In this case, illegals supply a required function that the existing distribution of the laborforce does not satisfy.  Without that supply of illegals, the very existence of the industries that use them would be threatened.  There would not be as many restaurants, or as many different kinds.  And costs would be reduced elsewhere, maybe in unsafe ways.  Additionally, like illegal immigration (and black market economies), the situation exists in the open-- precisely because of its necessity.

So if the "loose" Percocet prescribers disappeared, two things would happen: first, Percocet would be pushed into the ilicit underground trade.  It is already partly there. It would become more abused with the removal of even the rudimentary medical supervision.  Second, the practice of medicine itself would change drastically.  It's easy to say that low back pain should only be treated with NSAIDs or physical therapy and weight loss, but it's easy only because the the 100 million scripts a year for Percocet are being written by people who don't say that.  It's also easy to be dogmatic that SSRIs should be the first line agents for the treatment of anxiety as long as you're willing to purposely ignore the 50 million combined yearly scripts for Xanax and Klonopin that non-academics are writing. (For reference: Lipitor is 63 million.)

These commonly maligned inner city doctors actually serve as a pressure valve on the rest of medicine, in the same way that illegal immigration is a pressure valve on labor and industry. 

I'm not judging the actual clinical utility of these drugs, nor am I saying that "loose" practice is good or bad.  I am thinking about what role, if any, such practice has in the larger scheme of medicine.  And I'm convinced that it's actually quite enormous.

What proportion of the inner city population, for example, goes for routine preventative medical care-- checkups,, blood draws, etc-- only because that doctor also gives them Percocet for back pain?  How many people fill the Glucophage only because there's a Percocet on the same script?  I  have no idea-- is it 15%?  If it is, is it ethical, by " not prescribing narcotics,"  to create circumstances in which patients self-select inferior medical care? 

Let's assume Dr. Smith is one of the best doctors in America and he has unparalleled success at treating diabetes in his patients.  However, he never gives Percocet or other narcotics: "that's my policy."  Patients looking for Percocet-- for abuse or legitimate purposes-- will not go to him, and consequently get inferior diabetes care.  Is this ethical?

I'm not saying that the Dr. Percocet won't give good care; I think you understand that my point is how seemingly unrelated medical decisions may in fact be the key medical decisions.

From a simple patient advocacy perspective,  I think it's possibly both disingenuous and unfair to have a "policy" on narcotics.  Whether the risks outweigh the benefits isn't something that can be answered a priori-- each patient's situation is different. So saying "I don't give Xanax" willfully ignores the uniqueness of each patient's existence, as well as the clinical data that says the thing actually does work. 

Treating the patient with maxims and guidelines that result in the patient leaving treatment to go elsewhere is, arguably, the opposite of medical care. Medical treatment often requires unorthodox and unproven maneuvers, but always with an eye towards the risks and benefits, and to the single goal of improving the patient's condition.

I don't have all the answers, but I have a lot more questions.