The median has to be lower than the average for “books read” because of that skew (and because there is no such thing as reading a negative amount of books).

It’s been beat to death, but people really want the median here. You’d line up every American from least books read to most books read. That 150MM person will have 50% of Americans who read more the prev year and 50% who read less. How many did that person read?

Like you I would assume it’s 0 or 1, but apparently someone linked a survey saying it was 4. Regardless that’s the number people are curious about.

The mean basically only would give the info of how many books were read in America along with how many people are in America (which we already know). It’s pretty meaningless at an individual level.

Maybe I’m missing something in your argument but I’d argue the reverse: OP absolutely meant to ask about the median American not the mean number of books.

When they say average (or typical American), I think this is exactly what they mean. Some people read a lot, some don’t read anything, find someone in the middle - how many books are they reading? If I was in a plane or at the DMV and tried to strike up a conversation about books, what is the chance they could participate?

That’s how I interpreted their question, which is all about the median/statistical quantiles.