Fourth installment in a two week discussion of the recent New York Times article “Generation Later, Poor Are Still Rare at Elite Colleges” by Richard Perez-Pena. Interested readers should check out our collection of posts about socio-economic issues related to admissions, from which I have plagiarized extensively.

β€œYou can make big statements about being accessible, and have need-blind admissions and really low net prices for low-income kids, but still enroll very few of those low-income kids, by doing minimal outreach,” said Catharine Bond Hill, president of Vassar College. β€œThere has to be a commitment to go out and find them.”

Since when did every college president decide that coolness requires three names? Cappy Hill, or in more formal settings, Catharine Hill, was happily associated with Williams for 20+ years without anyone ever using her middle name. But now we have to include “Bond?” Weird. And Morty Schapiro seemed to do the same for a while, with his regular reminders that his middle name is Owen. Anyway, back to the article . . .

EphBlog loves Cappy Hill ’76 something fierce
, but this is misleading. Hill is implying that there are hundreds (thousands?) of low-income students with Williams-caliber credentials who don’t apply to Williams or places like it because of ignorance. But there aren’t! This is a pleasant fantasy of those who like to believe that parental wealth and student academic achievement are not as correlated as, in fact, they are. Consider the flaws in Hill’s research (pdf):

First, “high ability” is defined as 1420 or above combined math/reading SAT scores. Recall that Williams uses a system of academic ratings (AR) and that ratings below 2 are automatically rejected unless they have some special attribute like race or athletics. From Peter Nurnberg’s ’09 thesis, here are the relevant definitions:

AR 1: “Valedictorian/top or close to top”, A record, “Exceeding most demanding program, evidence of deep intellectual curiosity, passionate interest in particular discipline”, Exceptional, Exceptional, 770–800, 750–800, 1520 — 1600, 750–800, 35–36, mostly 5s

AR 2: Top 5%, Mostly A record, “Most demanding program, many AP and Honors courses, highly significant intellectual curiosity”, Outstanding, Outstanding, 730–770, 720–750, 1450–1520, 720–770, 33-34, 4s and 5s

In other words, lots of the students (SAT 1420 to 1450) that Hill describes as “high ability” are pretty much automatic rejects at Williams (leaving aside whatever affirmative action Williams places on socio-ec diversity). Moreover, lots of the other students with SAT >= 1450 lack the grades and other scores that put them in AR 2. So, Hill is (purposely?) overestimating the pool of potential applicants.

Second, even such high ability applicants are not (meaningfully) under-represented! Hill reports that these high ability, low income students make up 10% of elite schools but 12.8% of the population. Not a big enough difference to get worked up about, I think. And certainly not enough to make me think that the returns to more “outreach” are particularly high.

Third, Hill does not control for the desires of the students. Imagine two students in Georgia, both with Academic Rating 1 (on the Williams scale), both admitted to Williams and to the University of Georgia. Both won free rides at UGA via merit scholarships. One is student is poor and one is rich. I bet that the the poor student is much more likely to choose UGA. This preference, alone, might be enough to explain why poor students are slightly under-represented at elite schools.

Fourth, Hill does not seem to recognize that a poor student, even if as academically accomplished as the rich student, might be better off at an (excellent) state school rather than a (2nd tier?) liberal arts college like Vassar. Leave the details of this debate for another day, but it is indicative of the attitude of people like Hill that they would, in most cases, think that a student choosing UGA over Vassar is making a mistake.

Print  •  Email