For the last several years, we’ve been hearing that college students can’t read as much or as deeply as they used to, even at elite schools.
Now, math professors are also complaining. “Some students are five to eight years behind,” wrote UC Berkeley professor Zvezdelina Stankova in an August op-ed. “Office hours that should be spent discussing integrals instead become lessons on fractions and basic algebra you would expect students to learn in middle school.”
A recent study of math performance at Rutgers, New Jersey’s flagship public university campus, found the same problem. One professor said some students can’t add fractions without a calculator and become lost if a problem introduces a variable like “x.”
Across University of California campuses, more than 3,000 professors have signed a letter calling on the system to reinstate the SAT and ACT for admissions, as some schools have already done, arguing that it’s impossible to accurately gauge a student’s college readiness without an objective, external measure.1 Grading standards vary considerably depending on the school, and especially since the pandemic, grade inflation is rampant. Students may have done well in math at their high school—even taken honors pre-calculus—but still flounder in college classes.
But They’re Graduating
“Students on University of California campuses may struggle with basic math,” author Miriam Pawel recently observed in a New York Times op-ed on the SAT controversy, “but they are graduating at record rates.”
That might sound like an endorsement of Stankova’s argument that the university’s admissions standards are flawed. But Pawel was defending those standards. Her argument is that California’s K-12 public school system is “one of the weakest, most inequitable” in the country, and the state’s universities need to go “beyond harvesting the best and brightest in a troubled system.” Instead, they need to emphasize their “service mission.”
That could mean providing students with the remedial help they need once they arrive on campus. But that’s a tall order. The Rutgers study found that while some students with gaps in their math knowledge pay for tutoring or take more time to graduate, many switch to easier majors or simply drop out. Berkeley has seen similar problems.2
In any event, Pawel doesn’t seem to be calling for remediation. She seems to be arguing that simply providing students with a degree, whether or not they “struggle with basic math,” is fulfilling the university’s service mission, because it’s enabling students from underrepresented groups to succeed.3
It’s true that people with college degrees, especially from a school like Berkeley, are likely to go farther in life than people without them. But if that degree doesn’t certify what people think it does, at some point that lack of knowledge is likely to catch up with the degree holder, leading to failure in the workplace—and possibly affecting society as a whole.
Do we want people who struggle with basic math designing our bridges, or administering medications in hospitals? While well-intentioned, Pawel’s approach isn’t doing underprepared college students, or the rest of us, a real “service.”
“Fish Can’t Climb Trees”
A similar debate is playing out at the high school level in New York State, which is abolishing its Regents exams—externally graded, standardized assessments in various subjects that have long been required for high school graduation. New York’s move brings the number of states requiring high school exit exams down to just six. It’s not clear yet what system will replace the Regents, but it’s likely to rely at least in part on teachers’ assessments.
At a recent meeting, Regents board member Felicia Thomas-Williams pushed back against the argument that the change will result in a lowering of standards. She said she was “thrilled all New York City students will now be able to demonstrate their competency … in a manner that showcases their strengths.” She added, referring to a cartoon she’d seen, “you don’t assess a fish on climbing a tree, because fish can’t climb trees.”
As with Pawel, Thomas-Williams’ argument is well-intentioned. Students do have different strengths and weaknesses, and nobody wants them to feel like failures. Unlike Pawel, however, Thomas-Williams is not blaming the K-12 system for students’ inability to perform well on externally graded assessments. She’s assuming that some kids are just naturally unable to master complex academic subjects, in the same way that fish are naturally unable to climb trees.
If that’s the case, there are a lot of “fish” out there. The literacy scores of high school graduates in the U.S. have been dropping, and it’s estimated that over 20 percent of young adults with high school diplomas don’t have “commensurate” skills. But is that because they can’t acquire those skills, or because schools haven’t figured out a way to teach those skills effectively?
If high schools and colleges confer diplomas on individuals who lack pretty basic literacy and numeracy abilities, those people generally don’t acquire them later on. Tests show that the abilities of American adults are also declining—and, just as with students, the gap between high- and low-scorers is widening.
Explanations for Declining Achievement
That mirrors what Stankova, the Berkeley math professor, has observed. Some students are still well prepared for calculus, she wrote, but the middle tier of ability has collapsed, and the weakest students have become the single largest group. That, she argues, refutes the claim, made by opponents of reinstating the SAT/ACT requirement, that Covid is responsible for the decline in K-12 achievement. If that were the case, we would see a decline in ability across the board.
Covid is only one explanation offered for declines in student achievement—and one that overlooks the fact that scores were declining, and gaps widening, before the pandemic hit. Some say the problem is with the way we measure literacy.
