I was browsing through the admissions section of Inside Higher Ed recently when I came across a brief article announcing that Caltech had decided to move from requiring two SAT IIs (one math and one science) to making the exams optional. Now, over the last few years virtually every selective college—with the exception of a few engineering schools—has downgraded SAT from “required” to “recommended.” The fact that one more school is jumping on the bandwagon might not seem particularly noteworthy, just one incidence of the backlash against standardized testing.
Because the story involves Caltech in particular, however, it’s somewhat more interesting than it might at first appear. Not only because Caltech has traditionally been seen as a bastion of uncompromising rigor, but also because it’s difficult to see the move as separable from the school’s downward trajectory in the US News and World Report Ranking over the past 20 years, especially over the last decade.
Although it’s highly unlikely that either today’s applicants or their parents would be aware of this—unless the latter happened to be working in higher education or college admissions 20 years ago—Caltech was briefly, scandalously, ranked #1 by USNWR in 1999. I was a senior in high school that year, and I vaguely remember the kerfuffle it caused. I did some poking around the internet and found a great Slate article (from 2000!) explaining how the traditional rankings formula was established by a statistician named Robert Morse to ensure that HYP came out on top. Essentially, the rankings were created to affirm people’s preconceptions about the “best,” but with a scientific veneer; there was never any real pretense of objectivity.
In 1999, however, a new statistician was briefly permitted to alter the traditional formula, with the result that Caltech ended up on top (oops!). The next year, the following year, Morse was reinstated so that Princeton was #1, the formula having been “readjusted” logarithmically in the necessary categories to ensure the top Ivies’ continued dominance.
I didn’t check for every single year between 2000 and today, but a glance at the rankings over time reveals a distinct downward trend: from #1 in 2000 to #4 in the early aughts, then #7-8 by the mid-aughts, #10 by 2012, and down to #12 (its current position) by 2016. That’s below Penn, Hopkins, Duke, and Northwestern.
Now, Northwestern is a fantastic school, but does anyone seriously think it tops Caltech academically? And would anyone seriously buy that Caltech is a significantly weaker place academically than it was 20 years, particularly given that it has consistently remained at or close to the very top of various other international rankings?
Interestingly, Caltech’s decline almost exactly parallels UChicago’s rise—UChicago being the school that has proved most adept at manipulating the rankings game and shedding its reputation for incorrigible nerdiness. (Because college should never be too much about, you know, school.)
A more likely explanation for Caltech’s slide, therefore, is that USNWR has periodically rejiggered the rankings formula (most recently to deemphasize freshman academic achievement and admission rates while placing more weight on social mobility), in ways ensured not to knock HYP out of the top spots but still penalizing schools further down the line whose other metrics haven’t been optimized for the new rankings.
Notably, MIT has not given any indication that it is dropping the subject-test requirement…yet. (Interestingly, MIT does not specify that a Math SAT II is required, although it’s fair to assume that the vast majority of the applicants submit Math II.) MIT, however, is still ranked at #3, tied with Yale and Columbia. One could also make the case that MIT is a more media-savvy institution, more outward-focused and attuned to the public side of its image (to see what I mean, read the MIT admissions blogs, which comprise thousands of posts; Caltech has nothing even remotely comparable). It’s a big enough name internationally that ranking it lower would be inconsistent with USNWR’s aim of confirming public opinion about who counts as elite. And while MIT’s acceptance rate is more or less the same as Caltech’s (Caltech is actually slightly more selective: 6.4 vs. 6.7% for the class of ’23), MIT receives almost three times as many applicants (about 8500 vs. 21,000) and has a freshman class nearly five times as large (235 vs. just over 1100). So for now, MIT doesn’t seem to have an overwhelming need to reconsider how it assesses applicants.
