THE ROLE OF MISTAKEN BELIEF: Leonhardt and Douthat attempt to do deaths!

FRIDAY, SEPTEMBER 18, 2020

Where do they find these guys?: How many people will die of the virus over the next four months?

We can't exactly tell you! That said, here's a fact-check of something Joe Biden said at last night's town hall forum. It comes from the Washington Post's Fact-Checker site:

“We should expect another 215,000 dead by January. But if we wore a mask, we’d save 100,000 of those lives, doing nothing but that.”

These numbers are on target. Deaths from the novel coronavirus are near 200,000 in the United States, and one influential group of researchers predicts the total will reach 415,000 by 2021.

The projection comes from the Institute for Health Metrics and Evaluation (IHME) at the University of Washington. The IHME model currently forecasts 415,090 U.S. deaths by Jan. 1, 2021.

As we noted yesterday, some modelers project a substantially higher daily/weekly death rate over the next few months.  

In its "most likely" scenario, the IHME has projected an additional 215,000 deaths in the next (roughly) four months. That works out to (roughly) 54,000 deaths per month—close to 2000 deaths per day.

(According to the Washington Post's numbers, we averaged 814.9 deaths per day over the past seven days.)

This projection by the IHME received a blip of coverage when it appeared last week. And, of course, it's just a projection. It could turn out to be wrong.

(More from the Post's fact-check: "Assuming public health mandates were relaxed, the IHME model predicts twice as many deaths before the end of the year: 400,000, for a total of nearly 612,000." In that scenario, we'd be experiencing an average of more than 3000 deaths per day in the coming weeks and months.)

These grisly projections received a minor blip of coverage. That said, our major journalists tend to steer away from reports involving numbers.

For Talking Barbie, math was famously hard. For our most "highly-educated" upper-end journalists, statistics seem to be even harder.

What follows will constitute a highly significant lesson in anthropology—in the actual capabilities of our actual species. Keeping that framework in mind, let's consider what happened when two "highly-educated," upper-end journalists tried to evaluate the coronavirus stewardship of one Donald J. Trump.

David Leonhardt took the first crack at this important topic. He did so in an on-line post, "America's Death Gap," which appeared in the New York Times back on September 1.

By way of background, Leonhardt graduated from Yale in 1994. (He'd prepped at Horace Mann.) During his rise at the New York Times, he's been branded as one of the smart ones.

Now he was trying to evaluate how good a job this country, and especially its president, have done in fighting the virus. He started with an utterly silly framework:

LEONHARDT (9/1/20): America's Death Gap

Here’s a jarring thought experiment: If the United States had done merely an average job of fighting the coronavirus—if the U.S. accounted for the same share of virus deaths as it did global population—how many fewer Americans would have died?

The answer: about 145,000.

That’s a large majority of the country’s 183,000 confirmed coronavirus-related deaths.

No other country looks as bad by this measure. The U.S. accounts for 4 percent of the world’s population, and for 22 percent of confirmed Covid-19 deaths....

That logical framework made little sense. Beyond that, when Leonhardt tried to apply the statistical measure he had chosen, he bungled the assignment. 

"No other country looks as bad by this measure?" We're sorry, but that wasn't true on September 1. As a matter of fact, it still isn't true today!  But let's move through this groaner quickly, so we can move on to Ross Douthat's attempt to analyze this same question:

As he started, Leonhardt made a silly assumption. He assumed that, if the U.S. had done an average job confronting the virus, our deaths to date would match our share of the world's population.

That assumption makes little sense. As Leonhardt later mentioned in passing, the virus arrived in certain nations much earlier than in others. 

Due to patterns of international travel, it moved with relative speed from China into western Europe and into the United States. It arrived much later in other less developed nations. 

In some of the world's less traveled realms, the virus has barely arrived at all.

For that reason, the countries to which the virus traveled first have had more time to pile up a gruesome number of total deaths. This doesn't necessarily mean that they've handled the virus more poorly than anyone else. It simply means that the virus has been active within their boundaries for a longer period.

As of September 1, the United States did have 4 percent of the world's population and 22 percent of the deaths. In part, that reflects the ridiculous way Donald J. Trump (and politically affiliated governors) have chosen, and continue to choose, to react to the virus. 

In part, though, it simply reflects the fact that the virus arrived here early on, from China and Europe both.

On September 1, our share of deaths did outstrip our share of the world's population, by more than a 5-to-1 ratio. But this was also true of other "early arrival" nations, and it remains so today.

"No other country looks as bad" by that measure? At present, Belgium and Spain look worse than we do by this measure (assuming the measure is properly applied) as does the United Kingdom, if only by a tad at this point.  

That is to say, Spain's share of world deaths outstrips its share of world population by a higher ratio than ours. We have a lot more people than Spain does, but their ratio of share of deaths to share of population  is still higher than ours. 

On September 1, that was still true of other major European nations (of Italy, for example). As our nation's high daily death rate continues, we keep passing other nations with respect to this measure. 

That said, ten other nations still outstrip us on this score. That includes Peru, Bolivia, Ecuador and Brazil, where runaway daily/weekly death rates have overtaken the late arrival of the virus on their shores.

