Can computer algorithms reduce America's prison population?
Some of the country's top criminologists, prosecutors, defense attorneys and judges gathered at New York University School of Law last week to examine new data-based tools and alternative programs aimed at helping judges with sentencing and pre-trial decisions.
“It's about learning to use our prison resources for those we're afraid of, not for those we're mad at,” Michael Wolff, Dean of the Saint Louis University School of Law, told the Friday conference.
Wolff, a former Chief Justice of the Supreme Court of Missouri, demonstrated a website used by judges in certain Missouri courts that compiles risk factors for individual defendants, such as prior convictions, age, substance abuse history, education and employment.
The information is entered into an automated form that generates comparative data on incarceration costs as well as recidivism rates for similar defendants who have been in the Missouri system in the last five years.
Wolff suggested that the program could combat inconsistencies between judges.
He cited a study completed by the Missouri Sentencing Advisory Commission three years ago which tallied the percentage of inmates who were non-violent offenders incarcerated in each of the state's counties.
They ranged as low as 27 percent in one county to as high as 80 percent in another.
“The information we give (judges) introduces them to the concept of risk assessment, because risk assessment is what parole and probation need,” Wolff said.
While risk assessment is an important factor in sentencing and probation decision, it's even more important for bail, argued Anne Milgram, vice-president of the non-profit Laura and John Arnold Foundation (LJAF).
Nevertheless, fewer than 10 percent of courts nationwide currently use pre-trial data tools, Milgram said.
“We spend a lot of time talking about all these great things that happen on the back end, with people who are on probation or parole or re-entry, but to me the game is at the front end,” Milgram said.
“Who are we putting into the system and how do we handle the first interactions that people have with law enforcement?
Development of a tool similar to the one used in Missouri, but geared towards improving judicial decisions on bail and pre-trial release, was completed last week by the LJAF, according to Milgram
She said the organization worked with programmers, behavioral economists and a statistician from the Massachusetts Institute of Technology to identify key factors that can be used to compute risk.
But since few judges have either the time or ability for complex statistical analysis, LJAF's program is designed to be simple to use.
The goal of both the Missouri and LJAF tools is to highlight specific data measures that predict violence or recidivism.
“People love using social histories, but none of that crap matters,” Wolff said.
Prison Alternatives for Those with Violent Histories
Mark Kleiman—a criminologist who was recently hired by Washington State to advise regulators who are crafting the state's new legal marijuana market—noted a pioneering Hawaii program that works with those whose violent social histories suggest a high level of risk.
Hawaii's Opportunity Probation with Enforcement (HOPE) program, introduced in 2004, doles out “swift and certain” sentences that include intensive supervision—within the community—for offenders who have committed crimes including domestic and sexual offenses, Kleiman said.
In a 2009 study, co-authored by Kleiman, HOPE probationers were about 55 percent less likely to be arrested for a new crime, compared to Hawaii's general probation population. Nearly three-quarters were less likely to use drugs.
Despite the HOPE model's successes, Kleiman cautioned that an over-reliance on programs like HOPE could be counterproductive.
“You'll wind up with kids caught for scrawling dirty words on the pavement on six- months community corrections,” Kleiman said.
But what happens when too few defendants qualify for an alternative program?
When Bridget Brennan, Special Narcotics Prosecutor for the City of New York, gave an upbeat report on the city's drug courts—which emphasize treatment— others noted that it excludes those imprisoned for a felony in the last 10 years, as well as anyone with previous violent offenses or who is currently on parole or probation.
Most other offenders in New York are relegated to the city's labyrinthine criminal justice system, said Gabriel Sayegh, director of the Drug Policy Alliance of New York.
He added that the restrictions reflect the “tough on drug crime” mindset that was popular in the 1970s and 1980s.
In contrast, he said, changes to health care regulation under the Patient Protection and Affordable Care Act (commonly known as “Obamacare”), require insurance companies to cover drug abuse as a health issue.
The effect on those struggling with drug addiction, he said, is confusion.
“When people are with someone who's overdosing, why don't they don't call 911? Sayegh asked.
“Because they're afraid of (law enforcement).”
Graham Kates is Deputy Editor of The Crime Report. He welcomes comments from readers. He can be found on Twitter, @GrahamKates.