Reviewing #6--How to make editors love you, or not

OK, you've accepted the assignment to review a paper.  I've already outlined the elements of a good review and the etiquette of turning down a review request.  Now I'd like to discuss the etiquette of reviewing.

Don't accept a review assignment unless you are positive you can do it within the time limit set by the journal.  To ensure that you do, set aside a time--put it on your calendar, even--to do it.  Don't fall into the trap of assuming that because the deadline is 30 days out, you'll surely find the time.  In fact, I would recommend having a self-imposed deadline of 2 weeks.  It's easier to look 2 weeks into the future than a month, you will get it off your desk faster, and you'll avoid some of the tactics used by some journals to keep the review process moving (see below).

Here are some of the things that can happen if you are late with your review:

1. Decisions are delayed.  Put yourself in the author's shoes.  You've submitted a manuscript, and you pretty much expect to not hear anything in a month.  Then two months go buy.  OK, you say to yourself, that's still within bounds.  Three months go by, and now you are really sweating what's happening to the manuscript.  You've put a lot of work in, and a quarter-year is gone.  If it needs to be revised, and you've got field season coming up, you don't know if you'll have time to do the revision (revisions often have deadlines).  Worse, you think another group might be onto the same discovery and you might get scooped.  And so on.  It's just not fair to the authors.
2. Meanwhile, the editor is tearing her hair out.  The other review is in, and it's a fairly positive review.  What if the second review--yours--is really negative?  Then a third review will have to be sought, and that will take more time.  Or, if the journal asks for three reviews up front, and two reviews have come in, one "minor revision" and one "reject"*, the editor really needs that third review.  Should the editor invite another reviewer?  That resets the clock.  A two-week-late review is better than starting all over again with a new 30-day window for another review, but see bullet point #1.
3. It makes you look bad and could even get you eliminated from the reviewer pool.  See Reviewing #3 for why you might not find that desirable.  The journal I edit for keeps track of how long it takes reviewers, and I suspect that's true for most journals.  I avoid reviewers who have bad track records.
4. Some journals have a definite deadline and, if you don't meet that deadline, they will automatically eliminate you from the reviewer pool.  If you have read the paper and are partway through your review, that can be pretty annoying, but it's not the journal's fault.  Some journals, I've heard, will send out many, many invitations and take the first reviews that come in.  That means there is no deadline, or the deadline might be a maximum.  Again, you might be partway through the review and be cut off.  Again, it's not really the journals' fault.  It is so difficult to get reviewers, and so difficult to get reviewers to turn their reviews in, that the journals feel they have to resort to such tactics to keep things moving.

Of course events can intervene.  You might get sick, or an unexpected opportunity comes up.  Figure out if you can still do the review within a reasonable time after the deadline (two weeks at the most).  Then contact the editor.  Chances are, he will give you the extra time, if possible.  Or, he will take it off your plate if the delay is too much.  As is so often the case, communication is key.

*You would be astonished how often this happens.  In my 1.75 years of editing a major journal, only twice have the recommendations been unanimous.