12 July 2020

COVID-19 as an Intelligence Test

COVID is a weird kind of intelligence test. Probabilities and lags obscure cause and effect.

There are some folks who regularly rush out to buy a lottery ticket when the jackpot hits $100 million who are dismissing the need for a mask. The odds of winning the jackpot are 1 in 100 million; the odds of dying from COVID once you've contracted it are 1 in 100. (And if you are over 70, those odds are more like 1 in 10.) There is an old quip that the lottery is a tax on people who are bad at math. Bad at math might also describe one of the high-risk groups for COVID.

Here's an interesting and important article for anyone living in a pandemic. Derek Thompson explores various explanations for why we're currently seeing a spike in cases but not in deaths.

1. Deaths lag cases. There is about a month between when someone contracts the coronavirus and dies of COVID. 
Image may contain: text
2. Expanded testing is finding more cases, milder cases, and earlier cases.
3. The typical COVID-19 patient is getting younger.
4. Hospitalized patients are dying less frequently, even without a home-run treatment. (This falls into the category of, We're getting better at treating COVID as we learn more. One more reason to - at a minimum - procrastinate on getting COVID.)
5. Summer might be helping—but probably only a little bit.

One of the reasons that COVID is an intelligence test is that cause and effect are separated by time as well as probability. Of 1,000 people engaging in the exact same activity, only 100 might get it and only 1 might die. These probabilities obscure cause and effect. On top of that, it might take weeks for certain policies or behaviors to result in illness and another month to result in death. That lag, too, further obscures cause and effect. When New York was suffering the worst of hospitalizations and deaths, the behavior of its residents might have been the best. (People were incredibly cautious.) Before that, when there were very, very few cases or deaths, New Yorkers were behaving most dangerously. If you look around now to see what is going on, you won't see any clear cause and effect. If you could collapse cause and effect at the peak of the outbreak in New York, you would see that New York was the safest place and Arizona the most dangerous, even if the data at the time screamed a message exactly opposite of that. Cause whispers. Effect screams.

One of the the reasons that levels of education rises with levels of income is this: people who learn only from personal experience don't learn much. There is far more to learn from studying millions of people than from "just living your own life." A life is too high a price to pay for information and insight, no matter how quotable are your last words.

Derek Thompson's article in the Atlantic is here.

No comments: