Does jail work as an intermediate sanction?

With so many jurisdictions now using some version of so-called problem solving courts, the question arises: What are the best intermediate sanctions to use for probationers as opposed to revoking them to prison? In particular, are short jail stints the answer?

Recently Marc Levin of the Texas Public Policy Foundation urged the House Corrections Committee to follow the lead of Hawaii's HOPE program, which routinely uses jail time to punish minor infractions, reportedly with great success.

By contrast, Doc Berman the other day wrote about a study out of Multinomah County in Oregon that reached the opposite conclusion, published this month in a new Vera Institute report (pdf): "The use of jail as an intermediate sanction was correlated with higher rates of recidivism, a relationship that should be examined more closely." I reacted to this news in the comments thusly:

With a massive shift occurring toward use of so-called problem solving courts, it's really important to identify what are appropriate intermediate sanctions. If it's jail, great - everybody's got one already. But if the Oregon numbers are right and using jails that way boosts recidivism, there needs to be more study and empirical testing about which intermediate sanctions DO work.

A commenter in response suggested that perhaps the issue is not whether jail works as a sanction but how it's used, with the HOPE program showing better results because of the "apparent genuine concern for keeping [probationers] violation free." That writer argued that "Judge Steven Alm's "warning hearings" alone are a cut above the usual assembly line indifference" and could account for the differing results.

Another possibility: It may also be that jail is not best used as the ONLY intermediate sanction but as one of an array of options. In Multinomah County, jail was nearly the lone alternative: "Of the 30 percent of the supervised population who were sanctioned, 92 percent received jail time at some point during their supervision."

Multinomah County sounds like its probation program is not using methodologies from drug courts and other strong probation initiatives like HOPE. "Most people (70 percent) did not receive any type of sanction or intervention while on probation or under post-prison supervision in Multnomah County." So perhaps jail sanctions in the HOPE program work better because they more closely supervise probationers in the first place, actively seeking to identify violations instead of just checking in once per month.

This question requires an evidence based answer. If it hasn't already been done, some researcher needs to define the entire array of possible intermediate sanctions used by problem solving courts and test them individually for effectiveness.

We've got enough specialty courts and strong probation programs in Texas where CJAD (the division in charge of probation out of the Department of Criminal Justice) should be be able to perform such an analysis if it hasn't already been done.

My own guess is that not that all uses of jail to sanction probationers increase recidivism, but that the problem lies instead with overreliance on jail and underutilization of alternatives. That's just speculation, though, in the face of conflicting empirical outcomes. If jail really does boost recidivism and other intermediate sanctions work better, practitioners need to know that.