How Many COVID-19 Deaths Is Ron DeSantis Hiding?
I had been told a couple of days ago, either here in comments or on Twitter, that Ron DeSantis had put his cronies in charge of the Florida database for COVID-19 cases and deaths. I hadn’t followed up on that, but then on Twitter last night i learned that the scientist who had been in charge of the site was fired back on May 1. She spoke last night with a West Palm Beach TV station which broke the blockbuster story of why she was fired:
Rebekah Jones said in an email to CBS12 News that her removal was “not voluntary” and that she was removed from her position because she was ordered to censor some data, but refused to “manually change data to drum up support for the plan to reopen.”
Jones made the announcement May 5 in a farewell email to researchers and other members of the public who had signed up to receive updates on the data portal, according to Florida Today. She said that for “reasons beyond my division’s control,” her office is no longer managing the dashboard, involved in its publication, fixing errors or answering any questions.
Wow. Note that DeSantis “reopened” Florida on May 4. So the timing here, coupled with Jones saying she was ordered to change or censor data, shows a clear intent by DeSantis to game the numbers and create the false impression that the reopening would be a success. Just how stupid can a governor be?
Well, in his case, pretty stupid. He can’t even figure out how to wear a mask:
So who would trust this clown, compared to the scientist he fired? Here she is in a photo she supplied to Florida Today:
Near the end of the story filed last night, Jones notes that the Florida database still hasn’t been repaired. Here’s what I got when I checked at 1:40 this afternoon (refreshing 15 minutes later gave a plot for cases but not deaths, so there may be some “repair” work underway as I write this; and both plots were visible at 2:50):
There are big holes where the plots of new cases by day and deaths by date of death for the last 30 days would show up. Back on May 5, when Jones talked to Florida Today about her firing (but without mentioning the orders to censor data), Jones noted that near the end of her time in the job, the database suddenly started to malfunction:
Late last Friday, the architect and manager of Florida’s COVID-19 dashboard — praised by White House officials for its accessibility — announced that she had been removed from her post, causing outcry from independent researchers now worried about government censorship.
The dashboard has been a one-stop shop for researchers, the media and the public to access and download tables of COVID-19 cases, testing and death data to analyze freely. It had been widely hailed as a shining example of transparency and accessibility.
But over the last few weeks it had “crashed” and gone offline; data has gone missing without explanation and access to the underlying data sheets has become increasingly difficult.
The site was created by a team of Florida Department of Health data scientists and public health officers headed by Rebekah Jones. She announced last week her removal as of May 5 in a heartfelt farewell note emailed to researchers and other members of the public who had signed up to receive updates on the data portal.Citing “reasons beyond my division’s control,” Jones said her office is no longer managing the dashboard, is no longer involved in publication, fixing errors or answering questions “in any shape or form.”
Note that the story from the West Palm Beach TV station says Jones announced her firing on May 5, but this Florida Today story makes it clear she informed people on May 1 that she was being removed May 5. Since she speaks of the database malfunctioning at the time of her firing, for the purposes of discussion here I consider May 1 the firing date and the time when fuckstickery of the database began. With today being May 19, it’s clear that the database has been malfunctioning for nearly three weeks at a minimum.
The big problem, though, is that the plot for deaths magically starts dropping right after Jones was fired. I captured this version of the death plot around 9 this morning and noted the date Jones was fired (did she insert a parting shot of a bit more reality on her last day of May 4?):
There should be one partial note of caution here even though it’s really hard not to get the impression death numbers are being artificially reduced. There is a note on the death graph that I’m pretty sure popped up fairly recently and may well have been added around the time of Jones’ departure. The note says that deaths are now counted on the day of death rather than on the day the report is entered into the database. Since it can take a while for deaths to be reported depending on the county involved, the last few days in the plot can be expected to show increases as more death reports get entered into the system. But it has been long enough now since Jones was removed for it to be clear that there is a discontinuity in the plot that coincides precisely with her removal.
I haven’t included a plot of cases by day, but I also find the current data there (site is here) not believable. With the partial reopening of the state on May 4, it’s simply incomprehensible that the number of new cases per day is holding steady rather than increasing.
