Certainly by those who apply there, vs. the number of women applying elsewhere. 
That’s interesting - and I wouldn’t have guessed that for those colleges.
Are those the total admission/enrollment figures from section C1 in the CDS?
At first I wondered whether co-ed colleges generally have a lower yield (because I looked at Barnard’s 66%) - but then Smith wouldn’t make sense. Or is that mostly a function of how many people commit to ED (and are admitted) - which will skew an “overall applications” calculation/comparison.