Caltech, on the other hand, is a smaller, more purely academic institution, less glamorous despite its LA location, and less concerned about its image outside of academia. Indeed, when people talk about the very top schools, they typically refer to HYPMS—not HYPMSC. With fewer than 250 undergraduates per class, Caltech is so small that it just ends up flying under the public’s radar. And to reiterate, USNWR rankings exist in large part to confirm what people already believe; if Caltech isn’t firmly established in people’s minds as “A Place That Counts,” then there is no reason for USNWR to continue to emphasize factors that would allow it to rank at the top.
The real elephant in the room, however, is Caltech’s minuscule enrollment of African American students. The admissions site is, understandably, very coy about this fact, listing only statistics for “underrepresented ancestry” and lumping multiple groups together to obscure significant variations between them. From the statistics I was able to find, it seems that only around 1% of Caltech freshmen (~2 students) in a given year are African American, with another 7.5% Hispanic (and 45% Asian). MIT’s current freshman class, on the other hand, is 10% African/African American and 14% Hispanic, 41% Asian.
There are obviously multiple issues at play here, ranging from the dearth of African American high school seniors with the requisite mathematical/scientific track record, to the lack of name recognition (as compared to MIT) among people unfamiliar with academia, to the fact that the African American students who are admitted may choose to go elsewhere because they perceive Caltech to be unwelcoming. I can’t find the link now, but I did come across one source pointing out that in one year, Caltech enrolled far fewer black freshmen than it admitted because almost all (or maybe it was all) of them chose to go elsewhere.
And speaking strictly anecdotally, I have seen race shape decisions about Caltech at a way I have never seen at other schools: one of my former students who was admitted (with thousands of dollars worth of test-prep and essay tutoring, plus a very questionable extended-time allotment) hesitated to enroll: there wouldn’t be any affirmative action kids he would be automatically better than, he told me, and he was concerned about his ability to handle the work. He was also just a little bit worried that Caltech didn’t have the name recognition he craved.
In this context, Caltech essentially has no choice but to take some steps that at least give the impression that the school is actively seeking out underrepresented candidates. The SAT II is the easiest target, but ironically also the one least likely to have a significant effect—on American applicants, at least.
First, the vast majority of applicants will continue to take and submit SAT IIs because two of Caltech’s biggest competitors, MIT and Harvey Mudd, to which many applicants also apply, still require them (MIT does not specify which ones; as of January 2020, HM still requires Math II). Given that any seriously competitive applicant to those schools will almost certainly achieve a 780+ score on the Math II test (the curve being substantial), there is no reason not to submit to Caltech as well.
Second, let’s assume (very reasonably, I think) that Caltech’s eschewal of SAT IIs does not indicate a loosening of its current admissions standards; indeed, given the rigor of the freshman curriculum, there would be no advantage to admitting students unprepared to work at that level—aside from the fact that it would be massively unfair to set students up to fail, a decrease in the freshman retention rate/graduation rates would offset any diversity gains in terms of the USNWR formula. All that is to say that the profile of future admitted students cannot deviate too much from the current profile.
Now, generally speaking, students admitted to Caltech tend to be ones who have been fairly aggressive about seeking out math- and science-related opportunities beyond the classroom; kids who just go to math class and do well, and then go home and sit in their rooms and do more math, and maybe take some STEM AP tests… they don’t get admitted, no matter how smart they are and no matter their ethnic background. The ones who get in are the ones who do competitions, who captain their high school math and physics teams, who do STEM-related summer programs and conduct research in professional labs… The point is, they pursue opportunities aggressively.
To say to applicants: we expect you to max out the STEM opportunities available to you, and to demonstrate your overwhelming passion for math and science through extracurricular involvement in these fields… but we know that it’s just too hard for you to spend two hours on a Saturday morning taking a couple of relatively simple multiple-choice tests? That just doesn’t fit. The kids who are able to take advantage of everything available to them, no matter their background, are generally the same ones who can figure out how to sign up for two SAT IIs in the space of three years, even if the exams are held in the next town over. The idea that a horde of amazing undiscovered talent will simply come swarming out of the woodwork if this one minor obstacle is dropped is a fantasy.