Will DeSantis ever face consequences for this egregious breach of public trust? Odds aren’t good. The Republican Party in Florida has a long tradition of doing whatever it pleases, rules and laws be damned. Just look at how they over-ruled the will of the people on the initiative overwhelmingly passed in 2018 to restore voting rights to felons who have served their sentences. It would seem that firing a scientist because she refused to censor data and mislead the public on a life or death matter would be the end of a normal political career. But in Florida, there is no limit to how criminal Republican officeholders can be.
That is odious. Morally repulsive. May this be investigated at some point and those involved (other than desantis) widely named and shamed.
More from Jones’ farewell note [from Florida Today article]
Georgia’s odious governor Brian Kemp is also juking the numbers, to the point his office felt compelled to issue an apology yesterday for the misleading charts up on the DPH. the Atlanta Journal Constitution has a great article detailing the shenanigans, with great reporting on how the numbers are being compiled. one key point is the p.r. victory Kemp has already gained by the release of the happy clappy charts:
“Wrong information about Georgia’s battle against COVID-19 is already shaping the way the public sees the state. A Friday column in The Wall Street Journal dubbed Kemp’s controversial decision to begin reopening, “The Georgia Model.”
It said the state is experiencing “a welcome trend of declining new cases and deaths.”
In fact, seven-day rolling averages of cases show only a slight decline over two weeks. Deaths appear to have plateaued, according to an Atlanta Journal-Constitution analysis of daily DPH reports. The impact of reopening Georgia’s economy is still too early to measure for both new infections and deaths because of the lag time between an infection, testing, diagnosis and, potentially, death.”
https://www.ajc.com/news/state–regional-govt–politics/just-cuckoo-state-latest-data-mishap-causes-critics-cry-foul/182PpUvUX9XEF8vO11NVGO
https://www.tableau.com/covid-19-coronavirus-data-resources/global-tracker
Use the TABS for United States > Georgia.
Not a pretty picture.
Jim, so glad that I swung by shortly after you posted!
Just to give you another glimpse, here is Tableau’s dashboard page for Florida Coronavirus — look at the steep climb in May (!)
https://www.tableau.com/covid-19-coronavirus-data-resources/global-tracker
TABS: United States > Florida
(Be sure and run your cursor over the graphs to get more granular information ;^)
The climbing stats for cases/deaths are so startling in April and May, it would increase the likelihood that DeSantis/GOP panicked.
As a practical matter, Rebekah Jones just made herself much more employable with some corporate entities who value, y’know, *actual integrity*.
Data is only ‘the new oil’ IF and ONLY IF the data is accurate, complete, and easy to read. Otherwise, it’s just more fuckery.
What legitimate company is going to hire a known liar, someone so weak they will cave to a moron like DeSantis and his ilk? The liability, the stakes in getting people to pay you for data with integrity and detail, are simply too high to tolerate the kind of stupidity that DeSantis exhibits. IMVHO, if Florida were a corporate entity, DeSantis would be asking Ms Jones to incite corporate suicide; it’s completely self-defeating.
In my experience, data scientists geek out on the quality of the data. I hope that Ms Jones has a wonderful, bright, productive future in her field. The world desperately needs her kind of talent and (what appears to be) integrity.
Having worked with databases for most of the last 40 years, and much of that in QC – data should not be messed with in that way. Incomplete data is one thing, but lying is much worse.
Agree completely.
If nothing else has made clear that elections matter, this does. Chilling.
1] May 4, 2020
1a] NYT reports [and others confirm]:
https://twitter.com/ScottGottliebMD/status/1257339598516170755
12:01 PM · May 4, 2020
1b] https://twitter.com/PhilipWegmann/status/1257392032181518339
3:29 PM · May 4, 2020
2] May, 7, 2020
AP Exclusive: Admin shelves CDC guide to reopening country
https://apnews.com/7a00d5fba3249e573d2ead4bd323a4d4
5/7/20
It appears for the GOP that the whole world is no longer a stage but the closed set of a reality t.v. show, and their mission is to manage the perceptions of the gullible masses. Meanwhile, the mainstream news media plays its part by daily promoting each new cliffhanger.
The GOP numbers game has been going on for a long time, along with the “it’s not that bad so get back to work, peasants” theme. One wonders if the Statute of Laborers will be resurrected from Merrie Olde England to add a touch of dignitude to the process.
It’s about painting the red states as better off than the blue ones.
The “leak” of Birx’s displeasure with the CDC numbers appears part of the same hide the data plan.