Education professor Rachel Gabriel has argued that if students were assessed on topics they’re familiar with from their out-of-school experiences, their literacy scores would be higher. To enable kids to learn, she told a reporter for The 74, we need to “teach skills explicitly,” but in “a context that is relevant and engaging and has a real purpose in the world.”
Gabriel’s argument assumes that comprehension skills—things like the ability to make inferences about texts—are transferable, so it doesn’t really matter what students can read, as long as they can make an inference. But cognitive science indicates that the ability to do something like make an inference is highly dependent on how much relevant background knowledge the reader has.
So yes, if a student knows a lot about hip-hop music, she’ll be able to make an inference in the context of a text on that topic. Even if she doesn’t know about hip-hop but sees it as relevant to her life, she may be able to learn to make inferences in a text about hip-hop more easily than in a text about, say, World War II. But without acquiring the general knowledge and vocabulary that come primarily from subjects like history, literature, and science, she won’t be able to make inferences with, and gain knowledge from, complex texts on topics she doesn’t already know about.
In the same article in The 74, an analyst for the OECD, which administers the international tests known as PISA to samples of 15-year-olds in developed countries, argued that scores are declining because the complex reading passages on the tests are outdated. People no longer need to “engage with longer texts and difficult texts,” he posited, “because the way we consume written information has fundamentally changed.”
That argument assumes we’re not losing anything if people become unable to read and understand lengthier, more complex texts. But evidence suggests that kind of reading helps develop our cognitive abilities. Neuroscientist Maryanne Wolf has argued that without “deep reading,” we’ll lose our capacity for critical thinking, reflection, and empathy.
The Quest for Fairness Can Backfire
As I mentioned, some students are doing just fine. The ones who are not are clustered at the lower end of the socioeconomic spectrum and are disproportionately Black and brown. The subtext of the arguments that dismiss the importance of declining performance on standardized measures is that it’s unfair to hold those lower-performing groups to the same standards as those who are more affluent and disproportionately white and Asian.
But the quest for fairness can backfire. Consider the case of journalist Nikole Hannah-Jones, who chose to send her own daughter, Najya, to low-performing, majority-Black public elementary and middle schools in New York City, in the belief that such schools would improve only if middle-class parents like herself enrolled their kids there. In a poignant recent article, she revealed that Najya, now 16, blames that decision for holding her back academically.
At her middle school, Najya was confident she was a math whiz. But when she enrolled at a selective high school, she kept failing her algebra exams because of significant gaps in her math knowledge.
One problem, Nadja told her mother, was that her middle school had never assigned homework. Many schools have stopped assigning homework on the grounds of equity, based on the belief that students from better-resourced families have an advantage in completing it. When Hannah-Jones confronted Najya’s middle school principal on the homework issue, the principal expressed regret.
“It is important to not, in our pursuit of being antiracist, lower expectations,” she told Hannah-Jones. “Because this is what happens, right? It’s racism under the guise of antiracism.”
How to Narrow the Gap Without Lowering Standards
Hannah-Jones blames the low performance of public schools like those Najya attended on things like inferior funding and the behavioral and academic challenges inherent in serving a high-poverty student population. Aside from the funding issue, she’s not entirely wrong.4 It is challenging to bring students from low-SES families up to the academic level of those who are more privileged. It’s probably impossible to close the gap completely.
But we do have evidence that schools can do much more to narrow that gap than they’re currently doing—and those who overlook that evidence are unintentionally doing a great disservice to the very students they’re trying to help.
What can work?
School-wide behavior policies and teaching methods that create the calm and orderly environment that enables students to pay attention.
The explicit instruction, including in writing, that helps all students succeed but helps those from disadvantaged backgrounds the most.
A content-rich, logically sequenced curriculum that both engages students and equips them with the knowledge they need to succeed at higher grade levels.
Ongoing support for teachers, delivered by those with some understanding of cognitive science, in adapting the curriculum to their particular students’ needs.
Ideally, all of this will begin in the earliest grades—because the problems that manifest in college, or high school, or even middle school don’t begin there, and they only get bigger and harder to address as the years go by.
I’ve been in schools and school districts that are adopting all or some elements of this approach and are thereby transforming the academic trajectories of their low-SES students. I’ve read and heard about others—and I’m sure there are more that I don’t know about. But these examples get relatively little attention from the media or even those within the education world.
One reason may be that while the approach I’ve outlined accords with evidence from cognitive science, it goes against education orthodoxy, which downplays the value of knowledge, foregrounds “skills” that aren’t transferable, and favors having students direct their own learning as much as possible. Another may be the argument that this sort of success is possible in isolated instances, with extraordinarily gifted educators, but it’s impossible to scale.
