Perfect Deterrence?

One of the initial problems I had with Bogard (and I am now realizing it might be with Foucault, whose discussion of panopticism he relies upon) was how the logic of panopticism “induces offenders to police their own behavior, transforming them into ‘subjects…’” And it is this turn on “subject” that Foucault, and arguably Bogard too, is making. It seems that panoptical surveillance creates at once a subject and object of the surveilled – object in so far as it is the “prisoner” on whom the disciplinary gaze falls, and subject in so far as that same prisoner ends up playing a proactive role in his/her own discipline, ensuring “docility.”
I’m not sure if I can wholly digest Bogard’s claim that in perfect form, simulation leads to perfect deterrence, which is “a state where deterrence is no longer necessary” (32). On the one hand, this might coincide nicely with the above, with the seeming hybridity that exists in the object/subject. If I am the object of my own gaze, my own disciplinary subjectivity, have I surpassed the point Bogard is referring to, where deterrence is no longer necessary? But isn’t that a type of deterrence in itself. Having an individual or a population self-monitor, I think, is not some form of perfect deterrence. It might be a sophisticated one, and it might reduce the amount of information processing or even pre-processing that authorities have to do, but it is still deterrence, more hidden but not, I would say, perfect.