A common error people make when they see an old vs. young split in opinion is assuming that it must be an age effect rather than a cohort effect. Thank you for avoiding that mistake by noticing that it could be either! (Maybe I could use that as a rationality litmus test, ha ha.)

Actually, as for this particular issue, a cohort effect was what I used to consider obvious, and I didn't hypothesize an age effect until I looked for a long-term trend and failed to see one. (Maybe I haven't looked in the right places, though.)