Similar story to Colorado now reporting two different numbers due to local GOP lawsuit.
https://www.cpr.org/2020/05/15/colorado-changes-the-way-coronavirus-deaths-are-reported-to-the-public/
https://www.tampabay.com/news/health/2020/04/29/florida-medical-examiners-were-releasing-coronavirus-death-data-the-state-made-them-stop/
Presenting deaths by day of death does not fool the data hounds. It does show up the variety of delay lags in reporting the numbers that some of us are learning about for the first time. Weeks ago the Florida dashboard the last day always was a partial day of reports. Excess death counts are being used as well. They are another target for manipulation.
This is tenuous time for covidtracking.com to start using CDC data.
Self correction: covidtracking checked out the CDC data and evaluated it. “We will continue compiling and publishing data as we have been” https://twitter.com/COVID19Tracking/status/1262571274317914113
Colorado is just separating those who die with Covid from those who of Covid
There was a case where a local coroner labeled the cause of death alcohol poisoning (the labs came back at almost twice the lethal level) and the state changed the cause of death to Covid.
The coroner make a stink, and Republicans threatened the typical “lock em ups” and the tracking was changed to account for both types of Covid related deaths.
Makes sense to me, I can’t say the same for Florida.
Makes sense to me, also.
Johns Hopkins added an exclusive US map a few weeks ago. It also goes granular and if you know all the pertinent counties in florida, it has rankings by counties on the left side. I would trust this over anything Florida whips up.
https://coronavirus.jhu.edu/us-map
My spouse really likes that particular tracker, can quickly glance trends for different states/countries, too.
per Eureka @2:19
and the JHU site show per capita rates, too.
The data is poor across the board. It is mostly due to inadequate legacy process and systems. That being said, this seems like a best effort:
https://www.rand.org/pubs/tools/TLA173-1/tool.html#current-impact
“We don’t do body counts.”
It’s beginning. Collateral damage does not “count” and therefore will not be counted. Gets in the way. That’s been established in foreign conflicts and will now be applied domestically. At least until November.
The first casualty in war is truth.
Science, and getting nearer to things “objectively” true, are anathema to those who find democratic processes inconvenient for their own agenda. It’s pathological and not planned, per se. It emerges organically in abusive relationships and finds social niches where it can fester and grow. Faith is the debased currency.
Markets and the Electoral College are all that “count.” All the other numbers need to be as plausibly deniable (pliable?) as possible–to serve the needs of power.
Bury votes and voters. Deny climate science. Don’t count collateral damage.
The market won’t hold up. Will the Electoral College?
We’re gonna need the German word for the morally-aesthetically repugnant sight of those round, wasted masks. It is a roundness which offends my sensorium.
On an unrelated but current-events note, I am so sorry for the folks of Midland, MI and environs:
https://twitter.com/search?q=Midland&src=tren
Jim, I love your work. It’s nice to have a scientist talking about this and not a politician.
I have seen several people posting on Twitter reposting this nonsense from the NY Post about Woodstock being held in “the of a pandemic”
https://nypost.com/2020/05/16/why-life-went-on-as-normal-during-the-killer-pandemic-of-1969/
I was 15 when Woodstock took place and I remember the Hong Kong flu in the winter of 1968.
But I’ve been frustrated by trying to find accurate information about the death toll. Several sources are saying it was 100,000 deaths in the US and others that it was 33,800. The 33,800 figure is sometimes given for the Summer 1968-March 1969 data. Since the pandemic came in two waves, it would be possible that the 100,000 was both waves combined.
However, this source https://academic.oup.com/jid/article/192/2/233/856805
Says 70% of the deaths were in the first wave in the US. Dividing 33,800 by 0.7 gives 48,285 which is nowhere near 100,000. The CDC mentions the 100,000 figure, but it does so in a summary and does not give monthly totals.
Do you know of any sources that would reconcile these statistics?
Thanks
[FYI, first URL edited to remove tracking. /~Rayne]
Thanks.
Wow. I had missed the whole controversy about Woodstock and flu.
After a quick look, I see Reuters did a fact check on it, getting it wrong at first and then nailing it on their second revision, which is here. That links to a key scientific publication, which shows in its Figure 1 that at the time of Woodstock, the pandemic first wave had died down and was only active in Australia. It was virtually gone in the US between March and November of 1969. That publication is the one you link and the Reuters fact check does a good job of explaining how Figure 1 debunks the Woodstock during a pandemic claim.
