Dylan Matthews: One point your piece makes very strongly is that the crime problem in the United States isn't mostly about illegal drugs.
Mark Kleiman: Drugs are an important part of the question if you include alcohol as a drug. Take any dimension of the problem you like, except for source country violence. All illegal drugs combined are to alcohol as the Mediterranean is to the Pacific. We have our whole navy in the Mediterranean. And that's true both of the drug policy machinery and those who are fighting the drug war, and of the drug reform movement, which, it seems to me, neglects the problem with the one drug we've legalized. Any sentence about drug policy that doesn't end with "raise alcohol taxes" is an incoherent sentence.
There's no question that the upsurge in heroin usage in the 1960s was a big contributor to that phase of the crime wave. I don't think there's any evidence that increasing marijuana use had anything to do that. Then for the 1980s and early 1990s, Blumstein is absolutely convincing about the correlation between the spread of the crack trade and the spread of homicide.
Yes, that would have looked very different if cocaine had been a legal drug. I think it's a perfectly fair point. On the other hand, we would have had a wave of violence around the combination of cocaine and alcohol.
Matthews: Is that because of the combination of lower inhibitions and increased energy?
Kleiman: Yes, and it turns out there's a molecule called cocaethylene that's the product of having both alcohol and cocaine in your bloodstream, and if you administer it to rats, they get more aggressive.
Matthews: Roughly how much of the crime problem would you attribute to alcohol, percentage-wise?
Kleiman: Half the people in prison were drinking when they did whatever they did...Of the class of people who go to prison, a lot of them are drunk a lot of the time. So that doesn't mean that they wouldn't have done it if they had not been drunk. It's just that being drunk and committing burglary are both parts of their lifestyle. Still, alcohol shortens time horizons, and people with shorter time horizons are more criminally active because they're less scared of the punishment. Most people who drive drunk are sensible enough to know when they're sober that they shouldn't be driving drunk. It's only when they're drunk that they forget they're not supposed to drive drunk.
We need to keep them from drinking, which is what the 24/7 program does. We could also require everyone to be carded. Maybe you still get carded, but I don't. But imagine everyone got carded, and if I had a DUI, I had a driving license showing I wasn't allowed to buy a drink. You'd make the alcohol industry regulate its own customers. And I think you'd cut down on crimes substantially. But if I say that, I'm a nanny state fanatic, and if I say adults should be allowed to smoke a little bit of pot, I'm a crazy drug reformer.
Matthews: Excuse my ignorance, but what's the 24/7 program?
Kleiman: This is something that a guy named Larry Long, who's now a judge but was a district attorney and then attorney general in South Dakota, invented. A second time DUI in South Dakota is prison time. A third of their prisoners are in for DUI. They have miles and miles with nothing to do and no other means of transportation, so they're the perfect storm for drunk driving.
Long got tired of sending people to prison, so he invented a program where they could agree to come in at 7 a.m. and 9 p.m. every day for a breath test with the promise that they'd be in jail if they failed. Forty-seven percent of the people screw up at least once, but more than 99 percent of the tests are taken and passed, and the result is reduced recidivism for DUI and assaultive crime. Beau Kilmer at RAND has a couple of papers showing the impact. It's a proven program for alcohol. We're learning that testing and sanctions is the right way to deal with drug use linked to crime, whatever the drug is and whatever the crime is.
If someone's drug use is a problem for other people, you should make him stop using drugs. Then you can stop putting him in jail. I think we could seriously think about decriminalizing personal possession for use for all drugs. We'll make it illegal to sell cocaine because we think too many people would be tempted to use it if it were legal to sell. But if you're just using it, that's your problem until you make it our problem. If you make it our problem, we'll stop you using it.
Matthews: Isn't there a risk that legalizing possession would increase demand?
Kleiman: There's very little evidence that possession arrests discourage drug use. You'd think they would, except the risks are so small. The probability of being arrested for a particular instance of cocaine sales is about 1 in 15,000. For use, it must be some small fraction of that. People who are deterred by such tiny risks are not likely to be drug users in the first place. There's just no evidence that banning possession reduces use.
