You probably already know the jaw-dropping stats about American prisons: that the U.S. has the highest incarceration rate in the world. That we're home to 5 percent of the world's population, but 25 percent of its prisoners. That more than one in every 10 American men has a felony conviction. That more than 60 percent of people released from prison are back again within three years.

"Prison is horrible for the prisoner and expensive for the state," write Mark A.R. Kleiman, Angela Hawken and Ross Halperin over at Vox. It doesn't have to be. Kleiman, Hawken and Halperin study criminal justice policy, and they've recently proposed an innovative solution to help cut America's incarcerated population, and ensure that those who are released don't come back. It's called "graduated re-entry." The idea is to let convicts out of prison early -- real early. Three-years-into-a-ten-year-sentence-early.

But here's the catch. Parolees wouldn't simply be dumped on the street with 40 bucks in their pocket and no clear path toward re-entering society, the way they are now. "If someone needed to be locked up yesterday, he shouldn't be completely at liberty today," the researchers write. "And he shouldn't be asked to go from utter dependency to total self-sufficiency in one flying leap."

Rather, the prisoner would be released to an apartment subsidized by the state. At the beginning, he would be subject to strict monitoring. Curfews. Cameras. GPS bracelets. No visitors without permission. "A prison without bars," they call it.

Living this way, under 24-hour surveillance and strict rules, wouldn't be easy or pleasant by any means. It's still a form of punishment. But good behavior would be incentivized with loosened restrictions. Keep your curfew for three months? Get it extended by an hour. Stay away from your old drug haunts? Get permission to travel more broadly. Break any of these rules, even once? The restrictions tighten.

Jurisdictions would have to decide which types of prisoners would be eligible for these programs. You could limit it to non-violent offenders for a pilot program, but in order for it to be truly effective at cutting incarceration rates you'd need to consider other criminals as well. "Solving mass incarceration requires releasing some seriously guilty and dangerous people," according to the researchers. "The problem is how to do that while also protecting public safety by turning ex-criminals into productive, free citizens."

Would this be expensive? Yes. But so is keeping people in jail. It costs about $2,600 per month to house, feed and care for a single prisoner in a typical facility. "The money you spend in the community is balanced by the money you're not spending on a cell," Kleiman and his colleagues write.

The prisoner would help pay his way by mandatory work at a minimally-compensated public service job. He would have to search for a regular job as well. Once he's able to find a regular-paying job, the cost to the state becomes much smaller. And his chances of landing that job -- even if it's just minimum wage -- are much higher than they would be otherwise, because he's already working.

The system might seem to be similar to the old-style halfway houses. Kleiman doesn't see it that way. In an interview, he said "a halfway house is a facility. It's staffed. It puts all the offenders together. Graduated re-entry people use their own apartments. They're not all in the same building. You're not putting all the offenders together. You're not staffing a facility 24 hours a day." All of this keeps costs down, and reduces risks of recidivism.

To get back to historic, pre-1970s levels of incarceration, we'd need to cut our current jail populations by 80 percent or more. And our current situation is not just a function of putting more people behind bars -- part of it is that sentences have grown much larger in the past few decades.

Kleiman attributes this to the drug war, and the spillover effects of harsh drug sentences on other types of crime. "The insanely long sentences really started in the drug era," he said. "Once you were giving somebody 15 years for a coke deal, 7 years for armed robbery wasn’t enough. The drug sentences dragged everything else behind them."

Recent research from the Urban Institute shows that the number of people in federal prison for a drug-related crime roughly doubled between 1994 and 2013. Statutory minimum sentences for many of these offenders are a major contributor to the rise in the federal prison population.

So a return to historic incarceration rates can't happen simply by arresting fewer people -- we also need to provide more ways out for the people currently in there.

A true graduated re-entry program has never been tested. But Kleiman says that other programs that incentivize parolees for good behavior have shown promising results, and that graduated re-entry could apply those same principles. Even better? Many jurisdictions could implement a pilot reentry program without having to write new legislation. Municipalities could literally start piloting this tomorrow if they wished.

Since it hasn't been tested there's no guarantee it would work. But one thing is pretty certain: it hard to imagine anything worse than the current state of affairs.