Study suggests dropping SAT/ACT requirements could boost access, but hinder admissions

BALTIMORE, July 16, 2026 — Dropping standardized testing requirements could make college admissions more accessible for some applicants, but a new study suggests it may also reduce universities’ ability to identify high-potential students, according to research published in Management Science by INFORMS.

The paper frames the admissions decision as a tradeoff between “information” and “access.” Test scores can act as a reliable signal when other application components—such as grades, essays, or recommendations—carry different levels of predictive power across student groups. At the same time, removing the test can lower barriers that prevent qualified candidates from applying.

Using a statistical discrimination framework, the researchers show how eliminating a feature like test-score review changes the informational environment faced by admissions officers. In practice, admissions decisions depend on how all application data are observed together and how policy choices shape the applicant pool that reaches the evaluation stage.

Crucially, the study finds that outcomes do not move in only one direction. Dropping testing may lead to improvements in both diversity and academic merit in some settings, while in others it can worsen both simultaneously—demonstrating that policy effects are context dependent rather than universally beneficial or harmful.

The analysis incorporates capacity constraints and fairness goals, and it examines both non-strategic access barriers and strategic behavior by applicants. The team uses calibrated simulations to explore realistic admissions scenarios where results vary sharply depending on how other signals perform and who actually applies.

One striking implication involves student test-taking behavior in multi-school environments. When one institution drops standardized testing while another retains it, some qualified students may decide not to take the test at all, potentially reducing their chances at schools they would otherwise have attended.

The authors also emphasize a modeling insight: changing whether scores are available can alter how admissions systems “learn” from the remaining data. This can shift the observed relationship between application inputs and academic outcomes, not just who is admitted.

“Our results highlight that policymakers and admissions offices need to consider both the informational environment and access barriers, rather than treating the testing decision as a simple ‘tests good vs. tests bad’ debate,” the study notes.

The paper concludes that institutions may benefit from investing in better signals for all applicants while simultaneously reducing access barriers, instead of relying solely on the removal of standardized test-score requirements.

Read the full study via the DOI link and associated materials.