Earlier in 2021, we--Ronit Stahl (Berkeley History) and Mitra Sharafi (University of Wisconsin Law)--found that we were both tweeting about the ills of the academic Letter of Recommendation (LOR) system. We were both dissatisfied with the general expectation that all applications (from jobs and postdocs to archive grants and summer programs) require multiple (often tailored) letters of recommendation and typically request them at the first stage of application.
We wanted to hear more from other scholars in law and history. Were they struggling with the LOR system as much as we were? How and why has the LOR system become more unwieldy over time? Most importantly, what can be done to make the system more manageable and worthwhile for everyone involved? (One notable difference between law and history is that law school admissions are run through a centralized system, requiring faculty to upload a single letter. Another is that unlike tenure-track history jobs, most tenure-track law jobs require recommendations by phone call, not by letter.)
We set up a poll and survey via Twitter and Google forms. We received 46 votes on the poll and 21 responses to the survey. We also received e-mails from three people, and had one in-person meeting. We share what we learned now, as a multi-part blogpost series. In parts 1-4, we describe some patterns and views emerging from our efforts at data collection. In part 5, we consider how the LOR system may be reformed.
The first thing to say is: *many many thanks* to everyone who took the time to share their experiences with us. Not only do many of you spend an inordinate amount of time writing LORs, you also then spent extra time answering our questions about them!
Secondly, it is not our aim here to make letter requesters feel guilty about asking for letters. Our academic world runs on LORs, although there are signs here and there that this may be changing. For example, TT job searches now occasionally only ask for LORs for finalists and some grant agencies are no longer asking for LORs. We are questioning the LOR system, not those who ask for letters. We recognize that in our current system, asking for letters is mandatory for anyone who wants to launch an academic career. But we wonder whether it needs to be so.
Now, to our results--after the jump.
The poll asked first approximately how many LORs (including online forms) the respondent writes per year. 69.8% said that they write less than 30 letters per year. 22.2% write 30-60 letters per year, and 7.9% write more than 60 letters per year. We also asked how many people the respondent usually writes for in an average recent year. Most (65.2%) write for fewer than 10 people. About a fifth (21.7%) write for 10-20 people, and over a tenth (13%) write for more than 20 people.
The survey consisted of seven questions inviting long-form responses, and six questions seeking information about academic position, stage, location, and discipline. We will present all responses anonymously to respect the wishes of a majority of respondents. About two thirds of respondents (approximately 67%) are in tenured or tenure-track faculty positions. 14.3% are students (graduate, law, or other). 9.5% are lecturers or contingent faculty, and about 5% each are postdoctoral fellows or emeritus faculty members. Most respondents (76%) are based in the US. And most (85.7%) are in History or Law. The three e-mails and in-one person meeting involved tenured professors in law or history at American universities.
In what follows, we offer a sample of qualitative responses. All questions were optional.
As in the poll, we asked in the survey how many letters respondents usually write, and for how many people. At the upper end of the spectrum, we heard this:
I have been a full professor at a research one institution for the past 20 years. I send in between 50-100 letters, for about 10 graduate students and former students each year. I also do letters each year for 10-15 undergrads applying to law school, which are through an online dossier system, although law schools are now shifting to individual letters through their own online system; about 10 -15 letters for various grants and programs.
I write for (on average) 6 PhD students a year, so it amounts to 80-100 letters a cycle...it's a lot. I write about 20 a year for undergraduates, for internships, admission to medical, law or graduate school. In the fall, it is usually an afternoon every week or every other week.
I am a recently tenured faculty member at R1 Institution, I write LORS for 20-50 individuals each year (these range from individuals who need a single letter for law school to those on the academic job market who I write 20-30 letters for annually)
I have written for close to a 100 students for internal university grants, which could have required a two line or one sentence approval.
As an assistant professor in a law school with a couple of cross affiliations, I have students across departments with a range of applications and I feel committed to their success and invested in being available to write / speak on their behalf. But at times it does not feel sustainable...I write/speak, on average, on behalf of 30-50 students each year. I'm also queer, POC, and am a faculty sponsor for more than one student group. This means I am invested beyond the classroom in a lot of my students' journeys.
