If Harvard admitted students based solely on the applicant scoring in the top decile of an "academic index" (=test scores + HS performance), the racial/ethnic composition of its freshman class would be expected to change as follows:
Asian +108% pic.twitter.com/goSlpHpRaf
— Timofey Pnin (@pnin1957) October 15, 2018
Pnin is calculating that if Harvard simply selected admittees randomly among the top 10% of its applicants (as measured on test scores and high school GPA) then
- the Asian share at Harvard would rise from 24.9% to 51.7%,
- the white share would drop slightly from 37.6% to 35.5%,
- the Hispanic share would plummet from 14.9% to 2.7%
- the black share would vanish from 15.8% to 0.9%.
This current black share number (15.8%) is extremely high. Is this related
These numbers come from Table 5.3R on p. 110 of
REBUTTAL EXPERT REPORT OF PETER S. ARCIDIACONO
Students for Fair Admissions, Inc. v. Harvard
No. 14-cv-14176-ADB (D. Mass)
Professor Arcidiacono of the Duke Econ department is a hired gun expert witness for the plaintiffs in the Harvard discrimination case. He’s battling Harvard’s hired gun economics expert David Card. So keep that in mind when evaluating their statements.
This offers an interesting riposte to the ACLU’s recent tweet:
Make no mistake: A lawsuit against Harvard that's purported to represent the interests of Asian-Americans would, if successful, primarily benefit white students. #DefendDiversity pic.twitter.com/f7ssda9O9c
— ACLU (@ACLU) October 15, 2018
I’m surprised that Harvard has let the black share get so high because my impression is that Harvard has largely been hard-headed about not doing self-defeating things with its admissions policy. Harvard, which is now 382 years old, has been pretty competent over the centuries in admitting students who will be good for the Harvard brand. But making one-sixth of their admittees black means they are getting deep in declining marginal returns.
Perhaps Arcidiacono is playing some kind of game with the data?
Perhaps there is a ratchet effect in terms of black shares? At some point, Harvard’s endowment is doing well, so they spend a lot of money to boost their black share. But then Stanford gets a big windfall and ups the bidding war for black talent. And then Yale jumps in with massive money. Harvard should respond by letting its share of blacks fall as other vastly rich universities go hog wild about bribing blacks to attend. But any decline in blacks could lead to Bad Publicity, so it doubles down. And around and around, until blacks wind up absurdly over-represented at Harvard relative to their academic skills.
Looking at how Asians would be a majority at Harvard under a more objective system, current Harvard president Lawrence S. Bacow must now understand how Harvard president Abbott Lawrence Lowell felt in 1922 when he imposed a quota system to keep the then rising ethnicity’s share down.
The worry in both cases was that if they just let in the top students, Harvard, like Yogi Berra’s former favorite restaurant, would get so popular that nobody would want to go there anymore.