The Problem with Letters of Recommendation | Hacker News Reader

The Problem with Letters of Recommendation

thepointmag.com

42 pointsdiodorus40 comments

You can always take the Benjamin Franklin approach:

"Sir

The Bearer of this who is going to America, presses me to give him a Letter of Recommendation, tho’ I know nothing of him, not even his Name. This may seem extraordinary, but I assure you it is not uncommon here. Sometimes indeed one unknown Person brings me another equally unknown, to recommend him; and sometimes they recommend one another! As to this Gentleman, I must refer you to himself for his Character and Merits, with which he is certainly better acquainted than I can possibly be; I recommend him however to those Civilities which every Stranger, of whom one knows no Harm, has a Right to, and I request you will do him all the good Offices and show him all the Favour that on further Acquaintance you shall find him to deserve. I have the honour to be, &c."

https://founders.archives.gov/documents/Franklin/01-23-02-03...

A piece of advice I would give to students: If you're applying to N universities, have your professors write recommendation letters to accompany applications to N+1 universities (including a university that you're lying about wanting to apply to). Chances are, you'll get N+1 copies of the same letter in sealed envelopes. Open one. If you dislike what they write about you, throw them all in the trash and ask another professor.

Why do I recommend such an unethical course of behaviour? Because some professors are just nice people who want nice things for their students and who will write nice things in letters of recommendations. And some professors will have a character deficiency that will make them feel better about themselves by writing mean things about other people. ...in each case, the "personality test" around the professor has nothing to do with you, so why should you suffer the consequences if you misjudged someone's character.

I have never and would never give a letter of recommendation directly to the person I'm writing it on behalf of.

I've attended four universities, and all of them had the letters of recommendation addressed directly to them - I never got a copy of the letter.

Really? I've always had them handed to me. If a professor refuses to hand you the sealed envelope, maybe that's the red flag right there. Even so: There's probably a way to game that system, too. (Handing them pre-addressed envelopes, and one of the addresses is a PO Box you own?) ...although you have to be careful, because at some point it presumably becomes mail fraud.

...I should add, that I didn't do this myself, back in the day, because I was lacking in the worldly wisdom required to have this idea. But thinking back of it, I think it's probably what cost me admission to the places I would have most wanted to get into.

> Really? I've always had them handed to me. If a professor refuses to hand you the sealed envelope, maybe that's the red flag right there.

Sending it directly is the norm in my experience, and it's directly intended to avoid the "hack" you suggested.

> Even so: There's probably a way to game that system, too. (Handing them pre-addressed envelopes, and one of the addresses is a PO Box you own?) ...although you have to be careful, because at some point it presumably becomes mail fraud.

That would work for finding out after the fact who has written you a poor recommendation, but that's not particularly useful, since by the poor recommendations arrive at your PO Box and the other receivers at the same time.

> If a professor refuses to hand you the sealed envelope, maybe that's the red flag right there.

As a professor, it would be a red flag to me if a student insisted on my handing it to him/her.

...you have to be aware that this business with the sealed envelopes and stuff that exists in the anglo-american cultural sphere is not practiced throughout the world.

One day, I asked my prof for a letter of recommendation when applying to a scholarship by an Austrian funding agency which asked for the recommendation to be written in cleartext, readable by anyone handling the application form, including the actual applicant. I had explained this to her, but I guess that by the time she got home she had already forgotten.

She handed me the letter of recommendation in a sealed envelope. I opened it to transfer the text into the form. When I saw the way she was writing, A LOT of things started to become clear to me with regard to past application processes that carried letters of recommendation from her. She wasn't trying to be mean or anything, but it was just so obvious that she regarded this as a tedious duty. She didn't put the slightest bit of effort into trying to remember any of the achievements of mine that she had been witness to and could have mentioned, and was being exceedingly British in the way she was tempering her style, and basically only mentioned things that were objectively provably true (so that I was mentioning them elsewhere myself anyway without any questions around credibility if they were coming from me). When a letter of recommendation tries to say something negative, it will mostly say it by way of omission, and the omissions in her writing were huge and gaping.

So, let me try to rephrase my initial premise in a less unethical way: If I were a Professor, I would make a point of showing the recommendation letter to the student, and go "This is the letter I'm going to write, let me know if you want me to submit it."

In some places there will just be an esprit de corps mentality: People handing positive recommendations to each other whenever requested, unconditionally, and everybody benefits. -- It's just a competition where any cluster of persons with low ethics end up creating an advantage for themselves. So, game theoretically, the only rational course of action is to throw overboard all considerations of ethics.

