In a recent links roundup post, Tyler Cowen repeats his claim that decriminalization is superior to legalization of marijuana. He links to a story by Malcolm Gladwell and another one by Alex Berenson and, adding his own commentary, says, "Decriminalization, not legalization, is the way to go." Nothing in either piece supports his policy recommendation. ("Decriminalization" generally means no criminal sanctions or light fines for users, but continuation of prohibition on the supply side.) I've written before that I see this as a major blind spot for Cowen.
It's true that both pieces suggest that the harms from cannabis use may be understated. The legalizers may have been too zealous in denying any possible harmful effect of the drug, not wanting to give drug warriors any hooks to hang their policy on. But what's missing here is any kind of cost-benefit analysis. Is legalization a bad policy if any social problem (crime/mental illness/accidents) increases at all? Is there any accounting at all for the enjoyment that people get from consuming cannabis? Like I've said before, pleasure counts. As I've also pointed out before, Cowen and Tabarrok's own economics textbook provides an ironclad economic case against drug prohibition. (See the link for the full details of the argument.) Cowen owes his readers some explanation of why that analysis, which was important enough to include in his textbook, doesn't apply to marijuana. He might have some deep cultural or "histeresis" argument, about how the standard economic treatment is missing something important. I'd like to hear it so I'd know if it's a thoughtful argument that I might actually agree with, or if it's a collection of un-curated, pasted-together factoids assembled by a news-junkie. (Sorry, Tyler.)
Another thing that's missing here is a thorough vetting of the claims. The Berenson piece claims that violent crime has risen in the states that have legalized marijuana, since 2014. It's also risen nationwide over that period, so it's hard to claim anything about the effects of legalization without a more thorough analysis. (Yikes! I haven't noticed this before. Hope it's just a blip.) Both pieces suggest that some people are susceptible to developing psychosis from marijuana use. I think this claim needs to be parsed. I have heard elsewhere that marijuana use does interfere with the treatment regimes of some mental disorders (schizophrenia in particular), causing them to deteriorate or making treatment impossible. But I think there's a huge difference between saying that a cannabis harms a small percent of the population with pre-existing mental health problems and suggesting that it's dangerous for normal people using at moderate rates. Good on Berenson and Gladwell if they are simply pointing out that some people are more susceptible than others to the hazards of drug use. I'd like to see more of that kind of thinking. Other obvious examples of this: some people are allergic to marijuana smoke, and some AIDS patients are susceptible to fungus that might be growing on poorly cultivated marijuana. (And it sure looks like people with chronic health conditions are more susceptible to opioid overdose than normal, healthy people.) These are useful things to know. It's important to identify risk factors, but this should lead to a nuanced, targeted drug policy. Not general prohibition or "decriminalization", which is still prohibition on the supply side.
[As I was writing this post, I saw Jacob Sullum's excellent piece in Reason on the two Berenson/Malcolm pieces. Apparently some other people have considered the recent rise in violence more thoroughly, and they've found that it can't be plausibly blamed on marijuana legalization. See also this Twitter feed that Sullum links to. Or don't. Ugh. Way too much snark. Legalization advocates, stop doing this. It poisons your moral credibility. ]
Berenson also points out that the black market still exists. This is because excessive taxes have made it more expensive than it needs to be, and onerous regulations make it difficult to obtain in some places. So of course a black market is still going to exist, just as black market cigarettes and black market moonshine still exist, even though these goods are legal. We are still way better off with legal markets in these goods. The black market counterparts to these legal goods represent diversions from clean, regulated, audited industrial processes rather than filthy illegal operations. So it's a cleaner illegal market that we'd have under total prohibition. (More so for cigarettes than for moonshine, but still true of both.)
Both pieces gave me plenty to think about, but they ultimately miss the mark. The notion that marijuana use increases rates of psychosis at the population level is alarmist in the extreme. There is a way of saying this that implies that the entire population is at risk, as if everyone who smoked were making a role of the same set of dice. There is a more accurate way of saying this that reveals to the reader/listener that most normal people don't have anything to worry about, but some people with pre-existing issues and impulse-control problems might experience harms from cannabis use. I don't think it's useful to talk about drugs use as if it exposed the user to randomly distributed harm, as if randomly sampling from a probability distribution. Usually the harm requires deliberate, excessive use, which means the user has a great deal of control over whether any harm occurs. (Of course, this is less true of, say, heroin, where the user can't know the concentration. Another hazard of prohibition.)