Ben Wallace-Wells spells it all out in the latest issue of Rolling Stone. This is a remarkable piece, you should read the entire thing. Here’s an excerpt:

There are a handful of battles in the War on Drugs that have actually been won, times when fresh thinking prevailed over politics - but they are not the kind of victories that the Bush administration is eager to trumpet. In the summer of 2003, the police department in High Point, North Carolina, held its annual command-staff retreat in a small conference center themed to look like the log cabins of the pioneers who settled the region. One topic dominated the conversation: an increase in violent crime that was concentrated in three drug-dealing neighborhoods in the city. “The place we were at was that all the traditional enforcement was making no difference,” says the department’s deputy chief, Marty Sumner. “We agreed we weren’t going to be able to eliminate drug use. We weren’t even going to try to go after drug use. We wanted to change the marketing of the drug.”

Sumner’s department called in the Harvard criminologist David Kennedy. The High Point police had worked with Kennedy before, adopting the Boston Gun Project’s policy of trying to break the link between drugs and crime. Now the criminologist told them that he had a new kind of project to propose, one that went beyond the Boston experiment. Kennedy’s pitch was simple: The trick, he said, wasn’t to focus on eliminating drugs but rather to shut down the most “overt” drug markets, the ones operating so openly that they attracted prostitution and violent crime. “Instead of looking at it as a drug problem, we decided to think of it as a drug-market problem,” Sumner says. “What the public really couldn’t stand was the violence associated with public drug markets.” Dealers operating in the open are targets for stickup men and other would-be robbers, and the public swagger and turf consciousness of street slingers can cradle violent, simmering beefs.

High Point police began in the West End neighborhood, one of the city’s three overt drug markets. A team of officers staked out the site, videotaping hundreds of hand-to-hand sales and mapping out a complete anthropology of the West End drug market. They found it was strikingly small: Sumner had expected as many as fifty dealers working there, but it turned out there were only sixteen. Before long, the ­officers had enough evidence to put away each of the sixteen dealers for good. But they didn’t. Instead, Sumner and Kennedy called them in for a meeting. They showed each of them the portfolio of evidence against them and said that unless they stopped dealing drugs, the whole file would be handed over to the prosecutors and they’d be in jail for years. Family members were brought in to urge the dealers to stop, and social-service providers pledged assistance with food, housing and job training.

“We didn’t think it would work,” Sumner tells me, “but the drug markets have disappeared.”

For five years before the program went into effect, the number of drug-related murders in High Point had stayed steady, around fifteen a year. In 2007, in the program’s fourth year, it has plummeted to two. Violent crime in the West End has declined by thirty-five percent. “The use of drugs isn’t something we could affect,” says Kennedy. “But the violence was.” His logic has an appealing clarity for overworked police departments: There are now more than sixty cities in the United States that use some version of Kennedy’s program, edging away from thirty-five years of punitive measures that have turned the United States into the world’s leading jailer to a social-work model that encourages communities and cops to engage the problem on a more human level. The real radicals of the War on Drugs are not the legalization advocates, earnestly preaching from the fringes, but the bureaucrats -the cops and judges and federal agents who are forced into a growing acceptance that rendering a popular commodity illegal, and punishing those who sell it and use it, has simply overwhelmed the capacity of government.