In that sense, I like the Austrian system a lot better because it doesn't create the pretense that the letter of recommendation is not the result of careful selection, when the Ango-American system creates the pretense of ruling out selection when this can actually not be guaranteed.

The whole thing also carries a very negative flavour with me for another reason: It always conjures up in my mind a picture of victorian-age London and its caste system. I imagine some rich families employing domestic servants. And I imagine some poor people who have to work as servants. The system entailed that when a servant wanted to leave the service of one family and go to work for another, then the new employer would ask for a letter of recommendation from the previous employer. The idea was that if a person was caught stealing or anything, then a letter saying that this person is a thief would follow that person around and make them unemployable. It is easy to imagine how this serves to cement the caste system and prevent social mobility.

I can't think of a single one of my fellow PhD students who didn't have a very complicated relationship with their supervisor, and I think it's a pretty horrible system if the deficiencies in your relationship with your previous employer follow you around to your next. -- It's actually one of the reasons why I dropped out of academia.

If there is an argument that letters of recommendations are arbitrary and basically political game student can't fully influence, it is this post.

I'm not sure why you wrote all this in response to my comment. I think you meant to address it to someone who is in favor of letters of recommendation.

Agreed, I have also never dealt with letters handed to the student. Nowadays the letters of rec are usually online - the student can send the professor a link to use.

These days, recommendation letters are all electronic.

It is possible, sometimes, to game the system. One year I served on a hiring committee while applying for a handful of other jobs. Although I didn't, I could have read my own letters by applying for the job opening in my own department.

Even if you don't do that, rec letters sometimes say things like "X is better than Y", with predictable results if/when they get read by Y.

As a professor, every recommendation letter I have written (or that was written for me) is never seen by the student. The letter writer is contacted directly and given a URL to upload the letter.

I've read a lot of recommendation letters too. None have been bad, but you can definitely tell when the writer doesn't know the student well when all they really say is "This student received a good grade in my course." I've seen a few highly opinionated letters that openly discuss a students weaknesses, but also sing their praises regarding their strengths. Those letters felt the most genuine, plus the students were all-stars.

Advice I've seen amongst professors: Just ask if they would be willing to write a strongly positive recommendation. Few professors are going to say "Yes" and then write a negative recommendation. You are not worth their time.

No doubt pathological cases exist, but they are not the norm.

This is an example of a phenomenon in our culture: people would rather lie to appear to conform to social norms, than have an honest, human conversation.

When I ask someone to write me a letter of recommendation, I ask them "Will you write me a positive letter of recommendation? I'm not asking you to lie for me, but if you don't want to write something positive, or just don't want to write me the recommendation, I want to know so I can ask someone else."

I've been doing this since high school (I'm 33 now) and I've had one person say "No" to me in that time. There have been many cases where people haven't offered me interviews or whatever, but I have no reason to believe that was because of letters of recommendation.

Typically you waive your rights to view your recommendation letters. With FEPRA, I believe you can review recommendation letters a year after you applied as schools may keep them on file. However, I believe if you did not enroll at a school, they are typically destroyed.

Recommendation letters are candid responses from teachers/professors about your abilities and what you will bring to a school. By waiving your rights, it allows them to be more candid in their writing instead of stating facts such as grades and other items readily accessible on your transcript.

Having written letters of rec as a school counselor for 8 years, it is a daunting and tiring process. However, the best letters I wrote were because the students befriended me and I was able to write them without them providing me any information.

Man in high school I needed a letter of recommendation and went to one of my teachers. They said yeah they'll be happy to write me one. They had about a month and a half to do it.

Kept checking in every few days and they kept saying they'll get around to it. Didn't want to rush them but as the weeks went by I was sorta running out of time. The last possible day I could get it out in the mail I went to him after class and asked if he had it, looked me right in the face and in front of a bunch of classmates said "You know what, I for the life of me couldn't think of anything good to write about you."

Man I was humiliated. I still remember the look of disbelief some of my classmates had. Some time later after thinking about it I got pretty angry because it sure seemed like he was trying to sabotage me, a high school teacher should know that a letter of recommendation was required to get into college. It wasn't even a particularly great college either, a dinky state school out in the boonies. I'm pretty sure you don't even need to say much in the letter. At the very least he could have said no, or said he was too busy weeks earlier. Simply told me he didn't know me well enough.

It was partially my fault. I was very naive for a large part of my younger years. I had a weird feeling that this guy didn't like me. But it was always just a feeling. We never argued and I never talked back or disturbed class or anything like that. But there was just something off about how he interacted with me that I always dismissed as being in my head because I didn't see any reason for him to hold anything against me. I was naive because I had no idea that there definitely are people out there looking to screw you over. Not only do they want to see you fail but are willing to help make it happen.

