The Test Is Back: Why Elite Universities Stopped Trusting Grades Alone

On October 9, 2025, Princeton University announced that it will again require SAT or ACT scores, starting with the 2027-28 admission cycle for students seeking to enroll in fall 2028^[1][2]. Its reason was empirical: a review of five years of test-optional admissions found that students who had chosen to submit scores performed better academically at Princeton than those who had not^[1].

Princeton was late to a trend. MIT announced the return of its requirement in March 2022^[3]. On June 12, 2026, Columbia said it would follow from the 2027-28 cycle after a multiyear faculty review, making it the last of the eight Ivy League universities to bring testing back^[4]. The exam has changed too: the SAT is now digital, and 65 percent of test takers in the class of 2026 sat it at school on a school day^[5].

The universities reversing their pandemic-era pauses argue, with their own data, that optional testing worked against some of the applicants it was expected to help^[3][7].

What the admissions files showed

MIT made the case first. Stuart Schmill, its dean of admissions, said the office's research showed that in most cases it could not reliably predict who would do well at MIT without test results alongside grades, even after controlling for socioeconomic factors. His equity argument was practical: most students can prepare for the SAT with free tools, but usually cannot make their high school offer advanced calculus^[3].

A wider study followed in January 2024. John Friedman, Bruce Sacerdote and Michele Tine, of Opportunity Insights and Dartmouth, analyzed admissions records and first-year grades from 2017 to 2022 at multiple Ivy-Plus colleges, meaning the eight Ivies plus Stanford, MIT, Duke and Chicago. Among students of the same gender, race, family income and high school grades, those with a 1600 earned a first-year college GPA 0.43 points higher than those with a 1200. High school grades separated students far less: a 4.0 predicted a college GPA less than 0.1 points above a 3.2. Students admitted without scores performed like students who had scored 1307^[6].

The applicants who hid good scores

Some of the least expected evidence comes from Dartmouth. At the request of President Sian Leah Beilock, four faculty members (Elizabeth Cascio, Bruce Sacerdote, Doug Staiger and Michele Tine) examined applicant data from 2017 to 2022, spanning test-required and test-optional years. They found that Dartmouth reads scores in context. Under the old requirement, a less-advantaged applicant (first-generation, or from a below-median-income neighborhood or a high-challenge high school) with a 1400 had twice the admission probability of a more-advantaged applicant with the same score^[7].

Applicants could not see that. Dartmouth's published score ranges made a 1400 look weak, so most students at or below that level withheld their scores, and less-advantaged students did so at about the same rate as everyone else^[7].

The cost was measurable. A January 2025 working paper that grew out of the memo, by Sacerdote, Staiger and Tine, estimated that at one elite college less-advantaged applicants scoring 1420 or above were 3.6 times as likely to be admitted when they reported their scores: 10.2 percent against 2.9 percent^[8]. The peer-reviewed version, published in September 2026 in American Economic Review: Insights, reports a smaller gain, from 2.9 percent to 7.2 percent, still more than double^[14]. The original memo estimated, under what its authors call a strong assumption, that about 1,000 less-advantaged applicants in each pool had scored above 1400 and not disclosed it^[7].

Why would a missing score hurt? Without a common measure, more weight falls on everything else^[7]. Economists Raj Chetty, David Deming and John Friedman found that children from top 1 percent families were more than twice as likely to attend an Ivy-Plus college as middle-class children with comparable scores. They trace the admissions part of that edge to legacy preferences, non-academic credentials and athletic recruitment. Those factors were uncorrelated or negatively correlated with outcomes after college, while test scores were highly predictive^[9].

The gap has not gone away

None of this makes the SAT a level field. Among the 1,954,833 test takers in the class of 2026, students in the lowest-income fifth of neighborhoods averaged 923 out of 1600 and those in the highest-income fifth averaged 1180, according to the College Board. That is a gap of 257 points by our subtraction. The measure is the median family income of a student's census tract, not the family's own, and 34 percent of test takers have no income estimate^[5].

So the case for requiring tests rests on a condition: scores have to be judged against the applicant's school and neighborhood, as Dartmouth says it does^[7]. Used as a flat cutoff, the same scores would simply reproduce the income gap.

The trend is also narrow. FairTest, a group that argues for test-optional admission, counts 2,021 of 2,182 four-year institutions with confirmed policies, or 92.6 percent, as test-optional or test-free for fall 2027. Only 161 require scores, one more than a year earlier. Executive director Harry Feder says "there has been no national return to required testing" and argues that Ivy-Plus colleges, whose applicants fall in a very narrow band of high grades, are not comparable with most institutions^[10].

