These are nit higher average SAT scores across the population. These are the average SAT scores for these colleges.
So the reason is far far simpler than trying to calculate proportion of people from different ethnic groups of SES.
There are two actual reasons. The first is that with more applicants, and a lower acceptance rate, the colleges can select students with high SAT scores.
However, there is more than that. “Holistic” meant that students of specific SES, ethnicities, race, and high schools were preferred, and college-bound students knew it. So a student for the right high school and the right family would apply to, say, Amherst, even if they didn’t have the best test scores.
So, in 1960, acceptance rates to Amherst of students with verbal SAT scores of over 700 was 36%, and for students with SAT scores 600-700 was 25%. In 1974, it was 25% and 16%, as the SAT scores of applicants started climbing.
That meant that a large number of students were being rejected while many students with lower SAT scores were being accepted.
Interestingly, Amherst seems to have held similar acceptance rates for a while. In 1960 it was around 20%, and in 1970 it was almost the same.
I just spoke of AMherst, because they have the data available:
1960:
https://acdc.amherst.edu/explore/asc:927535/asc:927547
1970:
https://acdc.amherst.edu/explore/asc:927329/asc:927330