I’m asked increasingly often to act as a referee, for jobs, grants or awards, or in promotion cases. In theory this should be relatively straight-forward. When writing references for individuals who request them, you would only accept if you know the person and their work pretty well. In those cases, having a copy of the cv, the application letter and the job spec makes writing a reference an important, but not especially time-consuming task. For some people it’s often just an updating of a previous letter.
But ones I’ve been asked to do recently include an odd request to write an evaluation of someone who I’d never met, never heard of, and who worked in a completely different field, or with unreasonable timeframes. If the short notice is the candidate’s fault – they’ve just discovered a new grant opportunity, for instance – then I tend to be less forgiving, because courtesy towards your referees seems non-negotiable. But often this is the fault of the institution, and then it’s really hard not to work by their shrunken timescale: to miss the deadline would seem to disadvantage the candidate, when it is clearly not their fault. In such cases I always meet the deadline, even when grumbling.
There is also the amount of documentation. North American universities are the worst at this. For tenure and promotion cases they don’t want a reference, they want an evaluation. There is a big difference in workload between writing a reference letter for someone you know a little, have heard give a couple of conference presentations, have read a few pieces of their work; and being sent a box by Fed-Ex that includes one or more entire book manuscripts, most of an edited volume, a big stack of offprints and various other things in order to write an evaluation.
This is not entirely unrelated to the comments I previously made on peer-review for journal articles. There I argued that there was an exchange economy going on: I produce the need for reports because I submit articles; therefore I will review articles. A similar case could be made for promotions: because people had to write on my behalf, I should write for others. It’s a form of academic service. Absolutely, and this is why I almost always agree to do them, do them on time, and do them with as much care as they deserve. But administration demands seem to be being ramped up – tighter timescales and more documentation required, sometimes needlessly complicated online forms – and it’s not clear that this is producing a fairer or better system. It’s not immediately clear to me that the letters hold that much sway, or, at least, that several additional hours of work on a letter and preparation for writing it (i.e. reading all the material sent) would make that much difference.