My response to Donald Trump’s Axios interview
If you have not yet viewed the interview Donald Trump gave to Axios journalist Jonathan Swan, here it is in full:
If you prefer, you can read the transcript here, although you won’t get the full flavor without watching the video. I won’t respond to every assertion made in the interview, but I want to respond to each of the COVID-related claims. The first claim he makes is that he called for a ban on travel to and from China in January, and that’s largely true. He announced his plan on January 31, making it effective February 2. Even with the exceptions for Hong Kong and 10 other destinations, and the fact that US citizens could still travel as long as they observed a quarantine upon their return, this was probably a prudent decision. If the entire world had shut down travel quickly, the quick spread throughout the world would probably have been significantly lessened. But with all the exceptions and the fact that people were still allowed to come into the country as usual from other destinations that, it turned out, already had cases of the virus, it’s very unlikely that it had a significant impact on the number of cases in the US, but that’s just Trump hyperbole. He’s right that he called fairly early for a ban on travel from China, and he’s also right that the World Health Organization and others were, at the time, against the ban. His point about Europe is also largely true. He imposed a thirty-day ban starting March 13 on travel of travelers from the Schengen zone, but excluded the United Kingdom, which, according to Oxford University’s Our World in Data, was already experiencing about 400 cases per day at the time of the ban, and, since it was known to be three weeks behind Italy, was likely to shortly surpass its Schengen-zone neighbors. As a matter of fact, it became the country in Europe with the highest number of daily cases on April 11:
Moreover, the ban applied only to foreign nationals, not US citizens or permanent residents, so, for example, my husband and I, who have lived in London for over 3 years, would not be affected by the ban even if it included the UK. This might make political sense, but it didn’t make scientific sense. Still, since most of Europe is now following suit and banning travel from countries with high rates of infection—the US, of course, being near the top of the list—it does seem like the concept was one that made sense, and the US was earlier than most at implementing it, at least partly due to the WHO, at first, asserting that travel bans wouldn’t seriously inhibit the spread of the virus.
Trump’s next claim is that “some” countries are suffering at proportionately greater rates than the US, which, of course, is true. But which countries? He gets into specifics about Spain, Russia (or Moscow, as he says), China, and Brazil. Well, again according to Oxford, the only one of these countries that surpasses the US in new cases per day, as of August 5, is Brazil:
But maybe he’s right and new cases aren’t a good indicator of “suffering,” since the US is one of the few countries that are testing asymptomatic people. Now we’ll do what Jonathan Swan tried to do in the interview—look at comparative death rates, both cumulative and daily.
In daily deaths, the US is, once again, close on the heels of Brazil, and approximately 4 times the Russian per capita rate. But let’s discount Russia and China as having suspect reporting, as Trump alleges. The US has 18 times the daily per capita death rate of Spain and almost 50 times the daily per capita death rate of Germany. That’s a huge differential, but maybe those other countries had such high numbers earlier that it’s an unfair comparison? OK, let’s look at cumulative deaths per million people since the start of the pandemic. I separated the table and the graph this time so you could more easily see the shape of the curves:
Here, finally, are statistics that corroborate Trump’s picture to a certain extent. We are not the worst. But note that Brazil, which is obviously not doing very well right now, is also lower down in this chart, lower, even, than the US. But when you look at the over-time graph, you get a different picture. Spain and Italy, because of their early death toll, before we knew a lot of what we know now, still have higher cumulative deaths, adjusted for population, than the US and Brazil. But look at the curves. Spain, despite some recent blips associated with reopening, is still largely level since the end of April. Italy’s curve leveled out around June 1, meaning that for both these countries, there have been so few deaths that the increase in deaths per million is not even observable on the graph. In contrast, the slope of Brazil’s curve continues to get steeper, as does, to a slightly lesser extent, the US’s. If everything keeps going the way it has going forward, the per capita deaths in Brazil will surpass the US’s very shortly, and they will both surpass Spain and Italy’s per capita deaths in six to eight weeks. But what of Trump’s contention that “you can’t do that” (compare deaths as a proportion of the population). Instead, he contends, the statistic you should use is deaths per positive cases. Here’s where we start to really go down the rabbit hole. In the interview, Swan agreed that the US had done a lot of testing (perhaps not “the most in the world,” but a lot). In fact, again according to Our World in Data, the US has done between 2 and 4 times the number of tests per person as most European countries:
Even with differences in how tests are counted (some countries count the people tested, some count the number of tests performed even if one person gets multiple tests, and it seem that the US is a hybrid of the two), it’s likely that the US’s testing numbers are higher than other countries. This is partly due to the fact that asymptomatic people are being tested in the US, while in many other countries, limitations on tests and facilities mean that only people with symptoms are being tested. This means that the larger the number of asymptomatic but positive people in the population, the more cases a country will show, even if these are not seriously ill people. That was the president’s objection to comparing numbers of cases. His logic, which is probably right, is that there are many people in all countries who don’t even know they have the virus unless they are tested, so comparing a country that tests all those people with a country that doesn’t and then saying the country with more tests is worse because it has more positives is an unfair comparison.
