One annoyance in drug research is the placebo response, that improvement in a client's well-being that occurs in the absence of treatment or that occurs in response to the appearances of treatment. The placebo response often is about 45%, especially in research on antidepressants wherein the typical drug response is about 65%. It's possible that our present drug structures are ineffective; it's also possible that something else is occurring.
De Wall wrote at length about succorance, the behaviors that care for an injured creature. Chimpanzees exhibit this trait as do we. The possibility exists that succorance may have some standing as a psychological adaptation.
Succorance likely, if an adaptation, has trigger stimuli and efficient, reciprocal behvior sequences between victim and consoler. Complaints from the victim are greater when associated with hopelessness and isolation. A hug, a smile, warmth, and food are evidence of an alliance and elicit a reduction in complaints. The consoler is reinforced.
The sugar pill is a symbol, perhaps even a supernormal releaser by Tinbergen's standards (given the exalted standing given by many of us to medication), and would be effective because it triggers the same mechanisms as traditional expression of alliance from a relative, an alpha, or an alpha-shaman ("My doctor is the best and she said this will fix me right up.") Because of the power of psychological adaptations, the placebo and the antidepressant may often trigger the exact same mechanisms and achieve comparable outcomes.