A few comments in that thread are guessing that it could be because the 1955 study might not have polled minorities. But they forget that even if that was the c

A few comments in that thread are guessing that it could be because the 1955 study might not have polled minorities. But they forget that even if that was the case, the US was only 11% non-White in 1955: https://en.wikipedia.org/wiki/Historical_racial_and_ethnic_demographics_of_the_United_States https://en.wikipedia.org/wiki/Historical_racial_and_ethnic_d... So even if all of that 11% were to say they self-censored, it would only sum to 13.4% + 11% = 24.4%, compared to 2019's 40% self-censorship.