I don't doubt it. They do sometimes work, which is why I don't like the way this paper is being passed around; people are drawing incorrect conclusions from it.

That being said, here's my anecdote: I was worried about the side effects and hesistant to try them, but everyone dismissed my concerns, insisting that SSRIs are the "gold standard", that they're the most effective, that the side effects usually aren't too severe, blah blah blah. But not only did everything I worried about happening happen, they also did absolutely nothing to help me. I would rather have wasted my time with bupropion or some other drug that didn't fuck my shit up at all. Drugs with less severe and less frequent side effects should be tried first (that's usually the standard in medicine!). I believe the pharmaceutical industry has had a hand in this, and I think it constitutes systematic malpractice.