Models failed. Miserably.
The IHME models the CDC used as their primary source to justify the destruction of President Trump’s beautiful economy turned out to be no better than guesses from three drunks at a bar. The IHME factored in strict mitigation (lockdown), but they still over estimated everything by factors as high as 40.
That models were unbelievably bad is not open to debate. The models were hideously bad.
But why?
As we said, the national lockdown was not the reason, because the models assumed a national lockdown. The model-makers were not math illiterates as far as we can tell.
One possibility, and to me the likeliest possibility, is that Coronavirus was spreading in the US long before the CDC started testing for it. Perhaps many millions of Americans have already been exposed.
There is new evidence out of Massachusetts that makes the early herd immunity theory more likely.
If millions of American have already contracted the Coronavirus in December and January, we either got over it or died. Those who died of flu or pneumonia during those months might have been misdiagnosed. Those who survived with mild or no symptoms would probably be immune. This mass immunity would explain, in part, why the IHME and other models were so woefully wrong.
This new evidence supporting the herd immunity theory was found in waste water in Massachusetts reported on New York Post.
Researchers from Biobot Analytics and MIT studied waste water from a Massachusetts municipality collected in late March. The samples contained particles of Coronavirus. At the time, Massachusetts had just 446 confirmed cases, but the concentration of these virus parts suggested an infection number much higher: 2,300 people.
That means that, as of late March, only 20 percent of cases had been positively diagnosed in Massachusetts.
The discrepancy between confirmed cases and actual cases could be attributed to testing, but that discrepancy still exists. Even with the Coronavirus test widely available, people with no symptoms are unlikely to get the test, and some people with mild symptoms choose to ride it out rather than have themselves tested. Which means the number of actual cases of Coronavirus in the United States could be five-times higher than the confirmed count of about 500,000. That would mean 2.5 million actual cases since testing began.
But the active Coronavirus test does not test for antibodies. Someone who tests negative for Coronavirus, and 2 million have already tested negative, could have had the disease and gotten over it before they were tested. (With lung infections, symptoms often persist after the body has beaten the germ.)
So, if we take the number of people have been test, 2.4 million, assume 50 percent were already recovering from the virus at the time of testing, we’re now at 12.5 million who have or have had the disease. And we’re still not counting those who have had the virus but were asymptomatic.
According to Dr. Anthony Fauci, the estimate of asymptomatic cases in the US is between 25 percent and 50 percent. If there have been 500,000 people diagnosed with the disease, there are between 250,000 and 500,000 asymptomatic people walking around right now. Take the median of 375,000, multiply by 5, and that’s another 1.9 million for a running total of 14.4 million active cases.
And those 14.4 million cases occurred after mid-February, when large-scale testing began. If the virus arrived earlier, like mid-December, we might be able to double that number to about 30 million total cases, or 8 percent of the US population.
This would be very good news. It means the infection is less deadly than presumed. It means many people are safe to go back to work, especially in large urban areas infectious diseases spread more quickly.
The downside is that people in places with lower numbers have lower rate of immunity and could get hit later.
The irony of the situation this: the CDC, state health offices, and the media have conditioned us to believe that the number of positive tests is the most important number, and that we should hope that number remains low. That conditioning could be completely wrong.
We should be hoping that something like 80 or 90 million Americans have already been infected and 98 percent have already fought off the infection. In that case, not only is Coronavirus far less dangerous than feared, it has almost run its course, at least for this season.
I’m going to hope I’m right.
UPDATE: Antibody testing in Chicago shows 30 to 50 percent of population already immune to Coronavirus. The Gateway Pundit has more.
ANALYSIS CONFIRMED Two researchers have confirmed my simple calculations presented in this post. At least 28 million Americans have already gotten over Coronavirus. The Economist has more.
Our work (@Alex_Washburne and I) on COVID-19 has been featured in @TheEconomist. I am extremely impressed by @DanRosenheck who wrote the piece.
— Justin Silverman (@inschool4life) April 8, 2020
Key points: We find strong evidence that COVID-19 is widespread (>28 million) in the US but don't panic. https://t.co/sk8FSUhakD