Who Benefits From Letters of Recommendation?

When the University of California, Berkeley, announced in 2015 that it would experiment with inviting some applicants to send in letters of recommendation, the news was surprising to many outside California. After all, letters from teachers and others who know applicants are common in elite college admissions -- public and private. But UC, which was for decades not nearly as competitive as it is today, had never adopted the practice.

Given that California voters in 1996 banned Berkeley (and all public colleges and universities in California) from considering race or ethnicity in admissions decisions, any changes in admissions policies in the state prompt scrutiny from those on all sides of the affirmative action debate. Since 1996, Berkeley and UCLA have struggled with black enrollment -- and while Latino enrollment has gone up substantially, it hasn't matched the growth of the Latino population in the state.

Some critics of affirmative action speculated that the letters would be a way for black and Latino applicants to have supporters point to their race or ethnicity or make a case for admitting them over students with higher grades and test scores. Others worried that teachers at wealthier high schools -- well schooled in the college admissions process and not overwhelmed by large classes -- might put more into the process and give yet another edge to those with many advantages to start with.

After just two years, the University of California is dropping the Berkeley experiment. But now UC is starting a new system, in which up to 15 percent of applicants at each campus (including Berkeley) may be asked to submit letters from teachers or others. The focus -- as was the case in the Berkeley pilot -- will be on applicants whose other materials (including grades, test scores and essay prompts) have left admissions officers uncertain about whether to accept or reject.

Nationally, admissions offices report that letters of recommendation matter -- although not as much as other factors. An annual survey by the National Association for College Admission Counseling asks admissions officers to rank various factors in admissions decisions. Only 15.2 percent report that teacher recommendations are of "considerable" importance, although 43.5 percent say that the letters are of "moderate importance." That is in contrast with the factor identified as being of greatest importance -- grades in college preparatory classes -- which 79.2 percent reported are of "considerable importance."

With the University of California Board of Regents slated to consider the issue this summer, Berkeley commissioned Jesse Rothstein, professor of public policy and economics at the university, to study the second year (the cycle that started last fall) of what ended up as a two-year experiment with letters. An important caveat he noted at the beginning of his report was that Berkeley is highly competitive -- only 19 percent of in-state applicants are admitted, and the UC system's clear requirements may have the impact of discouraging long-shot applicants. "In some cases, students who are very highly qualified and observably quite similar to admitted students miss the cut," the report says.

Likewise, he notes that Berkeley uses "holistic" review, in which each application is reviewed in its totality -- making it difficult to determine whether any one factor led to an acceptance or a denial.

Rothstein divided Berkeley applicants into two groups to weigh the impact of optional letters. For his "underrepresented" group, he included applicants who qualified for an application-fee waiver for low-income students, those who attended poor-performing high schools, those without a parent who attended college and those from underrepresented minority groups. (Of course, many fit multiple criteria.)

He found that Berkeley admissions officials were more likely to find underrepresented applicants on the borderline of admission or rejection, and as a result were more likely to invite letters from them than they were from other applicants. Forty-five percent of underrepresented applicants were invited to submit letters, but only 26 percent of other applicants were. But while 23 percent of applicants outside the underrepresented group ended up submitting letters (a very high percentage of those asked), only 30 percent of underrepresented applicants did so, a much smaller share of those who were invited to do so.

Generally, the study found that readers gave somewhat higher rankings to applicants with letters than to the same applicants without letters, and this impact was greatest with underrepresented applicants.

But because fewer underrepresented applicants submitted the letters, the impact on overall admissions numbers was negligible.

A New Option for the System

Under rules approved by the Board of Regents last week, all UC campuses will now have the possibility of what is called "augmented review," but only for up to 15 percent of the applicants at each campus, in theory, those for whom there isn't quite enough information to make a decision.

Jim Chalfant, chair of the systemwide Academic Senate for UC and a professor of agricultural and resource economics at the university's Davis campus, stressed in an interview that the new option isn't just about letters of recommendation, although most of the attention has focused there. Besides requesting letters of recommendation, admissions offices may also ask for first semester of senior year grades (not normally reviewed at point of admission) or answers to "a questionnaire inviting the applicant to elaborate on special talents, accomplishments, extraordinary circumstances and/or their school/home environment."

"It is important to stress that almost all applications to UC are complete as submitted and provide a clear indication of the applicant’s qualifications, without the additional information that an augmented review would provide," says the policy document prepared for the board about the change, which was developed with the Academic Senate.

Chalfant said that the Berkeley pilot "wouldn't scale." The university system received 171,449 applications last year for admission to the freshman classes that will start this fall.

In fact, Chalfant said that he is not sure that all the UC campuses will use the system, although there are indications it will be used at Berkeley, Davis, Irvine, Los Angeles and San Diego. And even at those campuses, he said, he doubts "anyone will get to 15 percent."

It's no surprise that some UC observers are once again wondering whether some groups -- and minority applicants in particular -- will benefit more than others from augmented review.

Chalfant said that the Academic Senate intends to monitor how the new system is used. He said there is "an obligation to talk about it" so people know why some reviews will be augmented, and others not.

But going in, he said, there is "no expectation that one group or another" will benefit more than others.