> The best guesses by outside observers is that for a lot of them about 80% to 90% of their graduates relapse within a couple of years. The problem with this s

> The best guesses by outside observers is that for a lot of them about 80% to 90% of their graduates relapse within a couple of years. The problem with this statement is that you can go back to using drugs, but in a significantly lower risk pattern than previously. E.g. if you were shooting up heroin, but switch to smoking weed. This is actually pretty common, so these top line relapse rates dramatically understate the efficacy of these programs. That's not to say they're super effective, but I think there's at least something to be said for intellectual honesty.