Fun with COVID statistics

 Two apparent facts about COVID-related statistics that I’ve read about in the last few days.

1) Average wages apparently rose in 2020, due to COVID – but the main reason was that lower wage earners disproportionately lost their jobs. So, with them taken disproportionately out of the workforce, those who still had jobs had higher average wages – wholly without regard to any wage gains from jobs that survived.

2) In the next ten years, it’s apparently expected that the surviving American population will be on average healthier, and have longer average expected lifespans, than if COVID hadn’t happened. But the reason is that older people who were less healthy than the average, and yet had life expectancies of, say, 5 to 10 years, were disproportionately killed now.

This makes the great human loss from these people’s losing years that they otherwise would have had look as if it were actually a gain. After all, if the population were the same and life expectancies increased, that would clearly be good. But just as people’s economic welfare declines if they lose their low-wage jobs – but this misleadingly causes average wages to increase – thus prematurely killing off the less-healthy makes it misleadingly seem as if constant-population lives were lengthening.

Mark Twain’s famous (and if anything over-familiar) line about “lies, damned lies, and statistics” draws mainly, at least as I have always interpreted it, on the opportunities those citing statistics may have to choose the deliberately misleading. No doubt that will go on here as well. But it’s also the case that we can in good faith mislead ourselves with (in these two cases) average figures drawn from a sub-group consisting of the more fortunate, when what we actually (do or should) care about is the larger group, including the less fortunate.