hold that letter!

I learned today that the American Anthropological Association has the following rule regarding searches its advertises: “Solicitation of letters of recommendation should occur only after an initial screening of candidates to minimize inconvenience to applicants and referees. Names of references may be requested, however.”

Sociology doesn’t have this rule, right? Should it?

Author: jeremy

I am the Ethel and John Lindgren Professor of Sociology and a Faculty Fellow in the Institute for Policy Research at Northwestern University.

11 thoughts on “hold that letter!”

  1. As a relatively recent applicant, the letter component was always the most stressful for me. I knew that out of the 40 or so places I applied many (most?) of the places wouldn’t even read my letters. And while secretaries did most of the school-specific modifications, I still felt like it was a burden on people I hate to burden (advisers and their staff). That being said, even if the rule existed I would always list who my letter writers are. I have only anecdotal evidence of this, but name dropping does seem to matter.
    Soon after I got my job offer I (and a few other recent job candidates) jotted down a list of things we would change when we were in charge. Luckily a few of the issues have been resolved with a better functioning wiki and rumor mill and greater reliance on internet and email. Considering this was only a few years ago I am hopeful that by the time I am a search committee chair most of the issues will much smaller. Of course, I may also care less, which may explain why these issues remain year after year.


  2. I think letters are a HUGE help to people who are not at the top ten depts. A great letter or two from well-respected people at more “minor” schools is what makes a candidate from UMass or Illinois-Chicago credible (along with some pubs etc of course). There is too much screening by school already IMHO and waiting on letters would make that worse.


  3. Absolutely not, and I’m thinking of it purely from the perspective of the letter-writer. Writing letters of rec has a high fixed cost and a low marginal cost. The only way this saves time is in the case of candidates who don’t make any short lists.

    Furthermore, what marginal cost there is to sending out the letters consists of dealing with the idiosyncrasies of various personnel computer systems. (“I didn’t get the URL to submit my letter for you, it might be in my spam filter, can you request that they send me another email?”). The best thing for the letter writer is to get a list of snail mail addresses all at once at the start of the hiring season. Anything that involves requests for letter trickling in over time makes for more hassle, not less hassle, even if it involves fewer letters in total.


  4. Responding mostly to Gabriel:
    1) The list of names/addresses cannot be fixed at the beginning of the year because job postings trickle in all year.
    2) The marginal cost of sending a paper letter is 50 cents including postage plus the labor cost of a staff person which, assuming a letter can be handled in a minute by an underpaid clerical worker who costs $32,000 a year including salary and benefits, works out to about 26 cents a letter. I’m not sure the labor cost really can be a minute, the person would have to work very efficiently. So let’s say somewhere between 75 cents and a dollar per letter as the cost the department has to absorb after the professor has written the letter. That’s $30-$40 for a student who applies relatively unselectively to 40 places. I guess whether that is a little or a lot depends on your department’s S&E and staffing budgets and the ratio of the number of graduate students on the market to that budget.
    3) But I agree that for the professor, almost all the cost is the fixed cost of writing the letter in the first place, and all tweaks thereafter are small marginal costs.

    To the main point: On balance, I think that the “letters” requirement is a waste of money and time. I am certain it is so for applicants for more advanced positions. For new PhDs I agree there is more information in the letters, especially for people who have not published or written anything.

    My own view is that a much more useful requirement is writing samples, which actually provide germane information about the quality of the person’s work. I realize it is more work to read and evaluate them and much more expense for the candidate if applications are on paper.

    My ideal would be electronic applications with links to writing samples.


  5. As a current jobmarketeer I can say that organizing things in this way would have saved me and many of my peers considerable anxiety and lost time. Because of the great diversity of the ways in which departments ask for these letters- and in my case because of one not-so-on-the-ball letter writer, the experience has been very frustrating and stressful.

    Having said that, though, in this market, I want ANY advantage that will get me in front of a committee or at least out of the middle of the stack. I would hate to have my chances diminished because my strong letters were not immediately on hand. From a broader perspective, I’d hate to have this particular change privilege or disadvantage some candidates relative to others- and I’d withhold final judgement until we could be sure that that was not the case.


  6. i am on a SC and my dept head is an anthropologist…it is saving a lot of work to not have to read all the letters of rec because he is following the AAA norm


  7. I’m with Gabriel on the hassle factor, but that’s partly because my department doesn’t provide admin support for letters of recommendation. And on-demand letters does stretch out the period when requests for letters are trickling in, getting lost in e-mail spam filters, etc.

    One issue that hasn’t come up yet in the comments is timing. A quick look at the Anthro job bank (which nonmembers can access, for free!) suggests that their junior market is in spring. I’d guess that this gives them a bit more leeway for a two-stage process, if for no other reason than they can get permission from deans and the ads posted the term before. In our (soc) searches, at least, the timing is always tight enough that an additional week (two week?) delay to obtain letters from long-list candidates would make it unlikely that we could get offers out before the winter break. And, assuming a fall market, pre-break offers are desirable for all concerned.


  8. I think letters should matter in the deliberation before a short list is completed. While most are simply bureaucratic propaganda, others are quite revealing and can completely swing committee rankings. I agree with Gabriel that the costs for letter writers are initial—I write a letter for a candidate in August, and it doesn’t really matter how many places I send it. Indeed, the “short list letter” could create more problems than it solves. Many of our colleagues are not very good at managing letters, and I have seen candidates screwed because letters didn’t make it in time. Backing up the deadline also increases the delay for interviews, which may not matter at top places, but it matters a great deal for less prestigious schools which can benefit from speed (and can see positions disappear into the budget hole if delays are protracted).


  9. So upon reading comments and thinking about issues, I’ve changed my vote to thinking letters are good thing to have in hand for entry-level searches (where there is usually not much else to go on) but: (1) I think still letters at application are a waste of everybody’s time for more senior searches where the person has (or should have) a track record. (2) Departments whose main focus is research should ask for links to electronic copies of papers and publications (not paper copies) or should liberally ask to have them emailed, because reading people’s work is truly the only way to give good people from lesser schools a fair chance. (3) I really wish our hiring happened later in the year, more like anthro. Sept is just too early for many reasons. If I had a magic wand to wave, I’d get us all to agree that nobody will interview for entry-level positions for new PhDs before November or even February. (February is too late for more senior positions, due to the May 15 offer deadline.)


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.