Are recommendation letters a bad thing?


In the past couple years, we have made progress in dropping the GRE, and now the pandemic has come through with a huge assist. Maybe that’s the final blow for the GRE, as programs are now dropping it permanently. Which means that people who cared about the GRE are now placing higher importance on other pieces of applications, including recommendation letters.

Which leads me to ask: Are recommendation letters a good thing or a bad thing? Of course, I don’t think this is a binary matter and there’s a lot of nuance involved here.

I have heard a variety of concerns about recommendation letters in the graduate admissions process, and I think it would be foolhardy for us to think that we’ve made a big amount of progress by getting rid of the GRE. The problem with the GRE is that it doesn’t reflect ability, research potential, or value to the academic community, as it’s more tied to wealth, access to resources, and the accumulation of cultural and social capital. You could say the same thing about recommendation letters, too.

Let me illustrate the problem with recommendation letters with a little, and wholly true, story.

An undergraduate Biology major from a regional public university participates in a Summer Research Training Program at a highly prestigious institution. They succeed during their REU, and the PhD student who worked with the undergraduate is impressed by their research acumen and potential to excel in a graduate program. After the undergraduate is headed back home, the PI of the laboratory refused to sign a letter of recommendation for the student.

Why? The PI said that the undergraduate “didn’t have the adequate academic preparation” for grad school.

This is really puzzling, because the PI simply had no basis to make this assessment. They might have seen the student’s application to the summer program before they were admitted (which featured a quite respectable GPA), and during the summer, the PI was occasionally in the vicinity of the student. But really, the PI didn’t know the student at all, didn’t work with them, and the only information the PI had was the student’s Latinx identity and where they went to college.

I’ve been living with this story for some while, and it still upsets me. Part of what makes it so upsetting to me is that it’s so pedestrian. It’s the kind of thing that happens all the time. It’s something that people just live with and move on. Living with and moving on with this kind of bullshit is what perpetuates the inequities that, all of a sudden, everybody is having zoom meetings and book clubs and departmental committees about.

This story illustrates a lot of what can be wrong about recommendation letters. While we can take steps to minimize biases while we are reviewing an application, we can’t do much to prevent the biases of the letter writer from manifesting themselves on paper. (Or in this example, not even sign a letter at all.) If we take the presence of positive stories and praise in a letter as a positive, then when comparing applicants, this gives a relative advantage to people who don’t have letters are as good. Since not everybody has equitable access to “good letters,” then what good are they?

People with social capital and access to the hidden curriculum know that it’s savvy to cultivate relationships that will result in strong letters of recommendation. And folks who haven’t mastered that hidden curriculum might be just as spectacular, but don’t have those letter-writers ready to go. Just as the GRE is a problem because the test rewards those who are supported to become excellent test takers rather than excellent scholars, are recommendation letters a problem because they reward students who are better schmoozers and know how to impress their letter-writers more? I know we all would like to think we are excellent judges of character, but I also recognize that there are some students who probably deserve quality letters that I just don’t know very well because they didn’t cultivate that relationship with me, and I’m not capable of writing those quality letters. (And this intersects with the problem of unpaid research internships, because this gives access to those letters too, to those who can afford to work for free. How often is it that a quality letter of recommendation is dangled as a reward for labor?)

Another problem with letters is that people often put weight behind the impressiveness of the institution on the letterhead or the name on the signature line. If a recommendation letter comes from Harvard or UCSF or somesuch, does that really mean that the applicant has better chops or is more worthy than someone else? If it’s signed by a Nobel Laureate or a National Academy member, what does that really mean about the applicant? To a lot of people, those things matter. Which is, in my view, another drawback to using recommendation letters.

But, on the other hand, I also see that letters of recommendation are a way for me to describe so many of the positive attributes of a student that might be hard to communicate but are still valuable. And they allow me to advocate for students and be more direct about some challenges students have faced in ways that they might not be positioned to advocate for themselves.

I hope that a bunch of y’all do trust me and my judgement about my students, and so if I explain their talents and capabilities and talent in a letter of recommendation, I sincerely hope that this sways you. That’s why I’m writing the letter. And if you don’t know me or my work, then I would hope that my words themselves should still be taken at face value.

Just like some folks think the GRE can help students who have experienced academic and personal challenges (by scoring high and explaining low grades), maybe the recommendation letter serves this positive function as well.

That said, when I’m looking at applications (for anything), I think recommendation letters aren’t that helpful, when I’m dealing with people beyond the scope of my own department. I’m not going to penalize an applicant when someone is “missing” something in a letter, and it’s very rare that anybody from the US chooses to share a concerning piece of information. While there might be some information about an applicant that allows you to see how they might be well aligned with the expectations of the position, that this is as much the product of the letter-writer’s acumen as it is about the qualities and capacities of the candidate. I would much prefer a 5-minute phone conversation with a reference over a letter of recommendation. If there’s a program that has 100 applicants for 10 slots, I think it would a big mistake to rely much, or if at all, on the letters of recommendation to pluck those at the top.

So, then, if letters of recommendation are a problem, and the GRE is a problem, and you can’t judge a person based on how prestigious the university that they got into when they were in high school, and experience on a CV is more of a product of opportunities provided to person than anything else (and if grades often don’t tell much of a useful story because of intersecting issues), how can we evaluate people who are seeking opportunities? That’s a hard question. I think it’s one we need to sit with for a while.

