus death on the Chinese mainland since May 17. There has been no community or domestic coronavirus case for 40 days. Since June began there have been only 2 limited community clusters of infections, in Beijing and Urumqi in Xinjiang, both of which were contained with mass testing, contact tracing and quarantine, and both outbreaks ended in a few weeks. Imported coronavirus cases are caught at entry points with required testing and immediate quarantine. Asymptomatic cases are all quarantined.
The flow of imported cases to China is low, but has been persistent. There are as a result 166 active imported coronavirus cases on the Chinese mainland, but of which only 4 cases are classed as serious or critical.
@ ltr
ltr, Do you take lessons in telling oafish children’s lies from the MASTER of telling children’s type lies, Huang Kunming??
https://www.medrxiv.org/content/10.1101/2020.05.28.20116012v2
https://www.news-medical.net/news/20200608/Cremation-numbers-reveal-possible-suppression-of-true-COVID-19-data-in-China.aspx
https://foreignpolicy.com/2020/05/12/leaked-chinese-coronavirus-database-number-cases/
https://www.nudtdata.com.cn/
https://www.nudt.edu.cn/kxyj/d322ac209bbc47508a408b453762b83c.htm
You can tell children’s level lies better than this ltr, go back to your personal tutor of lies Huang Kunming and get a refresher course in bullsh*t fables.
I should state that the research paper I linked to above is NOT peer reviewed, And apologize to the blog hosts and readers for not stating that in the original comment just above. However, since it is authored by 4 people connected to 3 different respected institutions, I think eventually the paper will meet these standards of peer review.
Just a comment on the chart – if you are going to put out a graph where you know some of the data is not complete, simply shading an area is not a very clear indicator of the ‘bad’ areas. I’d suggest maybe dashed lines plus shading… something more obvious and easy to ‘read’ without needing to refer to the notes which are too easily overlooked or which may disappear in later publication.
Menzie, as I have previously stated I track the CDC data daily by downloading each days complete set. Thus collecting the daily changes for the whole set. I then calculate the 14 day average to compare with the CDC provided 7-day average noting the up/down direction of the difference. I have yet to see a change in direction due to updates in the original provisional data.
Accordingly I have found no evidence in the CDC daily data that: “I read some claims that Covid-19 fatalities are declining. I want to — again — remind readers about the hazards of interpreting (1) administrative data, and (2) data revisions.” Nor is there an evidence of a directional difference in your various articles over the past months.
These findings can be confirmed by the closeness of the of the CDC fatalities data with the OurWorldinData trajectories. Accepting there will be some changes with the CDC data in the provisional period.
Could it happen? Yes, of course, but AFAICT it hasn’t yet. Inflection points and trajectories remain relatively close. Which is to be expected in similarly not identically sourced data. Trends developed using the CDC Daily Data actually seem to be more accurate than those using their excess deaths data.
“Inflection points and trajectories remain relatively close.”
This is from the same troll who does not know that the first derivative is another way of expressing the slope of a curve. Come on CoRev – anyone can cut and past big words but knowing what they mean in the context of a real world issue – that is not your forte.
BTW – did your dog eat your homework? You say you did some calculations but you do not show your work. At least Bruce Hall provides his useless dropbox links. Of course he now graphs 7-day averages of news cases but not of new deaths. Go figure!
PGL, the data are right there on the CDC site. Go for it. BTW we are in a cycle where the Deaths (7 day avg V. 14 day) are going down and cases are going up.
“we are in a cycle”.
A cycle! Yes – going down for now after going up earlier. Hey CoRev – ask some high school kid what sine and cosine functions look like.
The desperation. Did the world’s worst analyst have a point with the latest nonsense?
“Did the world’s worst analyst have a point with the latest nonsense?”
You were the dipshit who introduced the term cycle. Most people’s ideas of cycles are functions that have periodic ups followed by downs followed by ups. But thanks for confirming you do not know what a sine or cosine function looks like. Most people who took high school math do but yea – you are still struggling with pre-K arithmetic.