"No other country looks as bad?" That wasn't true then and it isn't true now. If Leonhardt's measure is correctly applied, we aren't the worst in the world even now! 

Leonhardt managed to bungle this point because he made a basic blunder in the way he proceeded with his analysis. Anyone except Talking Barbie and her clueless consort, Ken, should be able to see what it was. (Don't be misled by his graphic!)

Leonhardt comes to us straight outta Harvard. He's been branded as one of the smart one at the New York Times, a newspaper which brands itself as the nation's smartest.

It's astounding to think that a journalist with that profile could be so innocent of basic analytical skill. But this is an anthropological study—a study of the capabilities of our floundering species as it really exists, not as it's long been described.

With that framework established again, let's move on to the analysis offered by Ross Douthat.

We'll guess that Douthat is a good, decent person. Almost surely, Leonhardt is too. But that isn't the question at hand, and his profile looks like this:

Douthat graduated from Harvard in 2002. (He'd prepped at Hamden Hall Country Day.) In April 2009, he became a regular New York Times columnist—at the age of 29!

Douthat looked at Leonhardt's attack on our country's performance and he wasn't buying. In a column which appeared in print on September 6, he cited and linked to "America's Death Trap," then offered this:

"I’m not fully convinced by my colleague’s approach."

He proceeded to offer a typical upper-end press corps attempt at analysis.  Where in the world—where on earth!—does the New York Times find these guys?

The answer, of course, is "at Harvard and Yale." But let's not get bogged down there!

How well had Commander Trump performed against the virus as compared to the rest of the world? Bizarrely, Douthat compared our country's performance in fighting the virus to the performance of such "peer nations" as Colombia and  Peru, but not to that of Canada. 

He hopscotched all over the countryside in his choice of statistics, alternating between Total Deaths to Date and Current Daily or Weekly Deaths. Sometimes he switched from deaths to the highly amorphous "cases" and "infection rate" as his unit of measure.

Also, there was a puzzling side trip to "rate of mask usage," a statistic which—or so Douthat said—puts us "right in the middle of the pack" when compared with our "peer nations."

Did we mention the fact that Douthat's collection of "peer nations" includes Colombia, Mexico and Peru, while Canada never gets mentioned? At any rate, out of this puzzling melange came an assessment of Donald J. Trump which differed from that of his Yale-powered colleague:

The commander hasn't achieved greatness is his handling of the virus, Douthat was willing to say. But it seems that he has pretty much turned in an average performance:

DOUTHAT (9/6/20): [T]he peer-country evidence suggests that to take the pre-emptive, creative and draconian steps that might have actually suppressed the virus, and in the process saved that hundred thousand or more extra lives, would have probably required presidential greatness, not merely replacement-level competence. We can say without a doubt that Trump whiffed when this call for greatness came. But distinguishing between Trump’s incompetence and what an average president might have managed is harder, so long as so many peer-country death tolls look like ours.

So assessed Harvard's Douthat—to which we offer this:

Really? Is it really true that "so many peer-country death tolls look like ours?" 

Below, you see some current numbers for the nations we would regard as the most obvious peer nations. Who can look at these numbers and think that our president, and our nation as a whole, have done an average job?

Deaths from Covid-19, September 10-16:

United States: 6,258

United Kingdom: 78

Germany: 30

Canada: 35

Japan: 68

South Korea: 23

Taiwan: 0 

Australia: 46

European Union: 1,325

Those numbers haven't been adjusted for population. But would you say that those other "death tolls" look anything like our? Does our number, in any way, signal anything resembling average performance? 

(Key point: The population of the E.U. is one-third larger than ours.)

Those numbers have changed since Douthat's column appeared on September 6. If anything, our relative standing was somewhat more horrible at that time. (Several nations have been experiencing an uptick in daily deaths.)

Those numbers show where matters stand after 6-8 months of fighting the virus. Roughly speaking, they show where our nation's efforts have left us as compared to a range of peer nations. 

In that chart, we're looking at nations with roughly comparable economies and infrastructures. They're countries where the virus came ashore at roughly the same time.

In the roughly eight months since the virus arrived, does it look like we've done an average job combating it? On what planet would a skilled journalist reach such a conclusion, with some editor cheering him on?

Harvard and Yale and the New York Times stand behind the journalists we've cited. In the ways our culture reaches such judgments, these journalists are "highly educated." They stand at the top of the pile.

The work they did in assessing this question was bungled all the way down. We regard this as an anthropology lesson, and we ask you to keep that in mind. 

Alas! A nation whose ranking journalists display skill levels like these will be a nation routinely driven by false and mistaken belief.  We'll extend our point this afternoon with one more look at George Stephanopoulos, who we've actually met and chatted with on at least one occasion.

He's a high ranking journalist too. He  graduated from Columbia in 1982, ranking number 2 in his class.

This week, Stephanopoulos also whiffed on the data! "Statistics can be boring and hard," anthropologists have frequently said.

This afternoon:  In a bit more detail, the basic rejoinder not offered