How colleges benefit from inflated SAT scores

When discussing the redesigned SAT, one common response to the College Board’s attempts to market the redesigned test to students and families by focusing on the ways in which it will mitigate stress and reduce the need for paid test-preparation, is to insist that that those factors are actually beside the point; that the College Board can market itself to students and families all it wants, but that the test is about colleges’ needs rather than students’ needs. 

That’s certainly a valid point, but I think that underlying these comments is the assumption is that colleges are primarily interested in identifying the strongest students when making admissions decisions. If that were true, a test that didn’t make sufficient distinctions between high-scoring applicants wouldn’t be useful to them. But that belief is based on a misunderstanding of how the American college admissions system works. So in order to talk about how the new SAT fits into the admissions landscape, and why colleges might be so receptive to an exam that produces higher scores, it’s helpful to start with a detour.

To back way up, the modern American university is essentially a hybrid creature. At the graduate and faculty levels, it’s based on the nineteenth century German research model. Faculty and graduate students are hired/accepted based on their research and publication records (“publish or perish”), and are expected to win grants and advance knowledge in ways that will amplify their institutions’ prestige — academia’s principal form of currency. Intellectual firepower is paramount; that is why graduate students, or at least Ph.D. students, are typically chosen by faculty members in the relevant department.

Then there are undergraduate programs, which operate on a “best graduates” rather than a “best students” model. Excelling academically is not of course frowned upon; however, the real goal is not only to admit the most intellectually capable students, but rather those deemed most likely to ultimately reflect well on their potential alma mater. And since institutions can expend tens of thousands of dollars to educate a single student, they are, to put it crudely, seeking return on investment. Although few admissions officers would put it that way (at least publicly), their job is effectively to identify the students most likely to provide that return. 

In addition, while academics are (ideally) the primary focus of undergraduate life, they are also designed to be one of many aspects of college life. It is understood that only a tiny percentage of students will go on to pursue academic careers; while there are of course many wonderful, genuinely caring professors who take a serious interest in teaching, there is often a tacit understanding that undergraduates will be kept out of professors’ hair so that the latter can do their research and pursue tenure in peace, while professors will not make undue demands of their students who, after all, have numerous extracurricular obligations to fulfill. To manage that balance, teams of deans, advisors, and graduate students are required.

Lest you think this is nothing more than a caricature of academia, in the two-and-a-half-years I sat in on faculty meetings in a humanities department at a certain Ivy League university in Cambridge, MA (a university whose dean of admissions has come out quite prominently in support of the new SAT), I recall undergraduates being mentioned by name exactly twice — and one of those times involved faculty complaints.

I also recall once receiving a phone call from an admissions officer who wanted to talk about a recent admit (“great kid”) with a particular interest, but didn’t know that the field in question — a common one — was handled by a different department. Admissions people and academic departments are sometimes not even on the same page.  

Given the goal of providing undergraduates with an “experience” rather than just a series of classes and exams, admission to a B.A. program is in some ways more like admission to a club than to an academic institution; it’s a question of “fit,” and the desire to maintain institutional (brand) identity.

At most universities, the people who will actually teach students are largely absent from the admissions process. Indeed, at many schools, the people who actually teach those admitted freshman are more likely to be underpaid part-time adjuncts than they are to be tenured faculty. 

For anyone familiar with the college admissions landscape, this probably isn’t news, but when gauging the reception to the new SAT, it bears repeating: colleges are not simply admitting individual students based on academic achievement but rather attempting to shape a class, one with the requisite ethnic, athletic, geographic, extra-curricular, monetary, and social attributes. Even if outright quotas are illegal, schools nonetheless have unofficial “targets” to reach. Indeed, admissions directors are in some places now known as “chief enrollment officers.” 

Assuming a student meets a minimum academic baseline, always left deliberately vague and contingent upon the student’s background, secondary qualities are in considerable part responsible for determining whether a student is offer a spot in the incoming freshman class. (At least that’s the case at most selective private colleges — it’s a little different at large public universities, which tend to be more numbers-driven.)

This “holistic” approach to admissions of course has a long and sometimes sordid history. As detailed in Jerome Karabel’s The Chosen, among other places, the very concept was invented in order to cap the number of Jewish students admitted to the Ivy League. The 1920s fear of enrolling too many “grinds” has perhaps been softened into platitudes about diversity and well-roundedness, but one can still hear barely concealed echoes of it. Few colleges, after all, would publicly admit to wanting students who spent most of their time studying. 

