[mennis16]

I have a question about the abstract for anyone that looked deeper at the paper- does the following statement take into account rejected student stats or only those of accepted students? Because it seems plausible to me that white students "on the cusp" are more likely to get rejected to "make room" for ALDCs. Meaning that removing this preference may actually benefit a typical white applicant more than others. I could of course also imagine that not being the case though.

"Removing preferences for athletes and legacies would significantly alter the racial distribution of admitted students, with the share of white admits falling and all other groups rising or remaining unchanged."

I do recall the popular line about how everyone at orientation could be replaced with a new cohort that has similar academic profile, which implies that there is quite a bit of judgement call going on.

Overall, it is unsurprising yet disappointing that they estimate ~3/4 of ALDCs would not get admitted without that status. This system is probably a big reason that most elite schools have ~15% of their students from the top 1%, while MIT is closer to 5%. MIT definitely does not do legacies and of course gives little weight to athletics. I am not sure about children of faculty though.

[cameldrv]

The numbers wind up pretty grim. 43% of white students are legacy or athletic, 3/4 of those are otherwise inadmissible, so that's 32% of white students ALDC and inadmissible otherwise. Harvard is 44% white, so it's 30% (academically) competitively chosen white. That compares to 58% of the overall college population in the U.S.

[mennis16]

But at whose expense are ALDCs chosen? I suppose we cannot know, but are there rejected non-ALDC white students that could fill those spots who the model would consider deserving to be there? FWIW MIT is ~40% white without AL(DC?) [0], so that is closer to H's current distribution.

The whole admissions situation is of course an opaque, handwavy, crapshoot. But I am just curious in practice what would happen to the distribution if Harvard decided to drop say legacy status.

[0] https://mitadmissions.org/apply/process/profile/

[cameldrv]

I can't really say except that obviously non-ALDC white students are a much smaller proportion of Harvard's class than undergraduate students overall. My guess is that they have target numbers for diversity. Since the overwhelming majority the ALDCs are, due to economic and historical factors, white, it is quite difficult to get into Harvard if you're white and your parents didn't go to Harvard, dont' work at Harvard, aren't rich enough to make a big donation, or aren't moderately wealthy and savvy enough to get you into some obscure Harvardy sport like rowing.