To Test or Not to Test

For more than 75 years, two standardized tests, the SAT and the ACT, helped determine who would be admitted to most of the nation’s highly selective colleges and universities. But in 2020, when COVID-19 rendered access to testing difficult, most colleges and universities became test optional.

Critics of standardized testing rejoiced. In their view, test scores reflect and reinforce racial and socioeconomic inequities. In the pre-COVID era, roughly one-third of students from the top 0.1 percent of the income distribution scored 1300 or better on the SAT, compared to about 5 percent of middle-class students. Students from the poorest families rarely hit 1300; only one in five took the test. And the average SAT scores of Asian students (1219) substantially exceed those of white (1082), Hispanic (943) and Black students (908).

Since wealthy, white students on average attend better schools, have more time to study and receive more tutoring and test preparation than their less fortunate peers, critics claim that “standardized tests are better proxies for how many opportunities a student has been afforded than they are predictors for studentsʼ potential.” Test advocates insist, however, that standardized tests help colleges and universities identify the students most likely to succeed academically.

Both are right. The challenge, in our view, is to use tests responsibly, as one indicator among many, taking into account differences in circumstances that might influence an applicant’s score.

In March 2022, the Massachusetts Institute of Technology reinstated its standardized testing requirement, insisting that its “research shows standardized tests help us better assess the academic preparedness of all applicants.” Brown, Harvard and Yale Universities; Dartmouth College; the California Institute of Technology; and the University of Texas at Austin, among others, followed suit, reigniting debate over the utility and fairness of the tests, admission goals, and the purpose of higher education.

This debate dates back almost to the introduction of the SAT in 1926, a history recounted in Nicholas Lemann’s thoughtful new study, Higher Admissions: The Rise, Decline, and Return of Standardized Testing (Princeton University Press). A professor of journalism and dean emeritus of the Columbia University Graduate School of Journalism, Lemann laments the use of standardized tests as gateways to America’s most selective colleges and advocates instead a focus on equipping “as many people as possible for as broad a set of life circumstances as possible.”

Ironically, as Nancy Cantor and Earl Lewis note in their introduction to Higher Admissions, the SAT was originally intended to be “a tool to democratize entrance into the nation’s Ivy League institutions.”

Before World War II, admissions policies garnered little attention, since, Lemann writes, “relatively few Americans went to college”; even elite institutions were, “by today’s standards, close to being open access.” But with the vast postwar expansion of higher education, the association between college degrees and socioeconomic status “rose significantly.” By the 1970s, admission even to prestigious public universities became “highly competitive and therefore also highly contentious.” And concerns grew that the SAT would “reinforce, not overturn, existing systems of social advantage.”

At the same time, colleges and universities sought to increase racial and ethnic diversity, sometimes by accepting students of color with lower grades and test scores than their white peers, a practice that lacked broad public support and which the Supreme Court ruled unconstitutional last year.

The court’s decision heightened the stakes for colleges and universities considering making SATs or ACTs mandatory again. Critics, including Lemann, cite studies showing that the tests mostly predict first-year grades and that high school GPA, which “picks up persistent, sustained effort,” is a better predictor of academic performance. Moreover, since many high schools remain largely segregated, “GPA tends to factor out a lot of class and race automatically by measuring a student’s performance in an environment with people of similar background.”

A recent study, however, indicates that “high school GPA does little to predict academic success in college,” at least at Ivy-plus colleges, while test scores predict college GPAs and “correlate strongly with students’ post-college outcomes, including earnings, attendance at prestigious graduate schools, and employment at prestigious firms.” And while a Fordham Institute brief notes that grades may be “less correlated with socioeconomic status and race/ethnicity,” they still correlate significantly with “student background factors, including race and family income.”

In fact, the Fordham brief states, “all components of a typical college application packet,” from letters of recommendation to personal essays, reveal differences among socioeconomic, racial and ethnic groups that appear to be “driven by average differences in academic achievement and preparation.”

In this context, more information, including test scores, seems better than less information, particularly if college admission officers consider, as MIT puts it, what students “have done relative to what might have been expected, given their resources.”

Unfortunately, efforts to assist college admission offices in making such comparisons have not fared well. As Patricia Gándara notes in a commentary at the end of Higher Admissions, a “Strivers Index” developed in 1999 at ETS, which administers the SAT, attempted to place scores in the context of students’ socioeconomic backgrounds but “met with a firestorm of criticism.” Twenty years later, the College Board Landscape system, which took “into account neighborhood and schooling factors that were known to correlate with test scores,” faced a similar response.

While admission offices may consider socioeconomic status when making admission decisions, the Supreme Court’s decision on affirmative action precludes any generalized consideration of race and ethnicity. In the year following reinstatement of its testing requirement, MIT admitted its most diverse class ever. But this year, in the aftermath of the court’s decision, enrollment of Black and Hispanic students plummeted. As other highly selective colleges that have reinstituted mandatory testing report their admission data, we will get a better sense of the impact of standardized tests on student body diversity, but MIT’s experience is not encouraging. And a recent study of 186 selective and highly selective institutions during the 2020–21 admission cycle, which predated the Supreme Court decision, found evidence that “test-optional policies … were linked with increases in Black student enrollment at moderately selective institutions, and some evidence that they were linked with increases in low-income student enrollment at highly selective institutions.’”

Lemann rightly notes that the question of who gets admitted to highly selective institutions receives “far more attention than it deserves,” since “only about 3 percent [of undergraduates] attend the fifty or so colleges that accept fewer than 25 percent of their applicants.” Accordingly, “the selectivity of admission to college is not a major factor in the lives of the overwhelming majority of young Americans.”

A much larger problem with American higher education, Lemann observes, is that only 60 percent of students complete their bachelor’s degrees in six years. Instead of using tests to select students for admission, he suggests, achievement and other tests should be used to assess student learning to enable “a more effective, tailored approach to that student’s instruction.” To increase completion rates and improve instruction, Lemann also endorses much greater support for “all-poor, usually all-minority, severely underresourced” K-12 schools, as well as for community colleges and less selective public colleges and universities.

Although he resists “a test-based meritocracy” based on the identification of “a limited cohort of winners relatively early in their lives,” Lemann acknowledges that it’s “an illusion to believe that dropping” standardized tests will usher in “a new age of truly meaningful equal opportunity in America.”

While appealing, Lemann’s recommendations do not resolve the dilemma facing the nation’s most selective institutions: how to assign a limited number of slots while valuing both academic excellence and diversity. In striking that balance, it will be helpful to compare the experience of highly selective institutions that have reinstated testing with those that remain test optional.

We agree with Lemann, however, about the need to focus much greater attention and resources on students who are not completing college. For much of its history, the United States led the world in educational attainment. The U.S. dropped out of first place in college degree attainment among its workforce in 2008, and it is no longer even in the top 10 in terms of the percentage of 25- to 34-year-olds with postsecondary degrees. The wealthiest country in the world has the resources, and in our judgment now needs to summon the will, to ensure that a much higher proportion of its citizens complete a bachelor’s degree.