The possession laws, they're giving cops the opportunity to arrest people they want to arrest anyway. That's what Stop and Frisk in New York is about. I'd trade that for rigorous drug testing for people on probation and parole and pretrial release.
Matthews: Circling back to alcohol, it's not just DUI you'd cut down on if you restricted alcohol use. You'd cut down on battery and burglary, right?
Kleiman: If you cut down on heavy drinking you're going to cut down on battery. That's what you see in domestic violence cases. How much, it's hard to tell. You could raise the tax. The tax affects heavy drinkers, and it's heavy drinkers we're worried about. Have I ever told you my Johnny Walker Black story?
Matthews: No, but please do.
Kleiman: When I was a kid, my parents would get The New Yorker, and the back cover was always a Johnny Walker Black ad. And they were great ads. Just terrific ads. One of them shows lights in an office building spiraling up. Next to it there's a picture of Johnny Walker Black. And it says, "As you move up, the work doesn't get easier, but the rewards get better."
But another said, "If the difference in price between Black and ordinary scotch matters to you, you're drinking too much." And I regard that as the first principle of drug policy. Price matters a lot to people who use a lot, and so it's a very good way to regulate consumption. So here I am in Washington state, thinking about regulating cannabis, and a big question is how to keep the prices up.
Matthews: How big of a game-changer is GPS-enforced house arrest or probation monitoring?
Kleiman: It could be huge. We're going to start that in NYC. They're going to use GPS curfews or home confinement as a sanction for probation violations. It's enormous if it works. It started off with active monitoring for sex offenders, and it wasn't very effective, and quite expensive. There are a couple of technical questions, like, 'What you do if he goes in the subway?' You've just lost him. So do you say you can't go in the subway? Then again subways tend to have cell service, so if you have GPS and cell phone combined, you're better off. You need to make sure the system doesn't issue too many false alarms, where it looks as if the person broke the rules when he didn't.
My view is that if you know where someone is, you don't have to put them in the cage.
Matthews: A political problem is that the main revenue source for alcohol companies isn't you or me buying the occasional bottle of Johnny Walker Black…
Kleiman: Not that I'd be caught dead drinking a blended Scotch.
Matthews: No, of course not, single-malt all the way. But how much power do the spirits companies have? It seems like they'd fight any price increase.
Kleiman: Much power. The spirits guys are not really important because they're not the real market. The real problem is beer. The beer guys are powerful. It's two thirds of the market. Not only do they have heavy campaign contributions to politicians, because they're state regulated and thus have a stake in state politics, but customers don't dislike their beer company, so if they get a political message from the beer company, they'll respond.
Contrast that with tobacco, with a smaller number of lower status users who hate their providers. The cigarette companies have absolutely no luck mobilizing smokers. Smokers hate tobacco companies. It's easy to say it's just a tax on responsible drinking until you do the math. It would cost a typical beer drinker $36 a year. The man who'd get hit is the 10 beer a day drinker, and he's the guy we want to hit.
Taxation is just about the perfect way to control alcohol use. It's not complete, because you need controls for the real problem drinkers. But if we tripled the alcohol tax it would reduce homicide by 6 percent. And you're not putting anybody in jail. But instead we spend our time talking about doing marijuana testing for welfare recipients.
Matthews: I know you think the price should be much higher than it is now, but what's the ideal price, do you think, for alcohol once you've taken all the social costs into account?
Kleiman: That's a really really interesting and hard question. The idea of an externality, or specifically of a Pigovian tax, is based on the notion that the harms are homogenous. You figure out what the cost of each and every ton of carbon dioxide is and put a tax on it.
The external costs of drinking are heterogeneous because drinkers are heterogenous. The cost of the second wine you have every day is negative. It's probably good for the world. The cost for the next drink for a mean drunk is going to be dollars, while the average price of the actual drink is about a buck. So there's not a single tax that's the right number both for the casual drinker and for the mean drunk.
So, you can't do this solely with taxation. You need some regulation with the tax. But tripling the tax would add something like $17 billion a year in new revenue. When we were debating how to pay for Obamacare we talked about taxing soft drinks, about taxing tanning beds, taxing plastic surgery, and we actually tax medical devices. Taxing alcohol was never on the list.