Some work for some people.
I think that part of the issue is that depression has many causes, which are not fully understood, and I would imagine that much of it is caused by several factors. In addition the term depression covers a very wide range of mental health states. This means that measuring the efficacy of anti-depressants is tricky.
Some therapists are against medication, as it can prevent patients from really getting in touch with important emotions, which means that processing them is much harder. On the other hand, most of those therapists recognise that many patients need the support of medication just to be able to get out of bed and into their therapy session.
The recent research which placed anti-depressants on a par with placebo was purely with regard to mild depression (which shouldn't really be medicated anyhow). In moderate to severe depression anti-depressants do perform well against placebos in clinical trials.
This.