I don't see your suggestion as unethical at all. It's great advice. I did finally get that last letter of recommendation from another teacher. One that I only had taken a few classes from in the past but he was more than happy to write something out at the last minute and I did end up going to that school.

All of that may be true - but seems crazy that your first choice for a letter of recommendation was someone who you had a vague inkling that they didn't like you.

Shouldn't you start with someone who you know - or at least have the vague inkling - that likes you?

I really, truly, deeply dislike this author. This is the sort of thing which sometimes happens to academics -- in a position of absolute power over their students, divorced from any sort of checks-and-balances by tenure -- they become mean people. Over time, that leads to mean cultures, with increasingly abusive advisor-advisee relationships across entire fields.

This is most common in fields where a Ph.D is required to do meaningful work (e.g. biology, philosophy, etc.). It is especially common in the humanities, where many (most?) suffer from imposter syndrome (in many cases, rightfully) so this gets tied up in ego issues too.

On the receiving end of recommendation letters, I can say two things:

1) They're almost meaningless. Most are glowing. 2) They reflect more on the writer than on the student.

That's something students don't (and based on information they have, can't) realize. If they pick the wrong professor to impress -- someone like Agnes here who applies a philosophy of "hyper-rationalism" to recommendation letters -- their career is shot.

For the most part:

* If you're applying somewhere, I recommend talking to many recommenders. Ask whether they can and will write a strong recommendation. If so, engage with them. If they don't answer, don't take a recommendation

* If you're looking for a professor for graduate school, talk to their students first. Look over where they are now. A state school with a good advisor will beat an ivy league with a bad one, hands down.

* If you're looking for an undergraduate projects, do the same. The stakes aren't as high, but you'll do a lot better if you chance on someone who cares about you as an individual and works to make you successful (which is their job, as a professor) rather than someone who places their career first (which is the reality, of most professors).

If you want to take a hyper-rational approach to it, think of it this way: if better students students pick ethical, humane professors, you're giving ethical, humane professors an edge in the academic rat race. And you're giving sadistic bastards a disadvantage in the rat race. You're making the academy a better place.

I think the writer simply wanted to appear very smart. The boiled-down TLDR is

1) A profesor should write a letter of recommendation for you. It's part of their job, not a favor.

2) If students believe the teachers know them deeply and the teachers deeply care about their studies, the students'll study better.

Overall, I agree with your assessments of "letters are meaningless." I also think that the author's views are somewhat characteristic of the life in academia. I don't, however, think the author means, or does, harm. It's more of a 'meh' situation.

I think you’ve completely misread her. She is saying that she doesn’t know what her students are capable of or what they can do, so she doesn’t tell them her honest opinion — she merely writes the recommendation letter based on what she knows objectively about the student. If she were to be honest, she’d probably direct some students away from life decisions that they’d have been successful and happy with if they had taken her misguided advice.

I've always been disappointed by people's consistent refusal to voice opinions about my "career" decisions.

I'm perfectly comfortable dismissing feedback and advice when I feel that is the best course of action. I want other opinions.

It's very difficult to get anyone to give feedback in a meaningful way or opinionated career advice about what I might be well suited to.

This was particularly true of college professors.

Helpful sentences would start with "I think you would be good at..." or "While you might be able to improve, I think you should steer away from ..."

Agreed - particularly in my early career I wished more experienced folks would tell me what potential they actually saw in me, rather than constantly being asked what I wanted to do.

In particular, this kind of withholding of advice disadvantages people who are first generation college students, from the wrong social background, or lack access to networks. People who don’t know what their options even are, let alone which of those options is suited to them. Denying those people advice is especially cruel.

I have been asked to write all my letters of recommendation myself, and the professor/employer will simply make changes and print it on university/company letterhead.

I know what details are important to highlight in the letter, and have always been met with the logic that "therefore it only makes sense for you to write it".

But we all know the reason is effort. They have "better things to do".

That just means that the letters from your institution carry no value to the letter readers. The letters only carry weight when the people behind them have credibility

This is the best kind of recommendation letter. Who knows you more but yourself?

Even if I'm super happy with someone's performance, a bullet list of the achievements they want to highlight is extremely useful. Heck, I have a hard time remembering exactly what I accomplished in the last year, it's super risky to assume your recommender will remember for you or happen to mention what you think is important.

It's like... I keep a bullet list of things I've accomplished at work and add to it when I feel like I did something worth highlighting as especially helpful. Then during my performance review I've got more in active memory than to discuss what I did last week.

Plus your recommender won't know the context of how you're pitching your application. They don't know the other recommenders you picked or what story you want to tell, even if they have the luxury of enough free time to write something personal.

