I have been asked to review a paper that I know to have been authored by a friend and former colleague. Is it my responsibility to inform the editor and decline to review the paper?
Assuming you reviewing the paper would not violate any guidelines of the journal or conference in question, the primary issue is objectivity: can you review the paper on its own merits, or will your association with the author in question "color" your opinions?
Personally, I would never review a paper written by a former co-author or personal friend. However, I may accept the review of someone whom I know professionally but have not worked with in any formal capacity, provided that I am not working on a similar project.
In this particular example, you claim that the person is "a friend." If that is the case, then I would recommend leaning against accepting the review.
The issue of informing the journal editor is a separate matter. You may explain that the reason is you know the author, but you are not necessarily obligated to do so. If you want, you could simply say "I don't have the time right now," and that would be just as valid a reason as "it's a conflict of interest." But I don't think there's anything lost by informing the editor of why you don't feel comfortable providing the review.
All other things the same, it would be better to inform the editor and decline, yes.
Similarly, if the author is not a "friend", but some sort of opposite to that, ideally one would decline for the opposite reason.
It starts to become trickier when the author is "a competitor".
In any case, with time, it can easily happen that many people in a given field are well known to you, "good acquaintances" (or "bad"...) if not "friends". That is, the ideal of dispassionate opinion due to lack of personal connection becomes impossible, not matter the ideal.
Thus, in practice, although it is harmless to inform the editor, they probably won't be surprised, and most likely would not at all insist that you decline... although you are within your rights to do so in any case.
Some journals have rules about reviewing the work of former colleagues. For example, I've seen journals claim that papers will not be reviewed by people that have been at the same institution over the last five years. You have to check with the editorial policies of the journal to see about this but if the editor has made a mistake by sending the paper to you, they'll appreciate you pointing it out.
The bigger issues are that (a) you know who the author is so the review will not be blind and (b) the author is your friend which might reasonably be seen as a conflict of interest. If you know the author because you have seen the work in the paper before, that is a third problem. I would not review papers that I know are written by my friends for all of these reasons and I think it is in the interest of the journal and of the peer review process to defer review to someone else.
Is it my responsibility to inform the editor and decline to review the paper?
It is your responsibility to inform the editor of the situation, but the two of you can decide together whether or not your should decline the review.
The editor may ask you whether you feel that you can objectively review the work (the real answer is that none of us are truly objective, but the question here is whether you would feel comfortable rejecting the paper if you think it is bad and accepting the paper if you think it is good).
It depends. Is this a double-blind refereeing system? If so, how do you know the paper was written by a close colleague? (Technically, everyone working in your discipline is a colleague, but some colleagues are closer than others.)
In some disciplines, blind reviewing isn't used or single blind reviewing is used. In that case, you know because the author's name is on the manuscript.
In either case, I would notify the associate editor who asked me to review the paper that I have a (potential) conflict and explain the situation exactly. If I felt I could maintain a disinterested perspective on the paper I would say so and leave the decision up to the associate editor. If I felt I could not maintain a disinterested perspective, I would say so and decline to review the paper.
With all that said, my perspective on my duties as a referee is a little different from the norm. I review work for technical correctness and point out any errors and limitations I might find. I comment on whether I find the work interesting. I will not make publication recommendations. I see that as the editorial board's job, not mine. That perspective reduces the potential for conflict of interest in my mind. I basically won't tell an editor anything I would not say directly to an author.
It depends on the journal. Some Elsevier journals do not consider this as a conflict of interest:
Avoid a potential conflict of interest A conflict of interest will not necessarily eliminate you from reviewing an article, but full disclosure to the editor will allow them to make an informed decision. For example; if you work in the same department or institute as one of the authors; if you have worked on a paper previously with an author; or you have a professional or financial connection to the article. These should all be listed when responding to the editor’s invitation for review.
So, unless you feel that you cannot provide an objective (to the extent possible) review, you can take the job. However, when you write the letter with your decision, you have to specify your connection with the authors or the paper. I am taking the word "friend" in a more relaxed sense. If, for instance, one of the authors donated you or your parents one of his kidneys, then it may be difficult to provide a useful review.
I guess if this were a double blind review you would not run into this problem! I have been in a small community of scholars for more than 30 years. It is impossible for me to avoid this conflict. I have to decide for myself, or let the editor decide. If the conflict might be perceived as significant, I usually leave it to the editor, telling them of the conflict and asking if they want me to continue. The entire review process is inherently subjective, and there are many reasons one can have bias when reviewing a manuscript, for example, what if their results contradict yours, but are well supported and it's a good paper? You have to use your best judgement in all cases.
Inform the editor. If he is OK with it (and you are) than you may still review the paper. After all he knows you are conflicted and will weigh your response.
It is not the only form of conflict. Rival groups, subject of critical papers, etc. Good editors know how to involve such parties, including when to cut off unfair reviews, get another opinion, etc.