Yes, there are brilliant kids with parents clueless about the admissions process and guidance counselors too overburdened to remember to tell them to sign up for the SAT IIs, but given the small size and self-selected nature of Caltech’s applicant pool, we are probably talking about a handful of kids at most. What is more likely to happen is that kids who are excellent but not phenomenal at math and who couldn’t be bothered to take the SAT IIs will figure that they might as well throw in an application to Caltech because, hey, you never know. Then the size of the applicant pool goes up, and the selectivity rate goes down, boosting the school’s cachet—if not its USNWR ranking under the current formula.
Even Caltech’s justification for dropping the requirement seems half-hearted: the admissions site makes clear that all applicants must submit either the SAT or the ACT, which test a lower level of math than the SAT II, but then explains that the SAT II is no longer required because it “is not necessarily in line with the expectations of the Caltech math and science curriculum, and so does not demonstrate whether a student is prepared to take on the Institute’s rigorous first-year coursework.” By that logic, why would SAT math scores be required? (Perhaps only the English portion is considered?)
Furthermore (as I’ve pointed out before), if Caltech truly no longer believes that math and science SAT IIs provide any insight into applicants’ ability to handle Caltech’s curriculum, why allow applicants to submit them at all? Wouldn’t it be more logical to refuse to consider them period?
The reality, of course, is that many applicants now submit scores from exams at a level far beyond the Math SAT II (and AP Calc), as they will continue to do in the future. Elite competition math requires elite coaching—it’s a lot like being an athlete. These students will also continue to enroll in advanced and college/college-level courses (many of which are not offered at schools with high minority populations, or are out of reach financially) and summer programs, some of them quite pricey. For them, SAT IIs are indeed irrelevant, and the focus on achievement in advanced coursework will change exactly nothing. But as long as other applicants are given a fair shot in theory…
An additional clue to what Caltech is up to can be found in the wording on its admissions site:
“In reviewing our admissions requirements, we have come to the conclusion that the requirement for submission of SAT subject test scores creates an unnecessary barrier to applying for a Caltech education,” says Nikki Chun, director of undergraduate admissions, noting that only a small percentage of high schoolers globally take the SAT subject tests.
Now, if the goal were to increase the number of underrepresented (American) minority students, why would it be relevant that few students take SAT IIs globally? Unless, of course, one of the other goals is to increase the percentage of international students—who on the whole tend to be more prepared in STEM fields than their American counterparts, and who are frequently full pay. (Compare Caltech’s approximately $3 billion endowment, tiny for a school of its stature, to MIT’s $17.4 billion).
The point is, there are a lot of things going on here. A lot. And to turn discussions about standardized testing into simplistic narratives about discrimination and leveling the playing field—or, conversely, about sacrificing academic standards on the altar of political correctness—misses the point about how incredibly complex these calculations actually are, how many constituencies interests must be balanced and appeased, how much would be required to actually effect change on a systemic level, and how convoluted the backstories actually might be.
Fascinating analysis. Thank you, Erica.
“First, the vast majority of applicants will continue to take and submit SAT IIs because two of Caltech’s biggest competitors, MIT and Harvey Mudd,”
Neither of these school consider Subject Tests any longer, similar to Caltech.
(and I have to add – “SAT II” is a blast from the past. The name was dropped in 2005.)
An interesting article, but it appears Cal Tech’s endowment per student is only moderately smaller than that of MIT’s.
http://reachhighscholars.org/college_endowments.html
And, if they want to really help their diversity statistics they will aggressively recruit overseas, offering scholarships where necessary. I can’t imagine it would be impossible to find 10 or 20 potentially world class science/math stars in all of sub-Saharan Africa, and another 10 to 20 in the Spanish and Portuguese speaking world.