The both/and isn't discussed here, oddly. Many patients get a combination of counseling and medications.
Except for rare situations, antidepressants should only be used short term, followed by clinically monitored withdrawal. Long term kneejerk use has literally created a passive, hypocritical society of zombies unable to feel anything other than their own immediate desires, who erupt with murderous rage the moment they are thwarted in the smallest thing. For this reason, the street name for antidepressants are “soul suckers.”
These clinical aspects also, however, perfectly explain liberal behavior - and virtually all liberals are on antidepressants.
The elimination of placebo responders in the test sample is not wrong; it is an attempt to reduce error in the data set, as is the truncation of Hamilton scores. The mushy middle never proves anything, and only promulgates error in the literature.
Good research is hard, and shouldn't be tried at home by amateurs.