Elite Colleges Should Try Harder to Stay Elite
They remain wildly expensive while doing less than ever to prove their worth. This combination is not sustainable.
Dear readers,
Rose Horowitch wrote last week for The Atlantic about the explosion of disability designations at elite colleges and universities, mostly granted to students with mild and increasingly common psychiatric diagnoses like anxiety, depression and ADHD. Officially, 38% of Stanford undergraduates are registered as disabled, as are a third of those at Amherst and more than a fifth at Harvard and Brown. As the proportion of the student body that is designated disabled has climbed at these universities, so has the number of students who receive academic accommodations — most commonly, permission for extra time on exams. And while students who receive academic accommodations in middle and high school lag their un-designated peers in reading and math scores and are moreover less likely to attend college, this is not true for college students, many of whom had no history of such a designation before they enrolled in college.
Horowitch cites the research of Robert Weis, a psychology professor at Denison University, who studies these trends:
When Weis and his colleagues looked at how students receiving accommodations for learning disabilities at a selective liberal-arts school performed on reading, math, and IQ tests, most had above-average cognitive abilities and no evidence of impairment.
And that, of course, gets at a major reason why so many students are suddenly “disabled” — it gives them a leg up, academically, in a supposedly elite, achievement-measuring environment.1
A college education is supposed to both build human capital and send a signal about human capital to employers. The explosion of accommodations is one of several forces that are simultaneously undermining elite colleges’ value on both of those dimensions:
Admissions standards at many schools have gotten less rigorous. This is because many universities have made standardized testing optional for applicants; in some cases, colleges will not even consider standardized test scores that students voluntarily report. The University of California Board of Regents voted in 2020 to phase out using test scores in admissions decisions, and the result is an exploding fraction of students at UC San Diego who can’t do basic math. This undermines the signal value of the degree: having gained admission to a UC is not the signal of academic aptitude that it once was. It also undermines actual human capital creation, because top schools are enrolling an unknown (but increasing) number of students who won’t be able to learn advanced material.
Once students are in college, the grades they receive have greatly inflated — a majority of grades at Harvard College are now straight A’s. This reduces the signal value of a high GPA: if everyone gets the same A’s, you can’t tell from a transcript which students actually performed the best. It also reduces the fundamental value of a college education, because it gives students the option to work less hard and learn less while still getting good grades — an option many are surely taking.
AI tools are undermining education. It is now easier for students to outsource their thinking to computers and harder for schools to accurately assess mastery. As a result, students learn less, and it’s harder to tell from a GPA how much a student has actually learned.
Disability accommodations are reducing the signal value of both grades and standardized test scores, because students take exams under conditions of varying difficulty, and those conditions are not disclosed alongside either GPAs or test scores that were earned under them. How could a recruiter identify which of two seemingly identical applicants would have trouble delivering projects on time, or in high-pressure situations? Firms may be able to close this gap through rigorous evaluation during the interview process, but a shift toward those kinds of assessments makes degrees less valuable, because firms are likely to get better at identifying the best applicants without relying on their degrees as the signal.
All of these trends are making a college education less valuable. Additionally, the college-age population is declining, meaning schools have fewer potential customers. So it shouldn’t be surprising that college costs have finally started to lag overall inflation in recent years — the market will no longer bear an ever-rising price tag for a college degree. That’s a big problem for colleges and universities, whose labor costs continue to rise, and who face a tougher audience than ever when asking the government for money.
This is a full-blown, system-wide crisis for higher education. And as Matt Yglesias writes, neither academics nor college administrators appear to have a plan for dealing with it, in part because they don’t even have a clear mission to evaluate their performance against.
Obviously, elite universities have a research mission. But I don’t think it’s hard to encapsulate what had traditionally been their teaching mission: to help the brightest students reach their fullest potential. If that’s still their goal, then they should be focused on unwinding the four trends I describe above. They can increase the value of a degree by adding back rigor where it was stripped away; they can find stronger applicants by requiring standardized test scores again (as many, but not all, have chosen to do in recent years); they can reverse grade inflation and refine AI-proof ways to assess performance.
And they should rescind any academic accommodations regime that treats deficits in cognitive performance as disabilities to be adjusted for, rather than traits to be measured. Our disability regime has made a category error, expanding a system that was once intended to ensure that non-cognitive deficits didn’t interfere with measures of cognitive performance so that we now “adjust” for factors that are simply components of whether or not someone is good at learning, understanding and applying academic material. Congress can help with this last point by repealing overbroad disability accommodation laws that universities have interpreted as forcing them to accommodate fundamentally cognitive conditions like anxiety and ADHD: If you’re too anxious to complete a test on time, that’s something a test should measure, not something it should avoid measuring.
At every level in the American education system over the last decade, left wing forces have pushed — often successfully — to evaluate less and rank less. They have fought for less standardized testing in K-12 schools, less tracking, less advanced math made available to the smartest students, and college admissions that are less tied to aptitude. The problems in colleges are in many ways downstream of the anti-measurement agenda in K-12 schools. And many of the people who have pushed this anti-assessment agenda are the same people who lament declining political support for educational institutions and declining funding for their activities.
But that outcome should have been foreseeable — if you make education less effective and make its effectiveness harder to measure, what about it is worth paying for, with public and private dollars? The political “assault” on schools can only be expected to get worse if schools make no effort to prove their worth. Liberals talk a lot about defending institutions, but the most important thing you can do to ensure that a valued institution survives a political assault is to make the institution worth defending. That effort has been sorely lacking in our educational system in the last decade.
By the way, I had a conversation about this topic with Ben Dreyfuss and Megan McArdle on this week’s Central Air podcast, which you can find below.
Very seriously,
Josh
I am told that a reason Stanford has a particularly high rate of disability designations is that such designations are especially useful for getting better student housing assignments — single rooms rather than doubles — and many of the students who seek designations for housing purposes don’t get academic accommodations. So, sometimes the system being gamed isn’t the academic system, but a system is still being gamed.



Moderate Democrats are in an increasingly abusive relationship with elite universities and higher education in general. Democrats go to bat for their funding and prestige (largely out of a fundamentally conservative impulse), while these schools pump out junk ideologies, misleading research, and militant activist staffers that make it difficult for elected Democrats to govern and coalition build.
Matthew Yglesias is over in The Argument making a fantastic case that CRT has been an incredibly corrosive and undermining force for liberalism, but he never circles back to blame the academics that popularized it and gave it institutional force. The same can be said of climate policy, where weak renewable energy only models have come to dominate the academic discourse, while actively hamstringing the energy transition of blue states.
I think moderate Dems should take the same advice that Matt gives the billionaires funding liberal NGOs; reflect carefully on the people you're giving money and influence to, because it's increasingly at cross purposes with your own goals.
I am unreasonably anti-elite university, though I understand their historical value. At this point, they seem more like a luxury good and a gatekeeping device to ensure children of the elite are advantaged and pushed to the front of the line for high-paying jobs. Until that last part changes, they will remain "elite," no matter the quality of their students, faculty, or administrators.