Research Notes: The Impact of Test-Optional Policies on College Admissions and Enrollment

Since the pandemic, many colleges and universities have adopted test-optional policies, allowing applicants to decide whether to submit standardized test scores. A new study by Anna Kye at UC Irvine and Meng-Jia Wu at Loyola University Chicago finds that, at one private Midwestern university with a moderately selective admit rate, the policy coincided with a larger and more diverse applicant and admissions pool, but didn’t impact which students ultimately enrolled.

Comparing admissions data from fall 2020 (test-required) to fall 2021 (test-optional), the researchers analyzed associations, not causal effects, between the test-optional policy and application, admission, and enrollment outcomes.

Applications rose 26 percent after the shift, including increases among low-income students and students of color, suggesting that the policy lowered a barrier to apply. Test submission patterns varied: White and Asian students were more likely to submit scores than Black and Hispanic peers, and score-submitters typically reported higher family incomes and GPAs.

Applicants who submitted test scores were admitted at higher rates and received larger average scholarship packages, though the number of admitted students from low-income, Black, and Hispanic backgrounds grew under the test-optional policy. But when it came to enrollment, these groups were less likely to accept offers than higher-income, Asian, and white students. As a result, the demographic makeup of the student body remained largely unchanged, and the average family income of enrollees actually rose despite greater diversity in the applicant pool.

Kye and Wu conclude that while test-optional policies can broaden access at the application stage, they do not remove the financial and structural barriers that ultimately influence enrollment. They caution against generalizing across all of higher education, noting that the findings are most relevant to institutions with similar profiles to the one studied.