In my previous post, I discussed the recently published paper by superstar Harvard economist Raj Chetty, along with colleagues from Brown, confirming what I suspect most people involved in selective college admissions could intuitively report—namely, that the top 1%, and really the top .1%, enjoy a massive advantage in the college admissions process, largely as a result of a non-academic factors.

I’ve read the full paper, and although the statistical formulas used to calculate the effects are well over my head, the conclusions Chetty et al. draw are quite clear. (A detailed summary is also available.) Some of the main takeaways are as follows.

For Ivy-plus applicants who received equally high academic ratings (SAT 1510+ on the pre-2016 exam; ACT 34+, with the most challenging course load), across the full range of incomes:

  • Compared to applicants below the 99th percentile income-wise, applicants in the top 1% (>$611K) are admitted at a 43% higher rate than middle-class applicants.
  • Applicants in the top .1% are admitted at a 250% higher rate than middle-class applicants.
  • Admission rates are lowest for applicants in the 70th-80th percentiles ($83K-$116K), beginning to rise for those below the 70th percentile and increasing through the 20th (That is, elite universities admit a substantial number of poor students—the gap is in the middle.)
  • The difference in acceptance rates between affluent and non-affluent applicants is largely due to, in order, “personal ratings,” legacy, and athletic recruitment.
  • Legacy applicants at a given school are rejected from *other* Ivy-plus schools at rates comparable to non-legacy applicants, implying that legacy status is the key factor in their admission—not that they are superior candidates.
  • For non-wealthy graduates, attending an Ivy-plus school rather than a public flagship significantly raises the chance of entering the 1% by age 33; working for a top firm; or attending an elite graduate school.
  • SAT/ACT scores remain one of the best predictors for non-affluent students’ post-college success. In contrast, there is no correlation between high personal ratings and post-college success.

A few comments.

First, to be clear: Admissions officers are not merely looking at applicants’ family incomes and issuing automatic rejections solely on monetary grounds. A candidate at the top of the pool will be offered admission regardless of their financial situation. Rather, the problem for non-affluent applicants is accruing the types of “soft” factors that will allow them to make it to the top of the pool in the first place.

Next, regarding test scores: Although the paper does mention that their predictive effect might be understated because of the amount of prep upper-income applicants may receive, it does not mention the extraordinary percentage of affluent students who receive extra time on the SAT and ACT. Because the dataset was primarily taken from applicants who would have taken the SAT after 2003, when extra time stopped being flagged, the correlation between high test scores and success post-college among non-affluent applicants may be even higher than the authors realize. (And given that average GPAs at private schools have been inflated more than ones at public schools, the correlation between test scores—as opposed to overall academic achievement—and post-college success for public-school applicants could be even tighter.)

So, to reiterate from my last post: Lower- and middle-income students are not failing to gain admission because their grades or test scores fall short—rather, they are being rejected for subjective reasons. Essentially, wealthy applicants need high test scores so that they can be admitted on other grounds; the scores alone are not the determinant. Consequently, there is no reason to think that test-optional admissions system will, on its own, lead to an increase in the number of lower- and middle-class students admitted to Ivy-plus schools. Non-affluent candidates who do not submit SAT or ACT scores will almost certainly be less accomplished academically overall, and thus even less likely to gain admission—particularly with the dismantling of Affirmative Action.

Now, all that said, I’d like to take a slight detour into a recent op-ed piece entitled “Legacy Admissions Don’t Work the Way You Think They Do,” published in the New York Times by the Columbia sociologist Shamus Khan. Contrary to its title, the article reveals absolutely nothing insightful or justificatory about the practice of giving the daughters and sons of alumni a tip in the admissions process. Its thesis—that legacy admissions are valuable because they provide the bright but socially unconnected with the opportunity to rub shoulders with the elite—is self-serving and tone-deaf nearly to the point of satire. (It also ignores a basic flaw its own argument: there is absolutely no reason universities could not halt legacy admissions while continuing to favor children of the otherwise wealthy and connected.)

If ever the masses required confirmation that graduates of elite colleges are detached from mainstream reality, this is all they would need. It is, perhaps, the single most clueless piece about college admissions that I have ever read.

And yet, in its matter-of-fact assumption that of course, the purpose of elite universities is to create and preserve an elite, it is highly revealing—and, in its own way, more realistic than the Chetty paper, which operates on the assumption that universities seriously want to admit a more economically diverse range of students. Despite expanded financial aid policies, Chetty et al. point out that the economic composition of Ivy-plus students has not changed in 20 years. In fact, the income distribution they report tracks the exact phenomenon that Khan describes: elite colleges are skewed toward admitting the most affluent, plus a meaningful number of exceptional low-income students who will benefit from being fast-tracked toward positions of influence. A small percentage of middle-income applicants make it, but they are not the main focus.

