Because it isn't quite that simple. Overprescription of antibiotics isn't usually done only for the placebo effects, but other social reasons as well. A significant subset of patients react poorly to "it's probably viral and so we probably can't do anything". As bad as antibiotic overuse is, lying to patients and fraudulently prescribing a treatment is still much worse than giving someone antibiotics in the 5% chance that something is bacterial.
I don't think that lying to patients to falsely reinforce their untrue beliefs is a solution to the problem -- it'll probably create an additional problem by reinforcing beliefs of the legitimacy of pseudoscience.
Although not as effective as deceptive placebos. I personally suspect that the subset of people for which honest placebos work well overlaps significantly with the same subset of people who already believe "pill = good" above the expert advise of their doctors.
On a 4 page list of my notes on issues and treatments I mentioned that my symptoms always got better on doxycycline for a couple of months. My ancient neuro-ophtomogist oh, you have Sjogrens. I’m like what’s that? Apparently he had seen that many times before.
I don't think that lying to patients to falsely reinforce their untrue beliefs is a solution to the problem -- it'll probably create an additional problem by reinforcing beliefs of the legitimacy of pseudoscience.