A health policy for 330 million based on a massive data error committed among 11 authors and further reviewed by top-tier scientists for presentation to the most knowledgeable experts in the field. Who accepted it without noting the obvious errors.

From Essential Terms of the Authority Crisis by Matt Shapiro.  The subheading is Through no small amount of dishonesty and incompetence, institutions of authority are crumbling and the nature of expertise is changing.

Reading this account all I can think of is Institutional Ineptitude.  Sure, malevolence and ideological mayhem are probably in there somewhere, but the main ingredient seems to have to be ineptitude.

This last week, the CDC held their ACIP meeting to discuss whether or not they should recommend the COVID vaccines for children 6 months to 5 years old. While presenting on the danger of the virus for children, a slide was shown claiming that COVID presented as one of the leading causes of death for children.

Click to enlarge.

This should be rather astonishing as child death from Covid has been near zero in virtually all countries ever since the beginning of the pandemic.  Granted that children have very low death rates anyway but for Covid to be the fifth cause of death seems deeply improbable given quite detailed and reliable data from such countries as Israel, Denmark, Sweden, Britain, etc..

Kelley, who runs covid-georgia.com, saw this slide and immediately knew it was false. She has been tracking COVID data in excruciating detail in Georgia since the beginning of the pandemic and has recently become an expert on the CDC’s pediatric death data simply because it was such a disaster and she wanted to get down to the truth of the matter.

This slide above is no small error. Not only did it count the wrong number for pediatric COVID deaths, it compared all pediatric COVID deaths in a 26-month period to annualized deaths from other causes. This is a massive data error, and yet it persisted through a supposedly rigorous data check from 11 authors and was selected by top-tier scientists for their landmark presentation to the most knowledgeable experts in the field.

No one in any of these meetings recognized this error. This slide was presented uncritically to the nation’s top doctors and epidemiologists who are in charge of setting the national policy on COVID vaccines for children and no one even noticed it. It was spread uncritically by dozens more experts, including a former Surgeon General of the United States.

And this error was caught by a woman who tweets using just her first name only and runs fact-checker on the world’s most eminent scientists in her free time.

On the one hand, I’m delighted that what Kelley does makes an impact. It looks like her persistence will result in a re-evaluation of the paper from which this chart was taken. In that way, the system works. But the system utterly failed before it worked and it’s only working now because an internet rando is more knowledgeable and paying closer attention than our top scientists and doctors.

On the one hand this is a shocking indictment of the ineptitude and untrustworthiness of our top epistemic, epidemiological, and public health institutions.  They simply cannot be trusted to get anything right.

On the other hand, I love this country where your average citizen pseudonymously tweeting and blogging in Georgia can point out that the Epistemic Emperors have not a stitch of clothing on.  To give Kelley of Georgia her Twitter due, here is her account.  

In the last two years, we’ve watched this story play out with alarming frequency. High-profile experts who are running policy for the nation (or even the world) show themselves to be woefully uninformed in their field of expertise. Over the last two years, a huge portion of this very newsletter has been about looking at nationwide narratives concerning COVID data and asking if they are actually true.

I’ve started asking myself: Why are we continuously playing whack-a-mole with bad policy drawn from poor science? It is impossible to do this forever, and for many people, it’s impossible to do this at all. Most people don’t know where to go to get the data and they wouldn’t know what to do with it if they got it.

This is the core of the authority crisis. After so many devastating and public failures at the highest levels of expertise, it seems untenable to give them credence simply based on their credentials and institutional positions.

Indeed.  And it is hard not to notice that those most prone to grave epistemic errors are also those most confident in their own wisdom, accuracy and goodness.  And also the most inclined towards authoritarianism.  The two years of the pandemic when public health experts have gotten almost everything wrong from the beginning has been bad enough.  Far worse has the official turn towards mandating and censorship.  

Voting out mandating censoring authoritarians is not really an epistemic solution but it likely will clean up the communal pond of knowledge quite significantly.  

UPDATE:  Another piece on institutional proclivities towards affirmative error making:  Never Trust A Number by Steve.  The subheading is They're shortcuts to understanding, and there are no shortcuts.