What Prisoners Are There For

The Koch Brothers have teamed up with some liberal groups, including the Center for American Progress and the ACLU, to start a justice-reform initiative called the Coalition for Public Safety. Cue the stories about "strange bedfellows" and a "bipartisan consensus" that we're locking too many people up for too long.

There is a significant opportunity here: Most Americans are now okay with legalizing marijuana, for example, and surveys indicate overwhelming support for reducing sentences for "low-risk, nonviolent" offenders. But as I've noted before, the public is definitely not okay with reducing sentences across the board: More than 60 percent of Americans still think their local courts are not harsh enough with criminals, a 25- to 30-point drop since the '70s, '80s, and early '90s, but still a majority. Even the broad notion that we have "too many people in prison" nets just 45 percent support, with 41 percent believing we have the right amount or not enough. (The rest say they don't know.)

So, before we get too excited, we might want to get a grip on how many offenders are plausible candidates for reduced sentences. Here are some numbers we should bear in mind as we think through a strategy for reform.

First of all, here are the broad offense categories for the prison population -- state and federal, sentences of one year or more -- as of 2012. "Public order" offenses include immigration violations, as well as "weapons, drunk driving, and court offenses; commercialized vice, morals, and decency offenses; and liquor law violations."

For state prisoners released in 2012, the median time served was 28 months for violent offenses (ranging from 17 months for assault to 153 for murder and nonnegligent manslaughter), 12 months for property offenses, 13 months for drug offenses, and 12 months for public-order offenses. About three-quarters of released state prisoners, meanwhile, are arrested again within five years.

Given marijuana-law trends, drugs seem to have the most potential. What kinds of drugs are involved? Sometimes pot, but usually cocaine, which Americans, rightly or wrongly, absolutely do not support legalizing. Here's a chart from a 2004 Justice Department report:

What exactly were they doing with those drugs? What was their criminal history? How long were their sentences? Here's another helpful chart from the same report:

There's plenty of room for change here -- legal pot, lighter sentences for possession of other drugs, etc., could reduce the prison population by thousands. But most prisoners aren't there for drugs, and most drug prisoners have prior offenses and were charged with more than simple possession. It will be a challenge to sort out which offenders we can let out without risking a public backlash, and it's good to see some bipartisan cooperation on this project.

Robert VerBruggen is editor of RealClearPolicy. Twitter: @RAVerBruggen

[Note: The numbers in the paragraph following the first chart have been corrected.]