CoRev: The point of figure 2 is that the “peak” of excess deaths gets moving later and later, suggesting that we don’t know for certain that we are in the downswing of the cycle — i.e., we might be still having rising deaths.
If there is so much uncertainty associated with a proxy data set, then it isn’t all that good. For what use is it fit?
CoRev: Employment data (particularly household, at state level) gets revised a lot too… should we dispense with it? What a stupid question. As you can see from Figure 2, we can learn what happened in the past, and something about what will happen in the future. We correlate with other series, perhaps at higher frequency, so we can make inferences.
Menzie, “We correlate with other series, perhaps at higher frequency, so we can make inferences.” Inferences? Inferences says nothing. So far the only inference stated has been there is a provisional period of data change in the RECENT/CURRENT CDC data, and all other data sets. That this provisional data period may not reflect the eventual trend trajectory. Do you have others?
NO ONE QUESTIONS THE EXISTENCE OF A PROVISIONAL PERIOD. IT’S EVEN NOTED ON THE CDC WEB SITE.
I too have an inference that the daily data HAS NOT changed enough in the provisional period to change trend trajectory.
BTW, I never asked about any other data. Why the need to reference it, and then add an ad hominem?
CoRev: I’m sorry, but your repeated citation of junk I believe is pretty stupid. However, I agree I should not have said you are stupid, just what you write.
I have written an entire post just for you.
Again!?! I’m flattered.
“as I have previously stated I track the CDC data daily by downloading each days complete set. Thus collecting the daily changes for the whole set. I then calculate the 14 day average to compare with the CDC provided 7-day average noting the up/down direction of the difference. I have yet to see a change in direction due to updates in the original provisional data….These findings can be confirmed by the closeness of the of the CDC fatalities data with the OurWorldinData trajectories.”
Let’s unpack CoRev’s latest gibberish by noting the 7-day average of daily deaths this month using OurWorldinData. Yes it fell to 721 on September 8 to which CoRev claimed victory as if we had “turned the final corner”. But wait – this series started going back up reaching 860 by September. OK the rise reversed itself and is now back to 741 as of September 24. Still very high and not exactly dropping like a stone as CoRev and Bruce Hall would have us believe.
I guess we see why CoRev did not provide us a graph of this series that he claims he tracks. It seems it does not provide the clear picture that we have this under control after all – regardless of all the long winded word salad CoRev routinely spews.
I bet CoRev’s hero is President Harding:
https://www.merriam-webster.com/dictionary/bloviate
Warren G. Harding is often linked to “bloviate,” but to him the word wasn’t insulting; it simply meant “to spend time idly.” Harding used the word often in that “hanging around” sense, but during his tenure as the 29th U.S. President (1921-23), he became associated with the “verbose” sense of “bloviate,” perhaps because his speeches tended to the long-winded side. Although he is sometimes credited with having coined the word, it’s more likely that Harding picked it up from local slang while hanging around with his boyhood buddies in Ohio in the late 1800s.
Of course, bloviation is a style of empty, pompous, political speech which originated in Ohio and was used by US President Warren G. Harding, who described it as “the art of speaking for as long as the occasion warrants, and saying nothing”.
“this series started going back up reaching 860 by September” 16.
Sorry about the typo here but my computer is really pissed at me for letting the serial crap from CoRev show up on its screen.
Did I hear this right? Trump signs an executive order declaring he supports existing law under Obamacare. So his plan to replace Obamacare is to keep Obamacare. Hey that was the Repeal and Replace mantra Romney tried in 2012!
https://www.census.gov/library/publications/2020/demo/p60-271.html
September, 2020
People Without Health Insurance Coverage, 2007-2019
(Thousands without insurance for entire year)
2007 ( 44,088)
2008 ( 44,780)
2009 ( 48,985) Obama
2010 ( 49,951) (Affordable Care Act)
2011 ( 48,613)
2012 ( 47,951)
2013 ( 41,795)
2014 ( 32,968)
2015 ( 28,966)
2016 ( 28,052)
2017 ( 25,600) Trump
2018 ( 27,462)
2019 ( 26,111)
(Percent without insurance for entire year)
2007 ( 14.7)
2008 ( 14.9)
2009 ( 16.1) Obama
2010 ( 16.3) (Affordable Care Act)
2011 ( 15.7)
2012 ( 15.4)
2013 ( 13.3)
2014 ( 10.4)
2015 ( 9.1)
2016 ( 8.8)
2017 ( 7.9) Trump
2018 ( 8.5)
2019 ( 8.0)
Florida Governor Desantis says restaurants and bars can open without restrictions and that local government will have no say in the matter. So yes, deaths are going to rise.
