I have been asked to referee a paper for a middle-tier physical science journal. I have reviewed for this journal many times before, although it was several years since I had gotten a paper sent to me by this particular editor. (I thought that this editor disliked me, since I was fairly critical of one of his own research papers.) However, some weeks ago, he sent me a paper to referee, and I sent promptly sent back a report.
The peculiarity came the second time that I was sent a draft of this paper. After the first round, I asked for some significant clarifications, since the work being described in the manuscript seemed to have been done properly, but the justification was very poorly explained. The editor was not able to find another scientist to provide a second report, so the revisions were based solely on my comments. The authors made some changes, which I have not combed through that carefully yet, but they are probably satisfactory.
However, this editor told me explicitly in the letter he sent me that he will definitely accept or reject the paper, according to whatever "verdict" I return. This makes me a little uncomfortable. On the one hand, I am honored that he trusts my judgement sufficiently to put the entire decision effectively in my hands. On the other hand, I feel a nagging concern that the editor is, to some extent, shirking his own responsibilities in this case. Not that I would ever really expect him to go against my recommendation, but I am sort of uncomfortable with the level of responsibility that he has implicitly placed on me.
As a general rule, the way this editor works is somewhat idiosyncratic. For example, the journal has an online system for submission and refereeing, but its use is apparently optional for the editors. Every other editor I have worked with on this journal has used the Web-based system for assigning and collecting reviews, but this guy does everything entirely via e-mail. His e-mails can be rambling and mix personal and professional communication, quite different from most e-mails I get from journals, which are almost all boilerplate.
So I guess my question has two aspects: First, is it reasonable to be concerned about what the editor has said? Second, is there anything I can (or should) do about it?