Letters of recommendation

Image credit: Quang Ho/Shutterstock.com

One of the things that I find fascinating and almost horrifying about university life is how little it has changed since I stepped across the threshold of my undergraduate institution in 1976. This explains, in part, why we still write letters of recommendation. Everyone knows they are burdensome to all involved, and that they inject class and racial bias into any process.

Yet no one seems to know how to stop asking for them. Instead, we have created complex software systems to manage letters of recommendation, and by doing so, allow the system as it is to not just survive—but thrive.

Let me note that every institution in my life other than higher education has become almost unrecognizably different in the last half-century. As an example, a form I was filling out the other day asked for the location of my bank branch, and it stopped me dead. I remember opening that account in 1982 on the corner of Fourth Street and Second Avenue in Manhattan with the first installment of my graduate school stipend. Larger financial institutions serially gobbled up that bank over the next decade, and eventually, I became the Bank of America customer that I am today. That original building may still be my branch. But given that I have lived in four cities since then, I do all my financial transactions online and call an 800 number when anything goes wrong, how would I know?

So to return to the terrific five-part series at Legal History Blog by Ronit Stahl and Mitra Sharafi, which not only parses the current system but tries to imagine solutions for the massive work it requires, I want to argue for change. And here it is: get rid of letters of recommendation entirely. I think they are completely pointless, and biased in favor of those who are already in elite institutions.

But LORs serve many purposes, so let’s start with the low-hanging fruit: letters for Semester Abroad programs. We can ditch these because I have never known someone not admitted to one. Why? Because they are money-making propositions, sometimes for companies that coordinate a network of university and college-sponsored programs. But they are also a moneymaker for the sponsors: a student from College A who goes to a program sponsored by College B takes a piece of the tuition and fees they have paid with them. And what does College A get out of this? The rest of the money, plus an empty dorm room that the college can sell to another student for whom they would otherwise not have room.

Semester Abroad programs only want to know only two things, neither of which has anything to do with academic accomplishments or requires a whole letter. First, and I quote, “Does this student respond well to those who are culturally different?” (Translation: is this student a racist/xenophobe who will cause trouble?) The second thing is far more veiled, but it’s a version of the same thing: is this student emotionally stable/have a substance abuse problem?

Next, letters for graduate school. Interestingly, Sharafi and Stahl were almost exclusively focused on letters written for people who are already graduate students. But in fact, the vast majority of us teach undergraduates. I never wrote a letter for a graduate student until about ten years ago, but my fall was often consumed with graduate and professional school letters, not just for the students I had but many who had graduated, sometimes years ago. “Dear Professor Potter, you probably don’t remember me . . .” some of these anxious little emails began.

Yet, what do three faculty letters written on behalf of an undergraduate tell an admissions committee that an official transcript, and the large number of application materials students have submitted, do not? While they may amplify on what the student has presented, what LORs mostly reveal to a graduate admissions committee is whether the student already has a stamp of approval from an important person in the field. But applicants who go to non-elite schools are unlikely to have such a letter, not because there is anything wrong with them, or because their teachers are incapable of evaluating them, but because that school’s faculty may be completely unknown to faculty at elite schools.

And here is the uncomfortable truth: letters from prestigious faculty may, or may not, have been written by those people. I know for a fact that one of my grad school letters was written by the Big Guy’s TA—and probably the other two were too. At least one of Sharafi and Stahl’s graduate student respondents admitted writing their own letters for faculty who would then sign them. This is not news, of course: students offered me this option when I worked at a SLAC, and they still do, now that I write for graduate students. Clearly, some of my colleagues were asking them to do it.

But think about it: many faculty requiring letters from faculty elsewhere seem to care so little about letters of recommendation that they don’t even write the ones they sign.

Although at least one of Stahl and Sharafi’s respondents argued that these letters replaced the cozier arrangements in which a guy at Amherst would write to a guy at Harvard asking for “a good man,” I don’t think that is entirely true. It’s fairly clear to me that letters from faculty at prestigious schools propel students on to more prestigious schools, more fellowships, and more awards. So what makes this labor-intensive system so different from the old system?

The letter of recommendation also shows us, in microcosm, how elite institutions—universities, foundations, humanities centers, think tanks—gate-keep for each other. Throughout an academic career, work speaks for itself, but LORs only show us which people are well-networked and which are not. Even within elite networks, Stahl and Sharafi argue,

using letters as a winnowing tool may well exclude the wrong applicants: letters favor those bold (or at least comfortable enough) to ask, not those who are the best or most serious candidates. It entrenches a system that favors the confident and those already in the know, not those who are anxious or learning how the process works. (Notably, even as a tenured faculty member at an elite institution, I find it stressful to ask for letters of recommendation.) If programs want the best and most serious candidates, especially those from historically marginalized communities, requesting multiple letters of recommendation at the beginning of the process may well drive them away rather than welcome them in.

And you know what else letters do? Because LORs are confidential, they allow famous people to stick a shiv in a student they don’t like. A recommender can tank an application by not submitting a letter at all, or by simply writing two or three non-committal lines rather than the conventional 1-2 pages. For example, as a search committee member, I once read a LOR from a Very Famous Man that literally said: “X was my dissertation student. Sincerely, A Very Famous Man.”

If you wondered, this situation has an official name: it is called having “a fish in your file.”

Interestingly, practically everyone who responded to Stahl and Sharafi seemed to believe that letters of recommendation were useful—but that we are asked to write too many of them. Many thought that LORs should only be required for finalists. Others thought that each application for a job, a post-doc, or a fellowship should only ask for one letter.

But would either of these or some combination be better? Or would it only address the question of faculty overwork, without challenging an antiquated evaluation system?

I think so. In fact, I would go further: faculty time would be better spent actually teaching students how to write their own job letters, grant proposals, and fellowship applications better. If an academic cannot describe their own research in compelling terms and persuade an anonymous reader of the importance of this work to the field, how can any letter of recommendation save them?

Stahl and Sharafi have done a terrific job of starting this conversation. But, don’t let it stop because we don’t just need to address faculty overwork. We need evaluation systems that have integrity, and that support equal access for all talented people.

Claire Bond Potter is Professor of Historical Studies at The New School for Social Research and co-Executive Editor of Public Seminar. Her most recent book is Political Junkies: From Talk Radio to Twitter, How Alternative Media Hooked Us on Politics and Broke Our Democracy (Basic Books, 2020). This essay first appeared on her Substack, Political Junkie.