It’s Not (Really) About Diversity

Last month, Harvard University announced that it will reinstate the requirement that applicants for first-year admission submit scores on one of the two best-known college entrance examinations—the SAT and ACT. Harvard’s decision came after similar ones by fellow Ivies Brown University, Dartmouth College and Yale University, and was quickly followed by an announcement on the reinstatement of standardized test requirements from Cornell University.

Like many other institutions, Harvard suspended its standardized testing requirement during the height of the COVID pandemic, recognizing that many high schools and test administration centers were inaccessible to students seeking admission. In explaining the decision to reinstate the standardized testing requirement, Harvard cited data from its own Opportunity Insights research group, which has found that SAT and ACT scores are a better predictor of students’ academic success than high school grades. Harvard also framed its decision as a way to identify and admit more students from low-income backgrounds.

Other elite colleges have also claimed that reinstating admissions tests will increase socioeconomic diversity. Massachusetts Institute of Technology reinstated its college entrance exam requirement for students entering in the fall of 2023, claiming that these standardized tests help the institution assess applicants’ preparedness for MIT, and thus their likely academic success there. MIT also asserted that test scores can identify students from socioeconomically disadvantaged backgrounds who might otherwise be overlooked, claiming that standardized test scores “help us identify socioeconomically disadvantaged students who lack access to advanced coursework or other enrichment opportunities that would otherwise demonstrate their readiness for MIT.”

Dartmouth made a similar argument in reactivating its entrance exam requirement for students who will start college in the fall of 2025. Researchers at Dartmouth found that high school grades combined with standardized test scores are a better predictor of academic success at Dartmouth than either grades or test scores alone. They also claimed that test scores can be read “in context” of students’ local educational environments to help identify talented applicants from lower-income and first-generation backgrounds and poorly resourced high schools.

In contrast, institutions such as Princeton University and University of Michigan continue to have test-optional admissions policies. Michigan concluded that a test-optional policy is the best way to ensure access and fairness: since going test-optional in 2020, Michigan reports that it experienced a sharp increase in applicants from all backgrounds, which the university argues helped boost diversity on campus.

How is it that both critics and proponents of standardized testing can cite diversity in favor of their arguments?

The answer is that the decision to require admissions tests, or go test-optional, is not primarily a function of data. Rather, it is driven by values.

Much depends on an institution’s goals in crafting an entering class, and there are differing logics across institutions. Some institutions have the goal of a “well-rounded” class; others seek to admit the most highly qualified; still others place a high value on the social diversity of the student body, opening up opportunities to students who historically have been excluded from access; and others privilege identifying the students who are most likely to succeed at their institution. Some institutions care about prestige and rankings, and orient their admissions process to selecting students who will enhance their reputations in relation to peer institutions by seeking to maximize the average academic qualifications of admitted and enrolled students. These various goals can differ across institutions, and can take on differing importance within a given institution over time.

The tensions among these goals are readily apparent. There are group differences in students’ average performance on college entrance exams such as the SAT and ACT. This means that assigning a high weight to college entrance exams will shape which students are offered admission. If one believes that these tests are objective measures of students’ readiness for college, and that they predict how students will perform academically at a given college, a high weight will select high scorers, but not necessarily result in a diverse class that reflects students who have overcome school, home, and family conditions that may have limited their high school academic success.

Conversely, removing standardized test scores and placing a heavy weight on high school GPA, or class rank, will boost the likelihood of admitting students who have commendable high school grades, but may not be as skilled at gaining high scores on college entrance exams. But because research suggests that test scores are often better predictors of how students perform in college than are high school grades, placing a high weight on grades works against the goal of identifying the students who are most likely to succeed at a given institution.

Colleges that have chosen to reinstate standardized test requirements clearly value finding students who are most likely to do well academically. That’s a defensible rationale, but the underlying value is not diversity. When examining group differences on standardized test scores, those from higher income brackets tend to have higher scores. This also means there is a smaller pool of low-income students with high scores, to say nothing of the fact that low-income students are less likely to take standardized tests in the first place.

In concrete terms, research suggests that wealthy students are much more likely to score a 1300 or higher on the SAT than students from low-income families. Only 2.4 percent of SAT test takers from the bottom fifth of the income distribution scored 1300 or above, compared to 17 percent of students from the top quintile. The disparities only become starker as wealth increases, as students from the top 1 percent were 13 times as likely to score a 1300 or above than students from the bottom 20 percent. Thus, although requiring the test may allow universities to better identify lower-income students who would do well academically, it is unlikely that this will do anything to fundamentally alter the diversity of the entering class.

Harvard’s motto is Veritas, which means “truth.” But they, and other institutions that pursue it, should at least be honest. There are any number of reasons for requiring standardized tests, whether it is admitting the most academically qualified students or giving admissions officers more data to make decisions. If diversity is a top priority, however, then the decision to require standardized test scores rings hollow without other fundamental changes to the admissions process.