Clearly, marking a question as a duplicate of either my own question or of a question that I answered is not against the rules. But, there seems to be a conflict of interest in the context of the "reputation-points economy" because I clearly stand to "profit" if a question is marked a duplicate of my question or of a question that I answered because it would drive more people to my post, leading to an increased likelihood of me gaining more reputation-points. I don't think that rules should be restrictive enough to ban such voting, but I was curious if the community thinks that one should err on the side of refraining oneself from engaging in this kind of voting.
Here's what I do. I try to be "objective" as in – would I vote to close if the answer was by someone else? but a little bit more critical with my own stuff. So,
If the new question is very clearly and unambiguously a duplicate, and if I feel that my answer is decent enough, I will vote to close.
If there new question is not an obvious duplicate and there is room for doubt, or if I feel my answer is not above average, then I will only leave a comment: "related/possible duplicate:
link to old question".
Hope this helps.
I think you're over-stressing about this. Closing a question requires a consensus of five different users (potentially down to three at some point in the future) that the question should be closed. It is extremely unlikely that you'll get five (or even, really, three) users with conflicts of interest on the same thread. As such, in addition to AFT's existing answer, I would say: vote to close. If it's not a duplicate, the queue will say no, and if it is, it will provide an objective viewpoint, insofar as that is at all possible.
Things change, of course, with unilateral actions (say, unilateral close-as-duplicate once you have a gold tag badge), and there there is indeed a responsibility to be particularly careful that the action is appropriate (and to seek suitable review, say, in chat, if there's any shadow of a doubt). But that's not a common case.