September 25, 2020
Coronavirus
Israel
Cases ( 217,899)
Deaths ( 1,412)
Deaths per million ( 154)
July 4, 2020
Coronavirus
Israel
Cases ( 29,170)
Deaths ( 330)
Deaths per million ( 36)
Having apparently approached a containment of the coronavirus in June, the Israeli government incautiously opened schools and businesses, and the result has been a persistent community infection spread contributing to what are now 217,899 cases in the small country as compared to 85,322 in all through all of mainland China.
Israel has unfortunately doubled and gone far beyond double the number of coronavirus cases in mainland China. The need then is to examine what public health structure weaknesses have allowed for such an infection spread after what had appeared to be containment.
Interesting stats about California here: https://data.chhs.ca.gov/dataset/statewide-death-profiles
“Deaths by Month” Shows very large changes from month to year-ago month.
We’ve discussed herd immunity as it relates to Covid-19 on this blog a number of times. I have stated a pretty strong belief that I don’t think we would attain any kind of herd immunity before the arrival of a mass produced vaccine (again, post July 2021). I think this research fortifies (though not outright) my subjective thoughts on that:
https://www.nytimes.com/2020/09/25/world/covid-coronavirus.html?action=click&module=Top%20Stories&pgtype=Homepage#link-6512802f
https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)32009-2/fulltext
Yet Trump and his handpicked Fox journalist/radiologist adviser seem convinced that a turbocharged version of the failed Swedish approach is the way to go in this country. Remember Sweden sacrificed life and health of a large number of people yet never reached herd immunity (because many of the sheep that were supposed to “get it over with” refused to be sacrifices). Since anybody with a brain refused to ignore the dangers, the Swedes had the same economic damage as comparable countries who took much better precautions.
TOTAL ESTIMATED US COVID-19 FATALITIES AND WHAT COULD HAVE BEEN
Following is a link to and an excerpt from a detailed WSJ analysis of how South Korea handled the pandemic.
Let’s assume for the sake of argument that we had a rational President who did not lie on a routine basis, who did not ignore reality and who was focused on doing what was best for the country, and let’s assume that the US government chose to emulate South Korea’s approach to the virus. Let’s further assume that the government was successful in this regard.
In round numbers, this would probably imply that we would see a total cumulative COVID-19 death count of about 5,000 or so in the US, for the entire pandemic (confirmed COVID-19 deaths in the US are in excess of 200,000).
The following study, which so far appears to be more or less accurate, suggests a most likely US total death count of about 410,000, just through the end of 2020.
Note that the difference between 410,000 and 5,000 is 405,000, which happens to be equal to the estimated total number of Americans who died in the Second World War.
In other words, a plausible case can be made that Trump could easily be responsible, through his depraved indifference, for the unnecessary COVID-19 related deaths, just in 2020, which are equivalencies to the number of Americans who died in the Second World War.
Key coronavirus forecast predicts over 410,000 total U.S. deaths by Jan. 1: ‘The worst is yet to come’
https://www.cnbc.com/2020/09/04/key-coronavirus-forecast-predicts-over-410000-total-us-deaths-by-jan-1.html
WSJ: How South Korea Successfully Managed Coronavirus
https://www.wsj.com/articles/lessons-from-south-korea-on-how-to-manage-covid-11601044329
Excerpt:
SEOUL—South Korea appears to have cracked the code for managing the coronavirus. Its solution is straightforward, flexible and relatively easy to replicate.
