The Persistent Grip of Social Class on College Admissions

It’s hard to disentangle social class from the college admissions process. The University of California system says it’s trying, announcing recently that it’s dropping consideration of the SAT and ACT. (It was part of a settlement in a lawsuit alleging that the tests are biased along lines of race, wealth and disability.)

More than half of U.S. colleges have made the tests optional for fall of 2021 admissions, according to FairTest, a group opposed to college entrance testing.

Because those tests are receiving so much scrutiny, it’s easy to overlook the influence of socioeconomic background on other admissions yardsticks.

Take the college essay. It’s the most important “soft factor” and the fourth-most important overall factor — after grades, curriculum strength and standardized test scores — according to a 2019 survey of admissions employees.

But essays can be polished by a paid professional third party, or helped along by an upper-middle-class parent.

In another sign of the persistent pull of social class, a recent working paper from authors affiliated with the Student Narrative Lab at Stanford shows that essay content, when quantified through a computer program, is more highly correlated with household income than SAT scores are.

Researchers did not analyze whether these signs of status affect an essay’s quality, or speculate on whether they would make any difference in an evaluation by an admissions officer. But the research suggests that much of the socioeconomic information critics accuse the SAT of reflecting can also be found in essays.

The paper used software to classify essays written by nearly 60,000 applicants to the University of California system in 2016. The essays were quantified partly through syntax choices. The number of commas, total punctuation and longer words were correlated with higher household income, for example, although that doesn’t necessarily equate to better writing.

The content was also quantified by word choice patterns, which are associated with particular topics. Admissions officials might not look more favorably upon essays written on certain themes, but it’s still notable that there are significant differences in the topics associated with higher and lower household incomes.

The topics associated more with students from higher-income households tended to be “more thematically abstract: human nature, seeking answers and sensory experiences,” said AJ Alvero, an education Ph.D. candidate at Stanford and one of the authors of the paper.

Topics more associated with lower-household-income students “were about interpersonal relationships (e.g. multiple topics about family) and school issues like tutoring groups and time management.”

A prior study by the same authors found similar patterns in income difference. A co-author of the study, Sonia Giebel, a Ph.D. candidate in sociology of education at Stanford, stressed along with the other authors that the content they identified was not a marker of essay quality, but pointed to a broader theme: “Class patterns are likely to be present across all the elements used to make admissions decisions.”

Poorer students, beyond writing their reality, may also be more likely to write about “economic insecurity” and “abuse” because of trying to meet perceived expectations. Even without specific guidance from admissions offices, they might feel obligated to “sell their pain.”

In contrast with much of the rest of the world, American admissions officers have a lot of discretion. Relying on elements like the essay gives them leeway to judge merit away from close scrutiny. The history of the so-called holistic approach — looking at the whole applicant and not just academic metrics — has not always been encouraging.

As Jerome Karabel wrote in his book “The Chosen,” relying on nonacademic characteristics had its origins in policies starting in the 1920s that aimed to limit the number of Jews admitted to elite universities. More recently, the discretion and opacity in admissions have been seen by some as harming high-scoring Asian students by penalizing them based on “character” or “fit.”

Despite this, the holistic approach seems here to stay. “I do think that it’s very possible that in this first full year of test-optional being widespread, there very well could be more emphasis in some applications on the soft factors, with the essay being one of them, along with recommendations and extracurriculars,” said Robin Miller, a consultant at the admissions counseling firm IvyWise who formerly worked in admissions at Georgetown and Vanderbilt.

Colleges may want to pursue egalitarian goals, but they have other aims, too. They need to meet tuition revenue targets, and some colleges face a more dire financial situation because of the pandemic.

Analysis of data on recruiting visits by colleges has shown that richer, whiter high schools tend to receive more visits. The persistence of legacy admissions at some elite colleges — many of whom have more students from the top 1 percent of income than the bottom 40 percent — shows that though social mobility may be a goal, it can conflict with pleasing potential donors or attracting sufficient numbers of students who don’t need financial aid.

Shifting away from standardized tests closes down some avenues for class bias like test prep (even though research finds prep has only a modest effect on test scores), but leaves many others.

Standout extracurricular activities might be accessible only to the wealthy. Or they can even be faked entirely, as in the infamous “Operation Varsity Blues” scandal.

Colleges are caught between multiple goals: predicting the people most likely to succeed academically; identifying talent missed by conventional metrics; collecting adequate tuition income; enrolling a diverse class of students; encouraging and enabling social mobility; complying with legal constraints on affirmative action.

Supporters of the SAT say it’s effective at predicting college academic performance. But if predictive validity were the main goal of admissions, you could argue for directly using higher socioeconomic class as a qualifier for admission because wealthier students tend to transition to college more easily. (Jesse Rothstein of the University of California, Berkeley, made that point last year in testimony against the use of the SAT.)

Nonacademic factors like an essay don’t offer an obvious numerical pecking order like a G.P.A. or SAT score. Reliance on soft factors can allow college admissions offices to pursue their goals but deflect questions about which of the goals they prioritize.

Admissions officials can say they consider every individual’s unique traits, but it appears these traits are mostly inseparable from socioeconomic indicators in applications. Colleges still have to make tough decisions in showing what they truly value, but it seems their decision-making will now be more obscured from the public.

Arvind Ashok is a recent graduate of the University of Texas at Austin, currently deferring enrollment to law school. You can follow him on Twitter at @arvind__ashok.