In recent decades, of course, the purpose of holistic admissions has shifted considerably, but the fact remains that colleges are free to select their students according to their own institutional priorities, and standardized tests are only valuable insofar as they allow schools to create freshman classes with the desired profile.

That, I suspect, is why most universities would never switch to a lottery-based admissions system for applicants who met a minimum set of academic criteria. Even the theoretical possibility of fantastic applicants with low scores and all the right other attributes outweighs the chaos and stress inflicted by the current system. 

Furthermore, universities are businesses, nonprofit status notwithstanding, and like any businesses, they have budgetary constraints. While an extremely select group of elite schools (which, ironically, tend to enroll only a small percentage of low-income students) have virtually unlimited funds for financial aid, that is hardly the case for most schools. The higher the number of competitive applicants, the more leeway there is to choose applicants who can pay sticker price.  

This is where rSAT comes in.

Even if scores shift generally upwards, they can only shift upward so high; 800 is still the top score in both sections. The top students, the ones who would have scored at the top on the old test, will continue to score at the top of the scale on the new test. In addition, however, students who would have obtained strong but not exceptional scores on the old SAT will be more likely to score in the stratosphere.

I suspect that this particularly true for students who are extremely strong in Math and merely above-average in Reading. (Many of my former students who obtained ACT Reading scores in the 34-36 range after failing to break 700 on the old SAT would very likely have fallen into that category.) The difference between the very top students vs. the merely very good ones is thus blurred, leading admissions committees to rely more heavily on “soft” criteria and making the overall process more opaque. 

As one commenter on my recent “Race to the Bottom” post pointed out, this is likely intended to be a feature of the system, not a bug.

One problem with the argument that colleges will reject a test that compresses the top end of the curve, making it more difficult to distinguish between high-scoring students, lies in the now-universal acceptance of the ACT. Colleges that had previous resisted accepting that test eventually gave in for fear out of losing too many applicants (Harvey Mudd was the last to capitulate, in 2007). And in case I need to spell it out, more applicants = lower acceptance rate = higher USNWR ranking = more applicants the following year (case in point: Northeastern University).

So the decision regarding which test would be accepted was ultimately driven by the applicants, not by the schools themselves. Even if scores rise across the board, individual schools can still tout their newly higher scores as a selling point; parents accustomed to the 1600 scale will assume the scores mean pretty much the same thing they meant in 1985. Not all schools will benefit equally either; schools with features that make them attractive to large numbers of applicants for other reasons (sports, location, etc.) will get the biggest boost. 

It also seems reasonable to assume that two groups of students stand to benefit most from score inflation: motivated underprivileged students from poor-to-mediocre schools who might have scored in the 400s on the old test but who could break the 500 or even the 600 mark on the new one; and somewhat above average middle-class-and-higher students who would likely fall short of the 600 or 700 mark respectively on the old test but manage to clear it on the new one.

The first group will help schools boost the number of minority admits while mitigating some of the pushback over test scores; the second will help them more easily accept full-paying students who in the past might have been eliminated, thus helping to offset the financial needs of the first group. Even if such students are at the low end of the range, they are still more likely to be in the “admissible” pool with the new test. For a college that charges upwards of $60K/year and needs more students paying full-freight, the sudden appearance of more students in that category is a boon. 

In contrast, students who would have scored at the very top on the old exam will find it more difficult to stand out if their abilities are not in evidence elsewhere in their applications. 

At the most selective schools, the superabundance of high-scoring applicants will also give admissions committees increased liberty to select the students with the most desirable non-academic traits without appearing to compromise the quality of their admits.

As things stand, only a relatively small number of students are admitted on academic achievement alone. Those students’ achievements tend to go far, far beyond the SAT, and even in an applicant pool filled with high achievers, they stick out. The SAT isn’t usually that big a deal for them; they’re more concerned with things like cancer research.

For students below that level, however, colleges will have an easier time justifying their decisions to deny applicants with exceptionally high test scores who don’t quite meet the extracurricular/personality bar. Both Harvard and Princeton have recently faced accusations of discrimination from Asian applicants who were rejected despite their extremely high scores and grades, while other, less academically accomplished students were admitted. Although both universities were cleared, that’s not the sort of publicity either school wants to invite. An increase in the number of non-Asian high scorers, which a flattening of the top of the curve would likely produce, would reduce the potential for those sorts of those allegations. 

Finally, colleges at the lower end of the hierarchy stand to benefit, but in a very different way. Some states have implemented policies to guarantee that students who are enrolled in public colleges and who meet a certain benchmark on the SAT will not take remedial classes. Such classes are costly to administer, and a lowered benchmark would reduce the need for them.