As I suggested earlier, paternalistic intervention involves a conflict between the risk hierarchies of a benefactor and an intended beneficiary. So paternalistic intervention is is based on the supposition that a benefactor's cost/benefit analysis (at least sometimes) trumps the assessment of a targeted beneficiary AND that the intervention will, in fact, leave the beneficiary "better off." Being made "better off" could mean either an unwanted benefit or an unwanted lessening or removal of a harm. The lessening or removal of a harm could mean either reducing the probability of its occurance or reducing its magnitude. Moreover, justified paternalistic intervention implies empirical evidence that the intended intervention, will in fact, leave the beneficiary "better off" in either the long-run or short run. So the question is whether a beneficiary will (in fact) be better off or not as the result of a given paternalistic intervention.
In assessing whether a beneficiary will be better off or not must take into account the full range of known costs and benefits. Paternalistic intervention invariably requires the use of coercive force, which is often an under-valued cost. The degree of coercive force is, obviously, contingent upon the degree of resistence offered by the intended beneficiary. Sometimes beneficiaries are relatively submissive and make a calculated decision to submit to the benefactor: "It's not worth fighting over!" When the beneficiary offers resistance intervention requires more coercive force. So some interventions require high degrees of physical coercion and inflict greater or lesser injuries on the beneficiary: black eyes, bruises, broken bones etc. Sometimes injuries result, not only from the interdiction process, but also from subduing the resistant beneficiary long enough to provide the benefit. Sometimes the interventions themselves inflict costs, which are euphemistically described as "side-effects. Most anti-psychotic drugs turn out to have serious long-term and short-term side effects, including suicidal thoughts, sexual dysfunction, and loss of energy etc. Thus the overall "redounding good" of any intervention must take into account the amount of coercive force required to get the job done and the side-effects of the intervention itself.
With the benefit of 20/20 hindsight, let's take a look at a few of classic large group paternalistic interventions. In the nineteenth century psychiatrists believed that the mentally ill were "better off" locked up in a mental hospital where they could be bled and force-fed emetics and purgatives. In hindsight, we know that those were unjustified interventions. In the mid-twentieth century, psychiatrists believed that electro-shock therapy and lobotomies left patients better off than bloodletting, emetics, and purgatives. Today we have a long list of modern drugs that have been proven to be either unsafe, ineffective, or both: DES, Fen Fen etc.
So despite the dismal record of well-meaning paternalistic interventions, and despite the fact that most interventions benefit the benefactors (and third parties), why do we accept paternalism as a social practice. I believe we all have a biological predisposition to submit to group-based authority, even when there is little evidence to suggest that authoritative interventions are any more reliable than anyone else. In other words, we blindly submit to group-based authority and rarely question it. This is probably a remnant of our Pleistocene hunter-gatherer brain. Therefore, it is very difficult for us to question socially-designated authorities. The social processes that we employ to manufacture experts is intriguing. Just think about how we ceremoniously bestow these "experts" with credentials: advanced degrees, licenses, etc.