Refereeing proposed journal articles is a crucial aspect of an academic's task.  Yet to my mind it is undervalued and is becoming more difficult as time goes on.

At one time I would have received a brief letter from the editorial staff of a journal enquiring whether I would be willing to provide comments on a manuscript.  I would phone or write back and would then receive through the post a full copy of the manuscript, along with any general guidelines the journal had for referees' comments.  I would be asked to respond within 1 - 2 months.

Yesterday the editorial team of a journal contacted me by e-mail, asking me to referee a paper within less than a month. I e-mailed back to ask for a few extra days (given future travel commitments: the 4 hour round trip to London by train provides an excellent opportunity to give serious first-stage consideration to a manuscript).  They allowed that, but I was then faced with the task of registering and logging on to the publisher's archive of manuscripts and then printing out a 47 page article already formatted in a style that might be very useful to those making up the final pages of the journal but which render it very difficult to read.  I would argue that, given the cross referencing between different sections of a paper that one has to do when one reads an article, it is impossible to do this satisfactorily on screen: a paper copy is needed.  But the effort and cost of producing this has been shifted from the publisher to the reviewer.

I then downloaded the form that I am asked to use to provide my referees' comments.  Many of the questions that I must answer areg irrelevant to the nature of the paper itself, whilst other issues that need to be addressed have no place on the form.  I'm also asked to tick boxes with very black and white possibilities, and not given room to add nuances to my answers.  The editors don't want me to send back an annotated version of the paper - yet I know from past experience that many detailed comments are best made against the actual words in the manuscript.

So the editors are going to get a poorer set of comments from me than would have been the case a few years ago. They have shifted quite a lot of the administrative burdens on to me.  And I also know that because of the pressures to publish and the demands of the RAE / REF (and because publishers can see the possibility of pushing up their income) the journal has moved from 4 issues per year to 12.  FInally, I also suspect that for many articles today that are read by three referees, they will constitute at least half of the total readership of the article.

And if this particular article is eventually published, I will not be sent an offprint of the article or even a message to tell me which edition of the journal it is appearing in.

Yet the whole peer refereeing system depends on the significant input of time involved in this whole process.  I have actually stopped accepting refereeing tasks for one research council since their demands for a speedy turn around (gerenally 7 days for a proposal of up to 100 pages) are completely unreasonable.

At least one side benefit of the RAE and the citations fixation is the fact that of the 47 page manuscript I received yesterday, half of the pages are the references - I suspect largely to the works of friends of the anonymous authors (which reduces their anonymity) on a 'you scratch my back, I'll scratch yours' basis.