About Letters of Recommendation

Consider this a space for the discussion of various issues related to letters of recommendations. Here are three:

(1) The difference between U.S. and British letters of recommendation. I put the following in the Heap of Links the other day but a reader suggested it would be worth discussing. Here’s Jonathan Birch (LSE) on Facebook last week:


The reader added, “I think this practice genuinely hurts British candidates for US positions.” It’s hard to come by evidence of this, apart from the testimony of those on search committees. Has anyone had experience in making hiring decisions in which the hyperbole of US letters of recommendation / understatement of British letters of recommendation had an undue effect? Has anyone come up with a standard way of accounting for these differences? Perhaps something along the lines of the following chart, but for academics?


(2) The number of letters required. It’s not unusual for job candidates to have five letters of recommendation in their dossier. One can’t help but wonder whether that number brings us to the point at which we’re no longer making good use of people’s time (the candidate who must request the letters, the authors of them, and of course those who are tasked with reading them). I’m tempted to think that we’re at that point above three letters, yet I see how competitive pressures can make it rational for candidates to seek more, particularly if their authors are well-known. If there’s to be a remedy here it will be on the demand side. Departments could just say: “we will accept no more than three letters of recommendation; if more than three are sent, only three will be selected–at random–to be read along with the candidate’s other supporting materials.”

I’d be interested in hearing the justification for requesting more than three letters, as some institutions do.

(3) What is valuable and what’s not in a letter of recommendationIn an essay a few years ago, Jonathan Wolff (UCL) wrote, somewhat exaggerating:

unless a reference is three or four single-spaced pages then the support is regarded as somewhat half-hearted. But usually there is not that much to tell, and so we receive, in effect, six different versions of the candidate’s CV in prose form, combined with summaries of his or her PhD thesis, and a few lines of detail designed to convince that the reference writer really knows the candidate well and has been truly, deeply, impressed… For all the effort and ink, what, as readers of references, do we hope to find? Only this: is the candidate better or worse than it appears from their CV?

Maybe these things could be shorter. Wolff thinks that the “business paragraph” is the one in which comparisons are made:

The rule is that the more the reference writer sticks out his or her neck, the stronger the recommendation. If you say the candidate is good, that means nothing. If you say “in the top three of the cohort”, again, very little. But if you say the candidate is the best this year, or for several years, or for a decade, that means something.

Do readers agree that this is the most important part of a letter of recommendation, generally? Anyone have reservations about the value of these comparisons?

Other questions, concerns, complaints, and advice about letters of recommendation are welcome.


There are 18 comments

Your email address will not be published. Required fields are marked *

Please enter an e-mail address