You’ve conceded that client variables are more important than treatment modalities but argued that this is exactly analogous to medicine, suggesting this is the case because, for instance, the sicker the patient the less likely there will be a good outcome. However, two important points seem to me to have been lost here… The first is that outcome studies (whether for effectiveness or efficacy) control for differences in parameters like “the sickness of the patient” — in other words, the conclusions of these studies refer to the effectiveness of different treatments given that it didn’t just so happen that one patient was in better health than the other. The second is that the “fifteen percent that makes all the difference” may differ wildly between the professional realms you cite, such as those of psychotherapists, surgeons, plumbers and electricians. If I had to take a wild guess, I would imagine that “plumbing outcome” is pretty close to 95% down to “plumber factors” such as plumber training, plumber choice of materials, etc. Likewise for “electrical wiring outcome”. Medicine broadly understood differs from either of these, because here client variables do emerge pretty noticeably (Hubble, et al. have some interesting material on the placebo effect which bears on this), but outcomes are still largely determined by the practitioner, once we have accounted for things like who was healthier to begin with. Psychotherapy, however, differs far more, because here the impact of specific practitioner variables dwindles to somewhere around the 15% figure you cite.

The argument for registration on the basis of effectiveness, however, really seems to me to come undone on the fact that the variables which typically figure in registration schemes for psychotherapy are almost entirely irrelevant to (and in some cases are inversely correlated with) clinical outcomes. This patently is not the case with electricians, plumbers, or physicians, where the types of skills and knowledge being measured are strongly correlated with outcomes.

Having said all that, I agree with you about the role of regulating agencies in addressing ethical problems. I agree that it’s very important to address the ethics violations you cite, such as beating clients or having sex with clients. Fortunately, I can fully support that specific aspect of regulation and agree with you on its importance — while still rejecting other aspects of arguments commonly put forward in this area that so often spuriously link practitioner variables with outcomes, in outright contradiction with empirical research evidence.