Olathe Ted Kim of Lenexa was busted for drugs last year but deserved a break. Statistics said so.
Forget the old gut feeling of a judge or probation officer. Kim's answers to 54 questions got him a far more limited probation than the long hours of programs he could have faced.
He scored "very low risk" on a statistical tool that predicts how likely an offender is to fail probation or parole. Putting such people in highly intense programs can actually make them more likely to become criminals, experts say.
For six years, Johnson County District Court has been a pilot project — the only Kansas district to test the statistical tool called Level of Service Inventory Revised. But state court officials say they finally are getting money from fees to teach the system to court services probation workers statewide and will start next year.
Alex Holsinger, associate professor of criminal justice at the University of Missouri-Kansas City, evaluates how Johnson County staff use what he says is among the best of many such statistical tools.
"They're doing it as well as anyplace in the country," he said.
Results mirror how workers score the offenders, with 6 percent failing among the very low risk group and up to 53 percent failing among the highest risk.
He also found that targeting more programs to higher-risk people appeared to work. The number of those sent from community corrections probation to jail or prison dropped from 40 percent in 2004 to 23 percent last year.
Amy Bond, a probation officer with the Johnson County Department of Corrections, sat recently with Kim, 25, to explain how she assessed him.
She gave him the test before his sentencing early this year, after he had pleaded guilty to selling the prescription narcotic oxycodone and possessing marijuana.
The questions are open-ended and provide answers that give scores in each of 10 categories, things like criminal history, companions, substance abuse problems and attitudes.
Kim, who works as a chef, scored zero risk on five categories but high risk on leisure/recreation and moderate risk on substance abuse and companions.
He took prescription narcotics for a burst disc in his back, he said, fell in with the wrong friends, used such pills recreationally and sold them.
A friend introduced him to what turned out to be an undercover police officer, and he sold him four pills for $200.
Kim scored at 11 percent risk level on the test. Based partly on that, a judge sentenced him to 14 days of shock time in jail and 18 months of probation.
Among categories that Kim scored zero risk on was attitudes, meaning he did not need long hours of work on changing criminal thought patterns. He doesn't see an unlocked car with a cell phone in the seat, for instance, and think, "That phone is mine." He doesn't make excuses like "everyone does it" or consider cops and judges out to get him.
He did not belong among more hardened offenders, Bond said.
"I didn't want him sitting out in the lobby with some other clients picking up new skills."
By now he has done drug treatment, goes to a therapist and still attends 12-step meetings.
"I got rid of all the bad friends and kept all the good friends," he said. "I see (probation) as almost a blessing as far as improving myself."
Such assessment tools have been used for generations but are getting better, and some liken the Level of Service Inventory Revised to a sleek new cell phone. Missouri uses a variation of a much older tool.
The Kansas Department of Corrections has also used the system for six years with parolees. It helped the state have the second-lowest growth in prison population nationwide from 2000 to 2009, said Roger Haden, deputy secretary of corrections.
Amid state budget problems, the tool will continue to target shrinking parole resources to those most likely to go back to prison, Haden said.
Edward Latessa, director of the School of Criminal Justice at the University of Cincinnati, gave a lecture to experts in Kansas City last month on best practices in corrections. Assessment tools, he said, "are the engine that drives correctional programs."
Latessa, an expert in evaluating such programs, cited new studies showing that high-risk offenders clearly did better with 200 hours or more in programs that address specific weaknesses.
But increasing the program dosage for low-risk people causes far more to fail, he said.
Throwing them into programs with higher-risk people can disrupt social networks and teach anti-social behavior, he said.
And forget corrections practices that flat do not work like yoga, physical conditioning, dancing, jogging, gardening, bowling or shaming people with things like pink underwear or diapers, Latessa said.
"Take a bunch of offenders and teach them bowling and you'll have a bunch of offenders who know how to bowl," he said.
Shame anti-social people, he said, and you get angry anti-social people.
Use the assessments and target weaknesses in the highest-risk offender, he advised.
"You got to go at this person with everything you got."