Slate links to an interesting 2003 study by University of Michigan researchers indicating that random testing of high school students for drug use has essentially no effect on the rate of student drug use.
I'll confess some surprise at these results. What I'd have expected is a mild net reduction in drug use, a fair amount of countermeasure deployment (Goldenseal, that sort of thing) and, most perversely, a gradual shift away from drugs like marijuana, which remains in the body for weeks, and toward drugs like cocaine, which can only be detected for a few days.
In any event, once we add to the objection that such testing programs are invasive and degrading the fact that they don't even appear to work, precisely what rationalization is left? Presumably whatever jollies administrators get from the prospect of taut adolescent bodies urinating on command.
Update: Jane Galt is skeptical about the idea that in the face of random drug tests, students might shift their drug consumption from marijuana to cocaine:
I can't see my (high-school) self saying "You know, I'd really like to get all my friends into my basement and smoke some dope and get all giggly and watch The Wall . . . but with these random drug tests, I guess we'd better spend eight times as much on cocaine and go out to a club where we can talk real fast and listen to our hearts race."
In the short term, of course, she's mostly got a point: Hippie kids who like dope smoking are unlikely to abruptly switch to uppers. But when I said a "gradual shift," I was thinking on the scale of a few years, not a few months. Consider: When I was in high school, plenty of kids smoked pot, but there was, relatively speaking, not very much cocaine use—which there had been in the late '80s. When I came back to visit my old journalism advisor a few years after graduating, I learned that there was once again a burgeoning cocaine scene. I'm not sure exactly what cultural or economic or other factors prompted the shift, but what's worth noting is that the least plausible hypothesis is that the new cohort of students all just had a greater intrinsic preference for coke than pot, while previous cohorts had the opposite preference. Especially in high school, drugs are schelling points: While obviously people will have preferences for the effects of one drug or another, what's more important in many cases is that it's something transgressive that cliques can all do together. For any particular group of friends with an equilibrium drug already chosen, or for a particular person with settled tastes, consumption choices may be relatively inelastic—but over time, I wouldn't be surprised if something like random testing were able to shift the most common equilibrium drug.