As a recommender, the one that I find most obnoxious is that they usually ask you to "rate people relative to others you've worked with" which is just a ridiculous question. I'm at MIT, if I find someone seems to me like they'd fit in as a first or second year grad student at MIT but maybe in the lower half, do you actually want me to write down that they're below average?

It's pretty clear that it is not the question you intended to ask when you don't bother to even ask me what kinds of people I've worked with before. You can't ask people to numerically rank performance when there's never been an agreement on the scale! And knowing that you're likely using it as a first pass filter makes it even worse; what, are you going to somehow correct for the difference between the people I've worked with versus everyone else according to some magic objective scale?

Even if you know that I'm comparing to MIT students, do you realize MIT has 10000+ students with skill levels all over the map? Do you know that I value practical skills more than most because I've seen what makes someone successful in lab? It's just a part of the recommendation letters I wish they'd just leave out. Filter based on GREs or something else that while equally pointless is at least uniform and then just read the letter and think what you want. Don't ask me to do half the thinking so that you can do none.

Sadly, that hurts talented great people who are modest or bad self promoters.

If you know you're bad at self-promotion, ask a friend or family member who is good at it to write the letter for you. If you either don't know you're bad at self-promotion or don't know anyone who is and would be willing to help you, you're probably not as much of a "talented great person" as you think you are. To me, this is the system working as intended.

I've been asked this when requesting a letter of recommendation, and honestly I find it a bit insulting when I worked with the person closely.

I see writing letters of recommendation as one of the responsibilities that comes with more powerful/prestigious positions. If you work closely with someone, a recommendation request is a way of giving back and helping further someone's career. If you ask me to write my own letter, it implies that either you haven't paid much attention to my contributions or that you don't think that I'm worth your time.

To be clear, I'm talking about cases where the person should be familiar with me and my work, such as a direct supervisor or research advisor. If I'm asking someone who isn't familiar with me, then I think it's far more fair to ask me to write my own letter.

I always enjoyed one of my friends grad-school supervisors approach to letters of recommendation:

"Such and such was adequate for my lab therefore they would be more than adequate for your lab."

I've been thinking about this from the perspective of the teachers who can be asked to write dozens of letters a year.

After some interviews, my hypothesis was that having students be more involved would lead to more engaged teachers and better letters.

So I started with this ( https://tinyrecommendations.com/ ) interactive blog post which walks you through the steps of requesting a letter (and even generates an example email to get you started).

As far as I can tell, these letters are just a mark that a candidate has enough social hygiene to be able to find a few people to say nice things about them. (Or perhaps are psychopaths.)

This system is utterly broken and useless. My response is to write glowing recommendations for anyone that asks me.

>My response is to write glowing recommendations for anyone that asks me.

I find this to be very unethical.

LoR's should be an honest accounting of the merits, flaws, and accomplishments of the subject - writing a strongly positive LoR for a subpar student harms the community by subverting meritocratic admissions, may harm the subject by placing them into a position they are unsuited for, and can possibly harm other, more qualified students by having the subpar one occupy a slot in a place-limited program that would otherwise be filled by a "better" student.

It may also harm your own career and students - if you get a reputation as someone who always gives out good recommendations even to subpar students, then it may reflect poorly on you. More importantly, it may harm your students - if you're known for giving good LoRs to mediocre students then it becomes impossible for you to give a good LoR to good students and have them taken seriously.

I urge you to reconsider this practice.

That's a reasonable position. But I've seen too many cases of people being left behind in the reference game due to simply being unusual in some way: shy/awkward, wrong demographics, uneven career, etc.

I believe in redemption and in judging people where they're at right now. If they can demonstrate their current skills and accomplishments, lack of refs shouldn't hold them back.

As for reputation, that ship has sailed. My refs aren't worth much, but for someone who just needs the box checked, I'm happy to do it.

(Most of my refs are for industry, if that makes you feel any better.)

I was once asked by a US company for "recommendations", that is people they would call to ask about me.

I am French and it is illegal to call you previous employers, so I guess this was the best they could get.

I was genuinly interested in that thing because the idea of such recommendations is so unbelievable to me that I must be missing something.

How can a company expect to receive anything else than close friends who will describe some kind of crossover between a genious and a saint? With, of course, some weaknesses such as the tendency to be too good at his work?

Why don't schools scan letters or request them in electronic format and use NLP to see if letters complement grades and test scores in predicting future grades and graduation rates?

Because, simply, we don't have that advanced NLP systems.

University professors (I am one), at least in my discipline (math), tend to be diehard traditionalists. For better or worse, we tend to distrust any formal assessment mechanism.

Reply on news.ycombinator.com