To paraphrase Bill Clinton’s famous campaign mnemonic of 1992 – the over-incarceration problem in America isn’t the war on drugs, or draconian sentencing – its overcharging by prosecutors. Criminologist John Pfaff lays out the argument in this Slate interview:

So why did the prison population keep on rising after 1991, when the crime wave ended? It seems like if your theory is right, that the increase in violent crime and property crime caused the prison boom, the end of the crime wave should have been accompanied by decreasing incarceration rates.

Three things could have happened. One, police just got much more efficient—they’re just arresting more and more people, with new policing technologies, new policing approaches—maybe they’re just arresting a bigger share of offenders. But we don’t actually see that. Arrests tend to drop with the crime rate. So the total number of people being arrested has fallen. The other thing it could be is we’re just locking people up for longer—but like I said, it’s not that. So clearly what’s happening is we’re just admitting more people to prison. Though we have a smaller pool of people being arrested, we’re sending a larger and larger number of them to prison.

Why would that be?

What appears to happen during this time—the years I look at are 1994 to 2008, just based on the data that’s available—is that the probability that a district attorney files a felony charge against an arrestee goes from about 1 in 3, to 2 in 3. So over the course of the ’90s and 2000s, district attorneys just got much more aggressive in how they filed charges. Defendants who they would not have filed felony charges against before, they now are charging with felonies. I can’t tell you why they’re doing that. No one’s really got an answer to that yet. But it does seem that the number of felony cases filed shoots up very strongly, even as the number of arrests goes down.

Isn’t the traditional explanation for why prosecutors tend to be overzealous is that their political careers depend on it?

The political question is interesting because generally the district attorney election is not very difficult to win. DAs tend to win elections pretty regularly. So, when Joe Hynes was defeated in the Democratic primary in Brooklyn, New York, in 2012, he was the first sitting Brooklyn DA to run for re-election and lose in more than a century. But that’s not to say that politics don’t matter: Maybe it’s that next election they’re looking at, that they remain tough on crime because they want to become attorney general or governor. There’s no clear data on this. We’re only just starting to look at this question. But that strikes me as a possible story. What might have happened is the crime boom made being a prosecutor more of a launch-pad position—it elevated the status of prosecutors, and perhaps elevated their political ambitions, and they remained tough on crime even as crime started going down.

OK. So why does any of this matter? Why is it important for reformers to have the right theory for why mass incarceration happened?

The reason it’s important to get it right is that if we’re trying to reduce the prison population, we want to make sure we do it correctly—and if you focus on the wrong thing, you won’t solve the problem. So if you think it’s the war on drugs, you might think, ‘OK, if we just decriminalize drugs, that will solve the problem.’ And, you know, it’s true that if we shift away from punishment to treatment that could be a huge improvement. But just letting people out of prison—decarcerating drug offenders—will not reduce the prison population by as much as people think. If you released every person in prison on a drug charge today, our state prison population would drop from about 1.5 million to 1.2 million. So we’d still be the world’s largest incarcerating country; we’d still have an enormous prison population.

And if we focused on cutting back sentence lengths, maybe that would weaken DAs’ bargaining power at plea bargaining, but since people aren’t serving the massively long sentences anyway, it probably won’t have that big an effect on prison population either.

What would?

Well, the real growth in the prison population comes from county-level district attorneys sending violent people to prison. And there’s a lot to be said for nonprison approaches to a lot of people who are in prison for violent crimes. But that’s a political issue that we haven’t even begun to address, in part because it’s politically scary.

Where does that leave reformers who want to see the prison population drop significantly?

What makes it very hard is that the person we really need to target now—whose behavior we need to regulate—is the district attorney, and the district attorney is a very politically independent figure. He’s directly elected, and he’s directly elected at the county level. So there’s no big centralized fix. You can’t necessarily go to Washington and say, ‘Here’s the law that’s going to control what the DAs do,’ because they don’t have to listen to the federal government at all. So you have to figure out how to go county by county and either elect DAs who have less punitive attitudes, or you can try to sort of change the incentives DAs face at the state level. But it’s very tricky.