Why the SAT Is a Poor Fit for Public Universities

Although the Ivy League and the Trump administration disagree on many things, they share one conviction: The SAT is a “big, beautiful” test.

Earlier this year, the administration offered its Compact for Academic Excellence in Higher Education to nine institutions, later extending it to all U.S. colleges and universities. Those that sign on will receive preference for federal funding—on condition that they require “widely-used” standardized tests like the SAT or ACT for admission, among many other stipulations.

The Ivy League, for its part, has provided intellectual cover for the SAT’s possible revival at hundreds of colleges that had gone test-optional or test-free during the COVID pandemic. When Harvard, Dartmouth, Brown and others announced last year that they were reinstating the SAT requirement, the news landed like a thunderclap. A front-page story in The New York Times, buttressed with data in a report from Harvard’s Opportunity Insights group, claimed that standardized tests are better than high school grades in predicting college success and, if used carefully, might even help diversify elite campuses.

That claim has undeniable appeal, especially now that the Supreme Court has struck down race-conscious affirmative action. If the SAT is both a crystal ball for academic success and a ladder for disadvantaged students, why not bring it back?

Because the evidence says otherwise. The Ivy League is a poor model—and the SAT a poor fit—for America’s public universities. Far from expanding opportunity, testing tilts the playing field against the very students that public institutions are meant to serve.

The Myth of Test Superiority

The first pillar of the pro-test case is predictive validity: the claim that the SAT forecasts college success more accurately than high school GPA. At first glance, that seems plausible. Tests yield neat, comparable numbers, while grading standards vary widely.

But decades of research show the opposite. Studies from my own institution, the University of California, and many other colleges consistently find that, despite differences in grading standards, GPA in academic subjects in high school is the stronger predictor of college performance. That’s unsurprising. GPA reflects years of sustained work in demanding courses—essays, labs, exams—the same challenges students will face in college. The SAT, in contrast, is a two-hour snapshot on a Saturday morning.

So why do Ivy League analyses seem to show otherwise? Because of a well-known statistical issue. If you compare SAT scores and GPA without accounting for background factors such as family income or race, the SAT appears more predictive because it correlates more closely with those factors—which themselves strongly predict college outcomes.

Buried in a technical appendix to the Opportunity Insights report is a table that underscores the significance of family income and race. The table shows what happens when background factors are entered into the prediction model: The predictive weight for SAT scores falls by more than 40 percent but jumps by more than 50 percent for high-school GPA. Controlling for student background substantially limits the predictive validity of the SAT and strengthens that for high school GPA. EDITOR'S NOTE: This paragraph has been corrected to fix a calculation error.

The Diversity Mirage

The second argument for testing is that, if used “in context,” it might enhance diversity. A 1400 from a low-income student, the argument goes, might be given more weight than the same score from a prep school graduate.

It’s a comforting story—but not one supported by evidence. Ranking applicants by SAT score systematically disadvantages top applicants—those with stellar records in high school—from underrepresented backgrounds.

In one UC study I conducted, all UC applicants were ranked into ten equal tiers by high school GPA. The same students were then ranked by SAT scores and the proportion of Latino and Black students in each tier compared. The difference was dramatic.  At the top of the applicant pool— where the chance of admission is greatest—Latinos and Blacks comprised 23 percent of the top GPA tier but only 5 percent of the top SAT tier. Similar differences were found for low-income and first-generation applicants. Far from leveling the playing field, the SAT steepens the climb.

Holistic review can offset some of this but not enough, especially in a race-blind context. At UC, holistic review increased underrepresented minority enrollment by about 6 percent—far less than the negative demographic effect of ranking applicants by SAT score.

A Natural Experiment in California

COVID turned this debate into a real-world test. With centers closed, hundreds of colleges went test-optional, and a few—including UC—went fully test-free.

The results were striking. Applications rose, and enrollment of low-income, first-generation and underrepresented minority students increased— without affirmative action, which California banned in 1998. Crucially, academic performance did not decline: Students admitted without test scores did just as well in their first and second years.

This natural experiment undermines both pillars of the Ivy League’s case. The SAT adds little predictive value beyond GPA and eliminating it doesn’t compromise outcomes. What it does is broaden access and create a more representative student body—even in a post–affirmative action era.

Why Public Universities Are Different

Why do public universities view testing so differently? The answer is mission.

Elite private institutions serve less than one percent of U.S. students. Their admissions offices build small, hyper-selective classes focused on academic distinction and institutional priorities such as legacy ties.

Public universities are different. They include land-grant institutions, state flagships and systems like UC. They serve millions. Their mandate is not just to educate but to foster social mobility, drive economic development and enroll student bodies that reflect their states.

That mission shapes admissions policy. Selective publics not only value academic excellence but also seek top students from disadvantaged backgrounds. Employing SAT scores systematically disadvantages those at the heart of that mission: high-achieving students from underserved communities.

There’s another key difference: Public universities are deeply linked to the K-12 system. Their admissions standards ripple backward, shaping high school curricula and instruction. Reliance on the SAT distorts that system, encouraging test prep over deeper learning. Valuing GPA, by contrast, reinforces classroom engagement and persistence.

The Wrong Yardstick

The SAT was once seen as a tool for democratization, identifying “diamonds in the rough.” Today, it’s more a measure of inherited privilege. Before UC dropped the test, more than 40 percent of the variation in applicants’ SAT scores was explained by family background—income, parental education and race—compared to just 9 percent for GPA. Treating the test as a neutral measure of merit is an illusion.

Public universities know this firsthand. The most able students aren’t always those with the highest test scores but those who’ve proven themselves over years in their schools and communities. These are the students most likely to drive social mobility—the core of the public mission.

A Renewed Focus on Public Mission

UC has now been test-free for four years. The sky hasn’t fallen. Academic standards haven’t slipped. What has changed is the student body: More low-income, first-generation and underrepresented students are earning spots without affirmative action.

UC’s faculty admissions board recently concluded that eliminating standardized tests has “demonstrated a way in which UC can lead in advancing access and opportunity for the state’s students.” That’s the right lesson.

Rather than emulate the Ivy League, America’s public universities should double down on their traditional public mission in this new, post–affirmative action era.