Broader research supports that caution. Christopher Bennett's 2022 study of nearly 100 private institutions that went test-optional between 2005-06 and 2015-16 associated the policy with a 3 to 4 percent increase in Pell Grant recipients and a 10 to 12 percent increase in first-time students from underrepresented racial and ethnic groups^[11].

California's six-year experiment

The University of California is a larger test of the argument. Its regents voted unanimously in 2020 to eliminate SAT and ACT requirements, overruling a faculty task force that had found the tests predictive across demographic groups, the Los Angeles Times reported. On June 11, 2026, UC said it would reconsider after roughly 1,400 science, engineering and math faculty signed a letter saying first-year students were arriving unprepared^[12].

A UC San Diego faculty and administration report supplies a stark figure: between 2020 and 2025, the number of entering students with math skills below high school level rose nearly thirtyfold. The report cites the pandemic, the end of testing, grade inflation and wider admission from under-resourced schools together, so the rise cannot be pinned on test policy alone^[13].

Critics answer with outcomes. UC's first-year retention has held at about 92 percent and its four-year graduation rate rose from 71 percent in 2020 to 74 percent, according to Jessie Ryan of the Campaign for College Opportunity. Even so, Janet Napolitano, who led UC when the requirement ended, now favors a reassessment. Any change would not apply before the fall 2028 application period^[12].

“It's been a six-year experiment, and it now needs to be revisited”

What it means in practice

• The strongest evidence comes from a handful of highly selective colleges. It shows that scores add information where applicants' grades are bunched at the top, not that every college needs them^{[6][8][10][14]}.
• Dartmouth's applicants were hurt less by the test than by not knowing what counted as a good score. Publishing how scores are read in context addresses that directly^[7].

The WAC view

The evidence supports a narrower claim than the trend suggests. At the most selective universities, a score read against a student's school appears to be among the least biased signals available, and making it optional left the applicants with the least guidance to guess wrong. Institutions that reinstate tests should publish how they read scores in context; those that stay test-optional owe applicants the same clarity about what replaces the score.

What to do with this

The events behind this article, in WAC News

• Princeton to require SAT or ACT scores again from the 2027-28 admission cycleOctober 9, 2025
• MIT reinstates SAT/ACT requirement for undergraduate applicantsMarch 28, 2022
• College Board announces the SAT will move to a digital formatJanuary 25, 2022

Sources

1. Standardized Testing. Princeton University Office of Admission, October 9, 2025.
2. Princeton Admissions Will Again Require Standardized Tests in 2027. Princeton Alumni Weekly, October 9, 2025.
3. Q&A: Stuart Schmill on MIT's decision to reinstate the SAT/ACT requirement. MIT News, March 28, 2022.
4. Columbia reinstates standardised test scores requirement. University World News, June 18, 2026.
5. 2026 SAT Suite of Assessments Annual Report: Total Group. College Board, 2026.
6. Standardized Test Scores and Academic Performance at Ivy-Plus Colleges (Friedman, Sacerdote, Tine). Opportunity Insights, 2024.
7. Report From Working Group on the Role of Standardized Test Scores in Undergraduate Admissions (Cascio, Sacerdote, Staiger, Tine). Dartmouth College, January 30, 2024.
8. How Test Optional Policies in College Admissions Disproportionately Harm High Achieving Applicants from Disadvantaged Backgrounds (Sacerdote, Staiger, Tine), Working Paper 33389. National Bureau of Economic Research, 2025.
9. Diversifying Society's Leaders? The Determinants and Causal Effects of Admission to Highly Selective Private Colleges (Chetty, Deming, Friedman), Working Paper 31492. National Bureau of Economic Research, 2023.
10. Overwhelming Majority of U.S. Colleges and Universities Remain ACT/SAT-Optional or Free for Fall 2027. FairTest (National Center for Fair & Open Testing), August 25, 2026.
11. Untested Admissions: Examining Changes in Application Behaviors and Student Demographics under Test-Optional Policies (Bennett). ERIC / American Educational Research Journal, 2022.
12. University of California System May Reinstate SAT Requirement. Los Angeles Times (via Government Technology), July 9, 2026.
13. Senate-Administration Workgroup on Admissions: Final Report. UC San Diego Academic Senate, November 6, 2025.
14. How Test-Optional Policies in College Admissions Disproportionately Harm High-Achieving Applicants from Disadvantaged Backgrounds (Sacerdote, Staiger, Tine), vol. 8, no. 3. American Economic Review: Insights (American Economic Association), September 2026.

Written by the WAC editorial team. Facts are drawn from the numbered sources above and were checked against them before publication; opinions appear only under “The WAC view”. The organisations and people named are not affiliated with the World Assessment Council unless stated. Send a correction