Experimental Evidence Is Hard to Get
I’d love to be able to point to high-quality randomized controlled trials that definitively establish this approach will raise achievement and narrow gaps, because that’s the kind of evidence some skeptics insist on. But that’s not possible. There are some rigorous studies out there, demonstrating the positive effects of combining content-rich curriculum with literacy instruction, but it’s extremely expensive and difficult to do that kind of research on a large scale.5
We do, however, have some evidence from the international PISA tests given to 15-year-olds. On the most recent tests, from 2025, the general trend was significantly downward, including for the U.S. One relative bright spot was England, which has overhauled its education system over the last 15 years or so to stress content-rich curriculum and science-aligned teaching methods and rely more on externally graded assessments, among other reforms. While England’s scores didn’t improve significantly, its international ranking rose into the top ten because so many other countries declined. And its socioeconomic gap was “significantly narrower than the OECD average.”
Some have credited the country’s education reforms for these developments, but those who are unenthusiastic about the reforms have raised objections. They point out, for example, that the sample of English students who were tested were less likely to be economically disadvantaged or in need of special services, meaning that the results may have been overly positive.6 But that’s likely the case in other OECD countries as well.
It’s true that PISA tests themselves can’t establish causation—and that international comparisons are difficult. But we don’t need to rely on just one country and one round of PISA for evidence. An analysis by a group of international education scholars, focusing on the 2018 PISA data, found that the most successful countries used an approach that stressed “curriculum coherence” and knowledge, among other factors.
I do feel the force of the argument that today’s high school students and graduates shouldn’t be penalized for the defects of an education system they can’t control. But rather than simply lowering standards, the best bet is explicit writing instruction, embedded in rich content across the curriculum.
That’s much harder to do at the college level, where professors are used to having more autonomy. If students have missed out on essential knowledge and skills but are motivated to do college-level work, we need to figure out a way to support them. But highly selective universities may not be the best place to do that. As Stankova suggested, it would make more sense for them to get the remedial help they need from community colleges and then transfer if they can demonstrate they’re ready to do so.7
But we also need to think of the low-SES children who are still in school or on their way there. If we lower standards on the assumption that those kids will never be able to meet them, overlooking the evidence that—given science-informed curriculum and instruction—many of them can, we’ll deprive them of the opportunity to reach their full potential, as we’ve done to so many in the past. We’ll continue kicking the can down the road, until one day it explodes and damages us all.
Update, 9.23.26: The post was updated to add a reference to grade inflation.
An additional 900 faculty members in non-STEM fields have called for also reinstating the verbal components of those tests.
Berkeley created special courses for students who were far behind in math, Stankova told NPR. But, she said, even students who take those courses face a 46 percent risk of failure.
Opponents of standardized admissions tests argue that the tests keep out many lower-income students, while advocates say they make it easier to identify high-achieving members of that group. But writing in Chalkbeat, Matt Barnum reports that studies indicate that “test-optional policies seemed to raise the proportion of lower-income students by [only] about 1 percentage point.” That finding, however, applies only to schools with test-optional policies, whereas the University of California has a no-test policy.
Hannah-Jones claims that nationally, “schools in predominantly nonwhite districts receive $23 billion less in annual funding than their heavily white counterparts,” citing a 2019 analysis by EdBuild. But that study did not include the billions of dollars of federal spending under Title 1, which is aimed at high-poverty districts, many of which are predominantly nonwhite. A Brookings Institution study that did include Title 1 funds found that over the past 50 years, high-poverty and predominantly nonwhite districts actually spent more per student on average than their wealthier counterparts. And Najya’s elementary school, P.S. 307, is funded at just over $52,000 per student, which is far more than the national average.
Matt Barnum has suggested that less academically prepared students would benefit from attending more elite schools, which spend more money per student than community colleges, but it’s not clear to me that’s the case. Elite schools may be spending more money, but they’re not designed or equipped to provide remedial education. Perhaps ideally we should have another tier of well-resourced schools specializing in such education, but it seems unlikely that will happen.
Oh god.. students arriving at Berkeley 5-8 years behind, with university office hours for calculus being used to teach fractions and basic algebra, and somehow the compassionate response is supposed to be lowering the standard further 😭. Let’s solve educational failure by redefining success..
Insightful.
I would add, that the behavior policies and teaching methods actually need to be implemented and part of the lived culture, not just posted on the walls to create "print rich" environments to satisfy the Whole Language advocates.
Also, I loved Maryanne Wolf's "Reader, Come Home" and wonder how to teach both forms of reading when students are engaged in academic learning for <5 hours a day (8am-2pm, minus lunch, art/music/PE).