How do I know? Because the number of counterintuitive findings is just skyrocketing! How do I know that? Because mentions of the phrase “counterintuitive finding” in political science journals have increased sharply.
The 1950s and 1960s were pretty boring. Lots of intuitive findings, it seems. But then, wow! There were 213 counterintuitive findings in the 1990s alone. And there are already 281 since 2000, and we’re not even out of the decade yet. Nor does this count include all the journal volumes from this decade that are not yet archived at JSTOR.
No doubt my intuition will become even more wrong in the years ahead.
[ADDENDUM: See a revised version of this graph and further discussion here.]