I write...about 30 a year. I am a tenure track faculty member at an R1 private university, and my classes are popular. As such, I am approached ALL THE TIME by students for letters (who rarely give me any lead time). I wrote 7 just last week for undergrads applying to summer internships (and two more for graduate students who are not even my advisees for summer fellowships). I am also approached by graduate students who have taken classes with me or TA'ed for me even though I am not on their committees. They often ask because their advisors and committee members don't always respond to them. So, besides writing for my own students, I find myself writing for everyone.
Each new one is a struggle. I write for about 8-9 graduate students a year (fellowships, jobs, postdocs, occasional law school admissions, clerkships), 4-5 former undergraduates (declining numbers of course, law school admissions, grad school admissions), and two to three others in a year (colleagues and former students for fellowships or grants or new jobs).
As a law prof, at a school where not that many students do clerkships, I don’t do an overwhelming number of student recs. I have some undergrads and a few clerkship or transfer recs for law students, but most law job references are phone calls. I do get asked, however, to do an enormous number of tenure and promotion reviews, lateral hiring letters, and LORs for fellowships and awards for colleagues. It seems to me those are a really different thing. Both may be out of control, but they have different answers. [Eds: we are not addressing tenure and promotion reviews or lateral hiring letters in these blogposts, although it is worth noting that these kinds of letters are required of many letter-writers in addition to the LORs addressed here.]
What is evident across these responses is that, generally, faculty are writing a significant number of letters for multiple students and spending a lot of time writing (and submitting) them. Students understand the volume quite well. One reported, for example, “In recent years, I have had maybe 150 letters per year sent out on my behalf (mostly through Interfolio, so they are not all newly written letters).”
Some respondents described the wide range of contexts in which LORs are now required. In the words of one, “For undergraduates there are applications for study abroad programs; university and national fellowships; internships; law school and graduate programs. For law students I write letters for clerkships. But the majority of letters I write relate to graduate students for fellowships and jobs.”
Many respondents wrote about the amount of time they spend writing and revising LORs, and the amount of lead time they are given (often through no fault of the requesters). Here is a sample of these responses:
Once I have drafted the letter, which always takes me 2-3 hours, I feel like I have a good foundation.
I’m a slow letter writer. That first letter takes me at least 3 hours and often more. For subsequent letters, I work from the first letter (~30 minutes). But as a graduate student progresses, I'm often rewriting dramatically (more hours) and then tailoring letters specifically for TT, VAP, teaching vs. research postdocs, specific fellowships, etc.
I used to tell students they needed to give letter writers 2 weeks advance notice. But in fact they often can't because they don't even get 2 weeks notice that a letter is needed. And there are always new things that pop up and take them by surprise. I have never said "no" to a student that I write for. Undergraduates are often just less organized and so there too I often have to fit letters unexpectedly into my schedule.
Spend a day every month dedicated to this. However, this can't always be controlled or set by me.
In addition to the quantity of hours spent drafting and revising LORs, a number of respondents also commented on the problematic distribution of letter-writing work:
The workload is uneven at best...the letter of recommendation system creates extra work for certain faculty -- many of whom are women, people of color, or untenured faculty -- who students view as accessible.
LOR are a form of unrecognized and unpaid service that disproportionately falls on the same people doing other kinds of unrecognized and unpaid service: untenured faculty, junior faculty, associate professors, faculty in popular areas (in History, this means US and especially recent US history), and sympathetic faculty. These are all disproportionately women and people of color at my institution. As a "woman-like man" (as labeled by a feminist friend for sympathy to students) I write many letters, ranging from 10-40 a year, each being multiple pages and almost entirely bespoke.
[LOR-writing] puts extra pressure on those who are most likely to be thoughtful about it.
For many people, in other words, the volume of LORs required is immense and requires a significant investment of time and energy. This is true for those writing letters and requesting them. But was it always this way? In part 2 of this series, we share observations on how the LOR system has changed over time.