[ylmm]

At least in the social sciences, journal editors delegate papers to reviewers. Your first sentence also suggests that you don't know what peer review is.

Peer review is not the same thing as replication, which is replicating the results of a paper after it's been published (e.g., confirming some groundbreaking finding). Peer review happens at the stage before publication of the original paper. Researcher(s) submit the paper to the journal. The journal editor sends the paper out to some reviewers, who review the paper (this is the "peer review" stage). Pending reviewer feedback and editor approval, the paper is published.

Edit: also the "benefit" that rsa4046 refers to probably doesn't mean networking. AFAIK, reviewers are always anonymous to the authors (which can generate its own problems e.g., if the reviewer gets a paper authored by someone he/she doesn't get along with). The benefit being referred to, I believe, is that of learning to write better reviews, and having reviewed other's work, learning how to improve your own.

[apathy]

Well, yeah, but learning to review is less difficult than learning how to communicate clearly, crisply, and forcefully. So in a well administered review, most benefits accrue to the authors even if the reviews are poorly done, so long as they are ethical. (Editors can step in when reviewers are being unreasonable or ignoring explicit instructions; unfortunately professional editors at glam journals can also step in to push exciting or politically expedient work into press before there is time to adequately vet it, which sucks and gives everyone a bad name).

[ylmm]

Well, yeah, there are other irrelevant things about reviewing that I also have the ability to bring up but my point is that to say that the "benefit" of reviewing is for the purpose of networking is misinformed.

[apathy]

I sometimes agree to review out of a sense of obligation, where I know damned well that the authors recommended me, but you’re right, that’s not networking per se. There’s no reasonable expectation of benefit and it would be unethical as hell to request any.

The authors most likely won’t (can’t) ever know that I agreed to review their paper; it’s more of a good citizen affair. Sometimes, afterwards, it will become apparent that a particular referee was someone familiar. I would like to think that the original poster was imagining something along those lines, but your take is probably closer to the truth. Oh well.