If legacy preferences in admissions—an advantage for alumni children applying to college—end soon, historians will identify two 2023 nails in the coffin.
The first is the Supreme Court’s June decision outlawing affirmative action in admissions. Because U.S. housing is sharply segregated by race and socioeconomic factors, the thinking goes, attempts to maintain some degree of racial and ethnic diversity under the new law might depend on recruiting outstanding students from under-resourced neighborhoods and schools—perhaps accommodating those applicants by eliminating legacy preferences in admissions. Given their families’ educational attainments, legacies tend to be from the upper tiers of the income distribution—and disproportionately from the very highest rungs.
The second, attracting extraordinary attention in academia and the news media, is a 124-page paper, titled “Diversifying Society’s Leaders? The Determinants and Causal Effects of Admission to Highly Selective Private Colleges,” released in July by Harvard professors Raj Chetty and David J. Deming, and John N. Friedman, of Brown—the first and third of whom are director and co-director of Opportunity Insights, a laboratory on economic inequality and social mobility. The scholars analyzed admissions to the Ivy-Plus institutions (the Ivies and Chicago, Duke, MIT, and Stanford) by comparing children with similar standardized test scores and found that those from families in the top 1 percent of the income distribution were more than twice as likely to attend an Ivy-Plus school as middle-class children.
As Chetty illustrated in a riveting presentation at the Graduate School of Education on September 12, that means, for an entering class of 1,650 (coincidentally, the size of a Harvard College class), there are 157 more matriculants from the top 1 percent cohort than if such differential rates of attendance were smoothed out (as they are at most public flagship institutions, like Berkeley).
So, do away with legacy preferences and, presto, sharply skewed student profiles vanish? Not so fast. Differences in application and matriculation rates account for about one-third of the varying attendance by family income. Another one-sixth of the difference is attributable to recruited athletes (think of elite schools’ elite sports, like fencing, rowing, and squash). Other factors thus account for half the difference in attendance rates for nonathletes—causes Chetty et al. further disaggregate into the effects of the legacy preference (the larger factor) and admissions officers’ evaluation of applicants’ nonacademic qualifications, such as extracurriculars and personal character (which tie closely to prep schools much in favor among the elite).
In the face of these findings, what ought a selective college do—particularly given the Supreme Court mandate to stop considering race as a factor in admissions? A number have done away with legacy preferences: Amherst, Johns Hopkins, Carleton, Wesleyan. Fair enough.
But for other institutions, the decision may not be so simple. One can quibble with the scholars’ data, given some colleges’ gains in economic diversity. For all its evident power, their model has sharply defined criteria: standardized test scores count for a lot, and the “outputs” (graduates’ predicted income, attendance at elite graduate schools, or employment at prestigious firms at ages 25 to 28) surely track with socioeconomic mobility—but might not tell very much about adults’ subsequent social contributions or life trajectories.
Elite universities are unique in part because of their overlapping roles as centers for research, education, cultural enrichment, and so much more. That can make it difficult to analyze their performance along any one dimension of alumni success—but it presumably contributes to professors’ creativity, students’ breadth, and, (one hopes) some graduates’ pursuit of goals not defined solely by high incomes. Legacy candidates tend to be academically well qualified, and precisely because their families have afforded them opportunities, immersed in some of the experiences and extracurriculars that enliven their colleges when they enroll. No wonder admissions officers like them.
Two further considerations. First, the Ivy-Plus cohort enrolls 0.8 percent of U.S. undergraduates; doing away with their over-representation of students from the families in the top 1 percent of the income distribution will have essentially no effect on income inequality in the country. (That might even be the case for students at all the nation’s selective institutions, which enroll well less than 10 percent of the total.) Second, ending legacy preferences will correct for only about one-third of the total skew toward the 1 percent that Chetty et al. teased out of their data.
That said, there isn’t much case for retaining legacy preferences: they are a rotten look for selective schools already the subject of considerable public disdain—an own-goal if ever there was one. And it would be far better to move on than to provoke legislatures or courts into imposing admission standards on private colleges.
But simply abandoning legacy advantages won’t transform selective colleges’ student bodies. If institutions like this one really want to share the education they offer with a broader range of students, a first step is immediately at hand. As part of its post-affirmative action admissions, Yale is using Opportunity Insights data to identify neighborhoods by their socioeconomic traits and recruiting applicants accordingly. A bigger step, intellectually, is what Chetty and others call “need-affirmative” admissions—harkening back to the idea that Princeton president emeritus William G. Bowen, LL.D. ’73, and his coauthors advocated nearly two decades ago: instituting affirmative action for students from lower-income families alongside, rather than in lieu of, other preferences (see “A Thumb on the Scale,” May-June 2005, page 48). That might, over time, make for more socioeconomically balanced enrollments at the expense of a few athletic or artistic recruits on the margins.
Claudine Gay, Harvard’s new president, is a self-confessed data junkie versed in the rigors of quantitative social science. Alongside the broader considerations of values that will factor in her thinking on Crimson admissions, she will make an ideal reader for the Chetty-Deming-Friedman paper. Whatever she decides will not, by itself, transform the larger society—but it might have very salutary effects within the Crimson community. The outcome will be a fascinating application of research to practice.
—John S. Rosenberg
