Illustration by Zack Soto

EVERYONE CLAPPED two months ago, as Kayla Ballew strolled up to a courthouse lectern in downtown Portland to accept a rarefied piece of paper.

The certificate announced the 22-year-old had completed START Court, a treatment program run out of Multnomah County Circuit Court aimed at addicts with a penchant for theft.

Ballew, a heroin addict who was caught pawning a stolen ring, had done what more than 80 percent of others in her shoes had not. She'd navigated a strict regimen of counseling, drug screens, and constant court dates while avoiding further trouble. She was a year clean, she told county staff on hand. A stint at a Subway had lasted longer than any other job she'd ever had.

Ballew earned her applause. But for skeptical prosecutors around the state, an uncomfortable question hangs over those good feelings: Is START Court, and scads of costly programs like it, actually worth the money?

As Oregon prepares to throw tens of millions of dollars at slowing prison growth in the next two years, that question is taking center stage for public safety officials, who have turned it into another front in an ongoing debate over prison reform.

District attorneys have long argued programs like START too often skate by on a combination of good intentions and protective bureaucracy. For years, some of Oregon's most-vocal DAs have urged strict and rigorous experiments to prove, once and for all, whether such efforts successfully cut down on crime.

"They're basic principles of science that are applied everywhere," says Lane County District Attorney Alex Gardner, "and for some reason in corrections and criminal justice, folks have been willing to accept very poor science."

Now, Gardner and like-minded colleagues could be on the verge of a big victory. Oregon is in the process of deciding how to distribute millions in state cash aimed at shrinking prison use ["Taking the Long View," News, Oct 8]. No one knows how the rules for spending that money will ultimately look, but the current draft requires the very experiments DAs have pushed—known as randomized controlled trials, or RCTs—to prove the effectiveness of any program funded by the money.

The only problem? Virtually everyone besides the DAs is railing against that requirement.

In Multnomah County alone, judges, probation officers, defense attorneys, and other officials say these rigorous examinations are often too complicated and expensive. They admit RCTs are the "gold standard" of research, but argue that insisting on them may prevent innovative new programs. And that, some officials tell the Mercury, might be precisely the point.

Prosecutors' concern "is not as pure as what is being presented," says Scott Taylor, who oversees probation and post-prison supervision services at Multnomah County, and therefore many of the programs DAs might call into question. By requiring RCTs, Taylor says, prosecutors might be hoping to delay efforts that would let more people out of prison, earlier.

"There are less rigorous tests that give you assurances that 'we're pretty sure this is working,'" says Lane Borg, executive director of Metropolitan Public Defender Services, Oregon's largest public defense contractor. "I believe it was just a roadblock that was put in to create impossible hurdles for reforms."

Other observers, and the DAs themselves, dismiss that charge.

"I don't buy that," says Craig Prins, executive director of the Oregon Criminal Justice Commission. "This is something they latched onto eight years ago. They truly believe that we should do as many RCTs as possible."

In Multnomah County's main courthouse—and Oregon as a whole—START Court is a rarity. The county counts two drug courts, a veterans' court, a drunken driving court, a domestic violence court, and a mental health court among its many offerings. And it pours millions of dollars into the programs every year (since multiple agencies are involved in each effort, exact costs are hard to nail down). But only START has been studied under the harsh glare of the randomized controlled trial—a years-long process that hasn't yet generated concrete conclusions but nonetheless suggests the program is effective.

The beauty of RCTs is also one of its detractors' chief complaints. Under the method, qualified participants are selected at random to participate in a targeted activity—say, a special court—while other qualified offenders are chosen to take some other path, usually the status quo.

Such nonspecific selection gives researchers unbiased data on how one group fared versus the other. But it creates ethical qualms, as offenders are left to the whims of fate.

"People who we know need treatment, housing, things that we know could change a life," says Taylor, "because it's a random controlled trial, they are not allowed."

The people pushing for RCT research point out that's the reality in many fields already.

"In medicine, it is absolutely recognized that, until the medicine has been proven, you can't give it to people," says Clackamas County District Attorney John Foote. "It's life and death oftentimes for people who are sick."

Officials like Taylor argue those tough choices aren't necessary. By comparing people who participate in a targeted program to offenders of a similar age, race, and perceived risk who have not, Taylor says we can already get a reasonable idea of an effort's effectiveness.

The debate goes on and on, and it's currently coming to a head, as a state commission holds hearings that will shape how money earmarked for prison reform flows to counties.

In 2013, the Oregon State Legislature passed a series of sentencing changes that have eased demands on state prison beds. The money saved—once projected to be as much as $70 million for 2015-2017—is supposed to help counties reshape their criminal justice systems. But just how they qualify for the money is an open question.

Currently, the draft rules state any new projects funded with the money must be backed by "rigorous testing."

"RCTs are the preferred method of rigorous testing and will be given the greatest weight" in applications for cash, the draft rules say. And if a proposed program doesn't have an RCT backing it? Counties would have to concoct one.

The requirement has chafed Multnomah County justice officials enough that a coalition—including the presiding judge, a county commissioner, several probation supervisors, and more—all showed up at a public hearing earlier this month to discuss the rules.

"To hear the amount of concern that's being raised gives me great cause for concern," County Commissioner Judy Shiprack testified at the hearing.

The rule-making committee plans to consider testimony from similar hearings around the state before releasing final rules in several months—a release that's being anticipated in courthouses throughout Oregon. And though DAs haven't spoken out at all of those forums, they've hardly been silent.

In August, Gardner arranged for University of Cambridge criminologist Lawrence Sherman—regarded worldwide as an expert on research techniques—to speak by phone to the committee drawing up the rules. A transcript of the conversation goes on for 53 pages, but Sherman summed up his position nicely early on.

"The point is that there's no free lunch," he said. "You can't get away from the question of, 'How do we know what works?' The only way we can know that is to do a randomized trial...."