LITTLE ROCK, Ark. — States are trying to reduce prison populations with secretive, new psychological assessments to predict which inmates will commit future crimes and who might be safe to release, despite serious problems and high-profile failures, an Associated Press investigation found.
These programs are part of a national, data-driven movement to drive down prison populations, reduce recidivism and save billions. They include questionnaires often with more than 100 questions about an offender’s education, family, income, job status, history of moving, parents’ arrest history — or whether he or she has a phone. A score is affixed to each answer and the result helps shape how the offender will be supervised in the system — or released from custody.
Used for crimes ranging from petty thievery to serial murders, these questionnaires come with their own set of risks, according to the AP’s examination.
Many rely on criminals to tell the truth, and jurisdictions don’t always check to make sure the answers are accurate. They are used inconsistently across the country, sometimes within the same jurisdiction. The same defendant might be scored differently in the same crime.
Supporters cite some research, such as a 1987 Rand Corp. study that said the surveys accurately can predict the likelihood of repeat offenses as much as 70 percent of the time if they are used correctly. But even the Rand study, one of the seminal pieces of research on the subject, was skeptical of the surveys’ overall effectiveness. It’s nearly impossible to measure the surveys’ impact on recidivism because they are only part of broader efforts.
Some surveys have the potential to punish people for being poor or uneducated by attaching a lower risk to those who have steady work and high levels of education. The surveys are clouded in secrecy. Some states never release the evaluations, shielding government officials from being held accountable for decisions that affect public safety.
“It is a vast improvement over the decision-making process of 20, 30 years ago when parole boards and the courts didn’t have any statistical information to base their decisions on,” said Adam Gelb, director of the Public Safety Performance Project at the Pew Charitable Trusts, which is working with the Justice Department to shape reforms nationally.
There are dozens of different surveys in use.
In the case of Milton Thomas in Arkansas, a parolee accused of raping a 70-year-old woman, the state’s predictions were inconsistent.
Three risk assessment surveys produced three different results. When Thomas, 58, was up for parole in 2013 after serving time for theft, the Arkansas Parole Board assessed him as a high risk to re-offend.
He was released in November of that year, and the state’s community supervision agency assessed him again. This one determined Thomas was a low risk, Thomas said, and required the minimum amount of supervision and no rehabilitative programs. After Thomas was arrested on the rape charge last July, the parole board assessed him again and downgraded his risk from high to moderate.
While these results seem to defy common sense, the board said the system worked and there was a simple reason for lowering Thomas’ risk: varying accounts of Thomas’ age when he was first arrested.
In 2013 in Texas, a routine risk assessment convicted rapist Darren Vann to be a low-risk to re-offend.
About a year later, in Indiana, Vann confessed to police that he killed seven women.