6 thoughts on “Are recommendation letters a bad thing?

  1. I did my undergraduate and master’s in a different country. Criteria for undergraduate study applications were high-school and national exam grades. Criteria for graduate study applications were undergrad GPA and references. I think the former were more fair criteria than the latter. So even though I hate taking exams and my GRE score was not great, I think it was a valid criterion for my PhD application, together with the essay I had to write. I did have letters of recommendation but I don’t know how important they were or should have been – since they were not from American sources, they didn’t say I excelled at everything.

    For me, the answer to your question is excellent public education in grade school which will provide better opportunities for all. That is not, unfortunately, the state of education in the US at this time.

  2. I have found a lot of wisdom in Freddie deBoer’s writing over the years and looking forward to digging in to his new book “The Cult of Smart”. The main thesis of his book is not fully germane to the question of graduate school admissions, but some of the ancillary ideas might be. In a recent podcast he talked about how our education system is centered around two expectations that are incompatible in many ways: 1) schools as drivers of social mobility and equality through equal access to education and opportunities; and 2) schools as meritocratic sorting mechanisms to identify the “best” students for university admissions, jobs, etc.

    Because grad school admissions are competitive and limited, point #2 will favor applicants from the most prestigious schools, with the most access to research opportunities, and with the most access to people / knowledge that can help them succeed within the system. People with means will always find ways to be more “competitive”…almost like a Red Queen dynamic where we might try to improve the “objectivity” and fairness of the grad admissions process but certain applicants will strive to find new ways to win the game. Theoretically a test like the GRE would identify students with the intellectual abilities to succeed in graduate school even if they haven’t been able to access these other resources. That was the original idea behind the SAT for college admissions. I’m not fully convinced that these tests are as evil as they are made out to be, but we as a society seem to be moving on from them so maybe not worth even discussing any more (although I find it odd that alumni and aspiring students to places like NYC’s specialized high schools are still deeply committed to such tests!).

    I fully agree with Terry that removing the GRE was not the great victory that many people seem to think it is. The system is not substantially more fair than it was before. It might even be worse! I don’t really have any great insights on how to improve the process. But a few things that would probably help: 1) restoring and increasing government funding for community colleges and regional, public higher ed systems like CUNY, Cal State, rural branch campuses, etc. Something on the scale of the land-grant universities. 2) I’ve generally been in favor of affirmative action programs that would help admit people that are not the traditional “winners” in the current system (with appropriate campus support to help them achieve success).

  3. I think letters are a bad thing. I recognize the issues with the GRE, but based on my experience as a faculty member at a regional public school, I think having to rely heavily on letters will be even worse. I’ve written so many lackluster letters for students I barely know. I try to meet with students and ask them to send me info about themselves to help write better letters, but there’s only so much you can do when you’re writing about someone you only know because they took a large class with you and you’ve never really interacted with them. The easy response is to refuse to write these letters at all, but I know they’re only asking because they don’t have anyone else and they need three letters to apply to a graduate program (admittedly, more of these are for various health related professional programs than Ph.D programs, but some are for masters and many are for REU or other research programs).

    Most of my students are first gen students who just don’t understand how to make these connections with faculty members until it’s too late and our classes are too big to make it easy. It also leaves students very vulnerable to bad advisers. I had one student who came to me at the last minute after her research adviser kicked her out of his lab for refusing to quit her job and work in his lab full time in the summer, a job she relied on to support her family. She needed a letter within a few days, and despite the fact that I had never even spoken to her before that, agreed to write the letter and that was enough to get her into her professional program, mostly thanks to her excellent GPA and test scores. That story had a happy ending, but I don’t know how anyone can think that’s a system that’s working well. I would be shocked if getting rid of tests benefits students like mine and doesn’t just make the already extremely problematic power dynamics in academia worse.

  4. I strongyl feel reco letter is a brand. If your brand sells high in the market (“celebrity” scientists” or institutes) then surely you have added confidence in your application. But this may not positively correlate with the students’ research ability, perseverance always. Some students cannot “show off” their skills like others. Clearly the “showy” phenotypes are more likely to be selected than those who are shy, embarrassed more hard-working but lacking communication.

    I strictly feel that reco letters are arcane for today’s world. I agree a hundred percent with Terry on this. When you can use video conferencing or short telephonic conversations with the student/ employee himself why rely on somebody’s say.

    Like the GRE, reco letters should be in a state of toss sooner (I feel).

  5. This is a problem with systemic racism and not so much with recommendation letters per se. Yes, letters have their own set of problems, including that very few of them are even informative (mostly just bad ones and ones that concur with other writers about some extraordinary feature of the applicant). Letters are a particularly common manifestation of the racist culture in academia, and ‘population’ biology in particular, and the story you describe is textbook discrimination.
    Consider that California has 15-20 million latinx and Black residents and are 41% of newly admitted students to the University of California this year. Our premier academic institutions (UC, Stanford, Caltech, USC, etc) have about 5000 faculty in Life Sciences and over 1000 faculty in population biology departments alone. By my last count, which was admittedly incomplete and biased, there are less than a dozen Black and latinx faculty in those departments. That number increases slightly if you include foreign national, overseas trained faculty of color.
    The numbers are higher in molecular/cellular/biomedical departments, but the proportions are only slightly higher.
    Houston, we have a problem.

Leave a Reply