To be honest, it strikes me as somewhat naïve to imagine that elite schools are unaware of the way in which the non-academic aspects of undergraduate applications systematically work to the advantage of the wealthy—especially since the holistic admissions process was introduced specifically for that purpose. Regardless of what particular admissions officers happen to think of Khan’s argument, at some level—perhaps conscious, perhaps not—they may subtly discount middle-class applicants on the grounds that attending an elite institution will not significantly alter their life prospects, whereas for low-income applicants it can have a dramatic effect.

Also, admitting low-income applicants is also good PR, whereas admitting ones who are merely middle-class does not carry the same whiff of moral showmanship.

Besides, admissions criteria are weighted to produce a class with the characteristics that a university has already decided it wants. Particular metrics are given more or less consideration not (only) because of their intrinsic merit but rather because of their ability to affect the desired outcome. (It has always struck me as vaguely hilarious when colleges announce that they have succeeded in admitting a class with a particular set of demographics—well, yes, because engineered their process specifically to select for those characteristics.) Thus, disproportionate weight is given to non-academic factors because they are what colleges value. If the goal were genuinely to select the most academically gifted students—those who are, according to Chetty et al., actually most likely to achieve post-college success—elite universities would not adopt test-optional policies. Moreover, they would actually push back against changes such as non-flagged extra time (which overwhelmingly benefits wealthy white students) and the re-scaling of SAT scores to obscure the difference between the very top performers and those a rung below. But they have made no attempt to use their position to do these things. On the contrary, they have more or less welcomed them—or at least they have done so publicly.

The question, however, is not really whether Ivy-plus colleges are primarily educational institutions or clubs for the elite and those who would benefit from proximity to them. In reality, they are both things at once, and often for different types of admits. This situation reflects the odd, bifurcated aims of elite undergraduate education in the United States (and, I think, the ambivalence with which Americans regard intellectual pursuits in general).

As I have written about before, in the late nineteenth and early twentieth centuries elite American universities began to shift toward the German research model while also continue to serve the social elite à la Oxbridge. This led to a system in which newly created graduate programs became focused exclusively on academics, whereas undergraduate programs sought to balance their traditional social clubbiness with a new emphasis on scholastic achievement.

The tension between those competing demands has persisted, and it largely accounts for the subjectivity and shifting emphases of the U.S. undergraduate admissions process; and the fuzzy definition of “merit” as something conveniently benefitting the most advantaged applicants represents an awkward attempt to resolve the problem while providing a veneer of fairness.

As Chetty et al. point out, “simply removing the admissions advantages currently conferred to students from high-income families (or offsetting them with corresponding advantages for students from lower-income families) could increase socioeconomic diversity by an amount comparable to the impacts of race-based affirmative action on racial diversity.”

However, this point ignores the fact wealthy families—and their apologists—are likely to perceive any attempt at removing those advantages as punishment. It is not difficult to imagine the howls of protest: Our children did what you wanted and became championship fencers/ published authors/founded charities aimed at empowering inner-city youth by teaching them to code, and now you’re going to deprive them of what they’ve worked so hard for? Despite claims to the contrary, elite college admissions is a zero-sum game: an acceptance for one applicant means a denial for 15-20 others. If there is one hard fact that virtually all elite-college applicants can agree upon, it is that.

If colleges genuinely wanted to address the problem, they could of course follow the lead of MIT —the sole Ivy-plus institution to admit students more or less equally from across the economic spectrum (Chetty, 16). They could, for example, tighten distribution requirements and require mandatory final exams; support professors in maintaining an appropriate level of rigor; and make the overall experience less enjoyable for all but the most serious students. Those types of changes would undoubtedly result in a more self-selected applicant pool, one from which wealthy prep-school students who wanted to work hard but not that hard in college would naturally remove themselves. But these measures, in addition to reducing the size of the applicant pools and potentially increasing acceptance rates—bad for USNWR rankings—would be at odds with the Ivies’ historical undergraduate identities and are therefore unlikely to be seriously considered.

In short, while Ivy-plus schools may eventually be nudged (or forced) into making some tweaks around the edges that result in the acceptance of a marginally higher of middle-class applicants, there is little reason to believe that the admissions process will be fundamentally overhauled. Even if the pressure to ban legacy admissions eventually grows to the point at which universities have no choice but to accede, development lists (which by definition are effectively limited to members of the .1%)  and Z-lists (which prioritize full-pay applicants) are almost certain to remain. The dismantling of Affirmative Action may have briefly exposed the extent to which universities have used racial diversity as a shield to obscure a lack of economic diversity; however, tuition and fees are now around $80,000/year, a figure that schools simply would not charge without assurance that a critical mass of students could actually afford to pay. And as long as they continue to offer their current level of cachet, they will not hurt for applicants for whom a quarter-of-a-million dollars remains a mere drop in the bucket.