The country has averaged about 77 new daily cases since early April and recently suppressed a spike in infections. Adjusting for population, that would be the equivalent of about 480 cases a day in the U.S., where new daily cases have averaged about 38,000 over the same period. Total deaths in the U.S. due to Covid-19 just surpassed 200,000.
South Korea halted virus transmission better than any other wealthy country during the pandemic’s early months. It was about twice as effective as the U.S. and U.K. at preventing infected individuals from spreading the disease to others, according to a recent report from a United Nations-affiliated research network. South Korea’s economy is expected to decline by just 0.8% this year, the best among the Organization for Economic Cooperation and Development’s forecasts for member nations.
The key to South Korea’s success came from blending technology and testing like no other country, centralized control and communication—and a constant fear of failure.
The nation fast-tracked approval of domestic testing kits as soon as cases began hitting. It tapped into its relative wealth and hyperconnectivity, blasting text alerts to citizens if infections occurred in their area. When the supply of face masks ran short early on in the crisis, the government seized production.
At twice-a-day briefings, health officials express worry when they can only trace the origins of three-quarters of confirmed cases. Virus experts stand at the podium of government briefings and frequently warn of looming catastrophe. Nearly everyone in the country wears masks. Every confirmed patient, even those with no or mild symptoms, gets isolated at hospitals or converted dormitories run by the government. Treatment is free.
As a result, South Korea never had to mandate a lockdown, so restaurants and business were able to stay open, cushioning the blow to the economy.
“No country has adapted to living with, and containing, the virus like South Korea,” said Dale Fisher, chairman of the World Health Organization’s global outbreak alert and response network. “You don’t need or want to eradicate the virus. But you modify your behavior and get on with life.”
. . . . In total, South Korea, with a population of 52 million, has reported 23,455 cases and 395 deaths.
For months when I wake at 4AM to get ready for my run in the park socially distant as few are up that damn early, I turned on ESPN to enjoy KBO (Korean Baseball). Not bad and they have even let their awesome fans in. Of course Major League Baseball is completing its 60 game regular season so it’s playoff time. GO BRAVES. Ronald Acuna Jr. jacked a 495 home run last night!
“Key coronavirus forecast predicts over 410,000 total U.S. deaths by Jan. 1: ‘The worst is yet to come’”
Back in early July I linked to some forecast that put this near 300 thousand to which Bruce Hall kept telling us that daily deaths would rapidly decline (they at first rose only to taper only a wee bit) followed by an unbelievable amount of ire from CoRev. Of course CoRev’s ire comes laced with a lot of gibberish and lies.
Now one would hope we do not even approach 300 thousand let alone 400 thousand but Trump and his crew of angry morons are dismissing social distancing and wearing masks which alas may mean this dreadful forecasts might alas bear out. And CoRev accuses US of wanting people to die?
In other news Trump travels to Atlanta to unveil his latest meaningless ploy:
The Platinum Plan for Black Americans
Part of this plan will be to link Antifa with the KKK. I kid you not. I guess the next time a police kills an innocent black person, Trump’s plan will make sure the cops get off light. After all, Black Lives Matter supported lynching. Or something like that. Why am I reminded of this scene from Blazing Saddles?
That Blazing Saddles scene:
https://www.youtube.com/watch?v=zkVl1zfM_G4
PGL: +1000 for Blazing Saddles
Antifa = KKK? Obviously doesn’t listen to his own counterterrorism experts.
This latest episode is yet another brought to you by tiny brains everywhere in cooperation with the RWNJ fever swamp. Belated redundancy alert.
September 26, 2020
Coronavirus
India
Cases ( 5,990,513)
Deaths ( 94,533)
Deaths per million ( 68)
There were 88,942 new coronavirus cases recorded in India today. Cases have climbed above 80,000 daily in the last weeks. The distressing point that should be made is just how undeveloped healthcare infrastructure is in India, for all the growth the last couple of decades. Indian infrastructure in general is poorly developed, as Amartya Sen pointed to in the wake of being awarded a Nobel Prize in economics, but there was little response to Sen from development specialists and even express criticism from Indian economists. Now, we find India in any day recording about as many or more coronavirus cases as China has recorded in all.