Let’s say, for example, we have two countries: Testmania and Notestia, each with 100 people, and in each country, exactly 40 people actually have COVID-19, but only 10 people are showing symptoms. In Testmania, in addition to the 10 people with symptoms, 50 additional people are tested, while in Notestia, only the 10 people we already suspected had the virus are tested. Because of the extra testing, Testmania catches a lot of those asymptomatic carriers of the illness, so you end up with something like 25 positives in Testmania and 10 positives in Notestia. In number of cases, Testmania looks much worse than Notestia, but really, they each had 40 people with the virus:
Now let’s say that each country experiences identical deaths. Five of those 10 people who’d originally shown bad enough symptoms to warrant testing died. Here’s where Trump’s contention that you should take deaths as a percentage of tests is in direct opposition to his claim that the number of US cases is inflated (or, more accurately, the rest of the world’s number of cases is under reported) because of more testing. In my two countries, I think we can all agree that the same percentage of people died, 5 out of the hundred, or 5%. This is the statistic that Jonathan Swan was using, and it seems like an accurate picture of identical countries with an identical experience. But what Trump is proposing is that we instead compare deaths as a percentage not of population, but of positive tests, or cases. What this would do is give us a rate of 5 deaths over 25 cases in country A, or about 20%. In country B, though, only 10 tests were conducted, so that rate would soar to 5/10, or 50%. We have two identical countries with identical COVID experience, yet one has a death rate that’s twice as high as the other’s—just because of the way the numbers are calculated. There is no way that this is an accurate or informative comparison. Of course, in the real world, we won’t have identical countries with identical experiences, which makes it essential to have a reasonable way to compare them, and that way is deaths as a percentage of the population. But Trump dismissed Swan’s reasonable comparison with a “You can’t do that,” and Swan finally let it drop.
This brings me to style. If you are already skeptical of or opposed to Trump and everything that comes out of his mouth, you probably saw, as I did, a man fumbling with some silly PowerPoint charts, asserting things that were either completely untrue or taken out of context. But if we want him to leave the White House in January, we should be pretty concerned right now. One of Trump’s techniques is tactical interruption. He may be fumbling and short of facts, but he knows how to cut the interviewer off when one of his points is refuted. For example, every time Swan brought up the current case rate (or some other statistic that would have painted a less-than-rosy picture of how the US is doing), Trump cut him off and just bulldozed on with his original, and generally unsupported, point.
The second thing he did quite well is evident in his discussion about the number of COVID-19 cases. He starts out with a valid assertion: The US has done much more testing than most other countries. According to Johns Hopkins, only Hong Kong has performed more tests per capita, and the US has performed almost three times the number of tests per capita as Germany. And testing is good. If we randomly test a population, we’ll be able to figure out what the actual current infection rate is, which will enable us to say with some accuracy what the likelihood of serious illness is for people infected. But everything Trump says after that start is refuted by the data. If, however, you were watching the interview as a Trump supporter, as someone with a predisposition to believe that the “MSM” (mainstream media, except for Fox) is out to get him, and as someone who doesn’t quite understand all the seemingly contradictory statistics being thrown around, you probably came away thinking that Swan was trying to pull a fast one by using country populations instead of positive cases in order to make the US look bad even though the US was actually doing better. Swan was a pit bull—he kept asking why Trump thought cases were a better denominator to use. And Trump did what Trump does: uttered a few half sentences, “Look, it’s there—,” acted superior, “You can’t do that,” and ended up by casting vague aspersions on South Korea’s numbers, leaving his supporters with the idea that 1) the way the statistics are being used does not paint an accurate picture, 2) countries other than the US are not trustworthy, and 3) he is going to great pains not to cast aspersions (“I won’t get into that because I have a very good relationship with [South Korea].” In the meantime, what he is actually doing is using statistics inaccurately and inconsistently—and casting aspersions all over the place, on US governors who haven’t done the right thing, on the “fake news” media who are treating him unfairly, and on other countries that are not doing as well as the US but are being protected by the media.
And this is why I think Joe Biden supporters should be worried. Jonathan Swan is a young, well-spoken, and well-prepared interviewer. He is also in a position of some power over the president because he is the interrogator and Trump is the interrogatee. Yet, several times, Trump talked over him and managed to divert the conversation to his advantage, even though a lot of what he said was devoid of either content or accuracy. Also, remember back to 2016. Hillary Clinton is not easily intimidated, yet his bullying style completely threw her. Biden was certainly not the best choice to run against Trump. He has too much of his own baggage, with his frequent gaffes, his son’s job in Ukrainian energy, and nagging questions about his inappropriate behavior towards women. Oh, come on, you are probably saying, Trump makes just as many gaffes, his family is far too involved in official business, and his attitude towards women is arguable quite a bit cruder than Biden’s bone-headedness. Yes, and Biden will take the advice of his people, he won’t tweet, and he won’t do anything more embarrassing than misremembering a person’s name. But about half the US voters voted for Trump in 2016, so we can’t just assume that all those people will suddenly jump ship. As a matter of fact, because the left has been so vocally critical of him, I know some people who did not vote for him in 2016 but are voting for him in 2020. I, for one, will be holding my breath on October 22, because it’s only 12 days before the election and, although Trump will never win with sound arguments, he very well may win the way he has done before—bullying, interrupting, and painting himself as the outsider, even after four years in office.