The Flynn effect mostly appears in only certain puzzle tests. There is no Flynn effect in vocab and numeracy, despite far longer hours spent at school. This indicates to me that we have gotten better at certain abstract puzzles, perhaps because the rising prominence of IQ tests made people study those puzzles more. The Flynn effect only proves that those kind of puzzle tests are very flawed at measuring innate aptitude.

I think that your objection is, rather, an indictment of schools and the surrounding bureaucracy.

Anyway, do people study such puzzles? For such a consistent, widespread rise, you'd expect it to be pretty obvious that they do, but I'm not sure.

Well, didn't you play with brain teasers and abstract shape puzzles and such as a child? Or edutainment software? Studying doesn't have to mean nose-to-the-grindstone.

I'm in my mid-30s. If there was anything we'd recognize as "edutainment" software in the 70s and early 80s, we couldn't afford it. Anyway, I have an average-software-engineer IQ, more or less, on tests in the 80s, and I can't remember ever doing much brain teaser or abstract shape puzzles -- those weren't interesting to me. I spent every waking hour I could reading past age seven or eight.

I think it's more likely the Flynn effect is a reflection of our getting better at taking tests.