Undeveloped Indian infrastructure has presented daunting problems in national efforts to limit the coronavirus spread.
India began the year as the third largest economy, and so far has experienced a fierce recession that because of structural dislocations domestically is going to take a considerable time to recover from. This means an important driver of the international economy has been lost already and will continue to be lost for a considerable time, and this will make an international recovery that much slower.
September 26, 2020
Coronavirus
US
Cases ( 7,276,752)
Deaths ( 208,949)
Deaths per million ( 630)
https://newseu.cgtn.com/news/2020-09-26/How-lethal-is-COVID-19-compared-to-cancer-flu-and-other-death-causes–U4LJxzQlA4/index.html
September 26, 2020
How lethal is COVID-19 compared to cancer, accidents, flu and other causes?
By Gary Parkinson
At its worst, COVID-19 killed almost twice as many Europeans per day as cancer – but the latest figures show the fatality rate is back down below accidental death and slightly above pneumonia and diabetes.
We’ve all been overwhelmed by the numbers flooding out of the news – cases and deaths, naked numbers and rates per 100,000 – and it can be hard to comprehend just how big a problem the pandemic is. Meanwhile, certain groups are claiming the virus is no deadlier than seasonal flu, with some using this as a reason to subvert the safety messages broadcast by worried governments.
Crunching the data from Eurostat, CGTN Europe has compared the various causes of deaths to get an idea of just how lethal the pandemic has been – both on the most recent figures and at its peak… so far.
How the numbers work
As its name suggests, Eurostat is the European Union’s statistical office and collates data from across the continent – inside and outside the EU. For ease of reference, these are given per 100,000 of population per year.
However, because we haven’t yet had a full year of COVID-19, CGTN Europe has calculated the daily rates of various causes of death and compared them to each country’s daily coronavirus fatality toll: the most recent (usually 23 September) and the highest toll so far. We’ve also sized up the figure to deaths per million of population, and beyond the EU28 (the current 27 EU members plus the UK) we’ve included European Free Trade Association members Iceland, Liechtenstein, Norway and Switzerland, plus Serbia and Turkey.
As a benchmark, the data reveals that from all causes, every day across the EU28 there are 27.27 deaths per million (although rates vary from Switzerland’s 22.2 to Bulgaria’s 43.77). Under normal circumstances, the biggest killer is cancer, with 7.08 deaths per million among the EU28 – again, ranging from 5.29 in Cyprus to Hungary’s 9.44.
[ https://newseu.cgtn.com/news/2020-09-26/How-lethal-is-COVID-19-compared-to-cancer-flu-and-other-death-causes–U4LJxzQlA4/img/5a2d481601f04074849305e755cd1392/5a2d481601f04074849305e755cd1392.png ]
However, peak COVID-19 killed almost twice as many Europeans as cancer: on each country’s most lethal day, it killed 12.95 per million across the EU28. This mortality has varied hugely between countries, though. Of the 34 countries only 12 have had COVID-19 fatalities peaking above cancer; the majority of European nations did not, which goes to illustrate the weight of deaths in those virus-wracked dozen countries.
Taking the most recent data, COVID-19 has slipped back down the mortality listings. The latest figures from Johns Hopkins have the EU28 suffering 0.75 daily deaths per million; again, there is a range, but only Bulgaria (1.73), Czechia (2.24), Romania (2.43), Spain (2.74) and Malta (3.89 – two deaths in a nation of just over half a million) suffered more than one death per million population….
Menzie, I believe you are always looking for interesting graphics regarding COVID-19 infections. Reuters has this unique set of graphics: https://graphics.reuters.com/world-coronavirus-tracker-and-maps/
Menzie, I should have added this to my previous comment: https://graphics.reuters.com/HEALTH-CORONAVIRUS/USA-TRENDS/dgkvlgkrkpb/index.html
Bruce Hall: Thanks – very nice interactive. FT’s is more up to date (running through 9/25 instead of 9/20). Of course there are tradeoffs. In any case, WI death toll rising in FT interactive, flat in the Reuters.