I was a freshman in college during the first wave – I recall one person on the dorm floor where I lived got it (though I misremember it as “Saigon flu”).
Few things would bring me more delight than seeing this doorknob DeSantis marched off to prison in an orange jumpsuit! That is, except Disinfectant Donnie being marched off to prison just ahead of him!
I know that Woodstock was between waves, but I cannot get a handle on the supposed 100,000 deaths. Wave 1 ended in March with 33,800 deaths. Schools were closed when it peaked, not by government fiat, but by the Christmas holiday closure, ergo it caused fewer deaths than expected given how infectious it was.
CDC has a summary paragraph quoting the 100k number in this sentence:
“ The estimated number of deaths was 1 million worldwide and about 100,000 in the United States. ”
https://www.cdc.gov/flu/pandemic-resources/1968-pandemic.html
This site has the number as 33,800:
https://www.tandfonline.com/doi/pdf/10.4161/hv.8.1.17622“The Hong Kong influenza pandemic broke out in Hong Kong in 1968. The first cases in the US were detected as early as September of that year, but influenza did not become wide- spread in the US until December. Deaths from the H3N2 strain peaked in December 1968 and January 1969, subjects over the age of 65 accounting for most deaths. The number of deaths from September 1968 to March 1969 was 33,800‘
Another source:
https://www.globalsecurity.org/security/ops/hsc-scen-3_pandemic-1968.htm
“The number of deaths between September 1968 and March 1969 for this pandemic was 33,800, making it the mildest pandemic in the 20th century.“
This next site says 70% of the deaths were in the first wave:
https://academic.oup.com/jid/article/192/2/233/856805
If that’s the case, that 70% were in the first wave, then 33,800 /0.7 = 48,286 should be the total for the two waves combined.
So, I cannot find where the 100k number comes from. I saw sole rolling 5-year numbers that might capture it if 1968-1972 were summed, but I’m not sure if that’s the explanation or what.
I just thought you might know of an authoritative source.
p.s. there’s a guy on twitter called “ethical skeptic.” His writings seem like what Popper called scientism, a lot of fancy-sounding words and cute graphs. His writing indicates that he has an agenda. But I don’t know the subject matter well enough to see if maybe he has a valid point or not. Thought you might find him interesting, even if only to debunk him.
Sorry, no, I don’t have any info on that part of the question.
Oops, I was replying to Jim’s reply to me. Didn’t mean to get it out of order.
The spike on May 4th most likely occurred because it was a Monday, with reporting from Saturday and Sunday being “caught up on.” This is the case on every Coronavirus tracker I follow – the graph drops a little on Saturday and Sunday, then comes back big on Monday or Tuesday.
Nonetheless, Florida’s incorrect tracking is a scandal!
Except that if they had already switched to deaths counting on the day of death rather than the day of the report that wouldn’t explain it.
Saw Jim tweeted about the difficulty of donning masks with hearing aids (& glasses).
We here use ear savers / mask strap holders:
https://duckduckgo.com/?q=ear+saver&t=hj&iax=images&ia=images
But worn below the ears/ slung across occiput, so no problems with other stuff our ears have to hold — getting the mask on (and off) is an independent procedure. It doesn’t seem to torque the masks out of place, also no ear pain.
hiding the truth is what they do in dictatorships. step by step trump and his gop supporters have been turning the U.S.A. into a failed state. Hiding the figures for the number of COVID deaths won’t help trump. people will figure out how many died by the number of their friends who died. Trump and his ilk can not hide the body bags.
Back in the AIDS epidemic days, I asked an acquaintance how they were doing. They said not well, they had been to a 100 funerals so far that year and it was Oct. or Nov. as I recall. (that was in Vancouver, B.C.) The figures for deaths were out there, but it really struck home, that some one I knew had been to a 100 funerals. It will become the same with COIVD, not that there will be traditional funerals, but people will notify friends of who died and eventually there wont’ be any one who hasn’t heard of some one they know who died. Trump needs to understand this isn’t an arranged “reality” show. this is real life and the truth will come out. Florida will be no different. Even if the GOP tries to hide the number of deaths there will be people wo start tracking the numbers by checking with individual funeral homes, hospitals, community groups, etc. You can not hide the truth.