It seems like most of not all news I read about COVID deaths, and cases in general, reports them in absolute numbers rather than per capita numbers. Is there a scientific or otherwise sensible reason for this?
Much is made of the US’s extraordinarily high number of cases and deaths, but per capita numbers put us further down in the top ten, after UK, Canada, and a few other countries. Given population differences, wouldn’t it make more sense to work with per capita numbers?
Any death is a tragedy, but is this simply a way to make the reporting more dramatic, or is the absolute number significant in a way I’m not getting?
It’s GQ, and perhaps it should be moved, but yeah, what sells more newspapers or gets more clicks? U.S #8 per capita in death or TENS OF THOUSANDS DYING!!! ?
There is a chart on this webpage with statistics for nearly every country in the world. It says that the U.S. is ninth in the number of Coronavirus deaths per capita (as of today). It says that the U.S. is eleventh in the number of Coronavirus cases per capita (as of today):
You know, if old PT was to unfortunately suffer demise due to COVD-19 it’d be merely a parting act of self indulgence that my shuffling off this mortal coil be counted amongst the published statistics at a value of one rather than 1/25,000,000 per capital thingy. After all it’d be an intensely personal matter.
I think it just makes more sense to people. Saying that something happened to 500,000 people in this country and 200,000 people in that country is easier to grasp than talking about those same numbers as 10 people per 100,000 in one country and 14 people per 100,000 in another country.
Also, keep in mind that a lot of people either don’t know that difference between ‘total deaths’, ‘death rate’, ‘deaths per capita’ and other ways of working with numbers, or even if they understand it, are going to mentally check out when they see too many numbers.
Back in March I saw an anti-quarantine/just the flu image going around. It showed the ‘death rate’ for several countries comparing the ‘death rate’ for 2020 Covid deaths compared to the 2016-2017 flu. Ignoring the fact that it compared a year of the flu to 3 months of Covid, the numbers weren’t the ‘death rate’, they were total deaths, as illustrated by most of the numbers being well over 100. If you looked at the numbers, the covid ‘death rate’ was about a quarter of the flu ‘death rate’, which made sense since they only used about a quarter of the time.
At best, the graph was incorrectly labeled by mistake. At worst, it was done on purpose knowing people would just take the headline ‘covid death rate much lower than seasonal flu’. In either case, it was a good example for me for when people say numbers don’t lie. Maybe they don’t, but they can be manipulated to present data in a different way.
And looking at what shows up as the #1 “per capita” country is educational in how, while per capita numbers can be better for some purposes, they can also be used to distort things. San Marino is #1 in deaths/million population, with 1,238 deaths per million! That’s Huge! Until you look at the absolute numbers, where they’ve had 42 deaths in total. The per capita number looks really bad, until you realize there’s only 33,928 people living there.
All statistics can be manipulated to tell a story in a way that you prefer. For instance, have you noticed that when Trump brags that the US has done “more tests than anyone”, he’s referring to absolute numbers? Per capita the US is only 28th in testing. You really need to look at all the numbers to figure out what’s going on.
They only worry about “per capita” when it makes them look good. When it doesn’t, suddenly absolute numbers are just fine.
True, but it DOES alter its rate to account for how densely those people are packed. Maybe what we need is a chart showing the number of cases per capita, and the lines are sorted by population density.
But the point is that you want to use those statistics to see what countries are doing a better or worse job of dealing with the epidemic. Even if the US were doing a great job, you’d still expect us to have more cases and more fatalities than most countries, just because we have a higher population than most countries.
[Moderating]
Oh, and moving to Quarantine.
per capita is the only way of measuring deaths based on the population size of a country. or for that matter, a state within.
Texas has an extremely low per capita count of 68 yet the current news makes it sound like New York which has a per capita count of 1,590. 23 times more people died per capita in New York than Texas.
Sure, but how are we going to use the information? Per capita is more actionable for the public. I’d like to see how the rate of new cases or hospitalizations looks on a per capita basis. I think this is the criterion that should dictate whether I stay home, or only do necessary errands, or return to the workplace, or hang out at Barnes&Noble and Cheesecake Factory (which is a very pleasant indulgence (I get the Cobb salad and read Sky&Telescope magazine)).
You are correct, I apologize. I did a quick bit of research before I posted and found this:
…in which I thought the first table was per capita deaths but was in fact case-to-death ratio. According to that table Canada has a higher case-to-death ratio than the US, but our case and death per capita numbers are definitely higher.
This pretty much encapsulates what I’m wondering about, and other posters have said something similar. If we’re talking about how effective a country’s anti-COVID action have been, doesn’t it make more sense to look at per capita numbers? I really hate Donald Trump and his entire performance during the pandemic has been offensive to me as a human being, but when news reports cite the absolute numbers of cases and deaths as a means to show how bad the US response has been, I get little question marks over my head. It feels like a convenient way to disparage DJT, but it also seems to warp our sense of whether we are successfully fighting the disease.
So I was hoping there might be a scientific/statistics reasoning behind citing absolute numbers instead of per capita numbers.
This stems from the fact that the vast majority of our cases (and deaths) were due to outbreaks in Long Term Care facilities. As of May 18, this article in the Washington Post shows 81% of cases are LTC related. We failed our seniors horribly on this one, hopefully it will causes structural changes in the system.
Of course it’s signficant. It’s the actual size of the outbreak that you are reporting on; the actual number of people dying.
The per capita figures are relevant only if you regard coronavirus as a sort of international competition, and you want to know how badly country A is affected relative to country B, and you need some reasonable basis for a comparison. And of course there may be a context in which that’s appropriate; e.g. you are writing an article about how well, or how badly, Country A has handled the outbreak relative to country B. But the default position, where you are just reporting on the fact that there is an outbreak and on how big it is, it’s the absolute numbers that are the most pertinent and the most informative.
Yeah, someone asked the OP’s question during an NPR segment not long ago, and the expert’s answer was something not unlike this, IIRC.
My analogy, not the NPR expert’s: Let’s say there are plane crashes in a large country and a small country on the same day. The former crash kills 500 and the latter kills 100, but on a per-capita basis far fewer people died in the first crash. Do you think the most relevant statistic is the per capita one?