Motivated by the How can an author get assurance that his work will not be stolen by journal staff or reviewers? question:
Authors submit their manuscripts to be reviewed by a journal or conference. Unless the submission is double-blind, the authors' names are listed on the submission, so the referees know who the authors are. Anonymizing referees during a review makes sense, since, for example, you don't want the authors contacting the referees while they are performing the review of the manuscript. But what about after the review has taken place?
Why don't journals and conferences make available to authors the names of the referees who have reviewed the manuscript?
It really makes no sense to me.
Other questions on this site related to the present question:
What to do if a referee plagiarises the result after rejecting a paper?
(And there seem to be many more related questions...)
One common gripe in related questions seems to be the worry that referees can "steal" ideas presented in manuscripts. (I've personally not worried about this at all in my previous submissions.) A possible tactic to thwart "idea-stealing" is to post the manuscript on a pre-print server (e.g., arXiv); however, not all fields/sub-fields think it is kosher to do this.
Thus, given that the "post-manuscript-on-arXiv" solution is not applicable to everyone, this question seeks to determine what are the main, logical reasons why referees' identities are not revealed to authors post-manuscript review.