Monday, August 08, 2011

LAPPL BLASTS COMPUTER PREDICTIONS OF FUTURE BEHAVIOR

California’s budget-saving prison releases and the attendant policy of keeping parole violators from being returned to prison have proven to be deadly. Time after time, California’s computer classifications of prison inmates as ‘non-violent’ and ready for release on non-revocable (unsupervised) parole have turned out to be dead wrong once those inmates have hit the streets. LAPD’s union is fed up with the state’s ineffective parole system.

PAROLE SYSTEM’S FATAL FLAWS EXPOSED AGAIN BY A DOUBLE MURDER
By Los Angeles Police Protective League Board of Directors

LAPPL Blog
July 29, 2011

Computers are a marvelous thing. They store information, churn data and generally make us all more productive. Often, though, there’s a need for human judgment and common sense to interpret a computer’s output.

A case in point is the California Department of Corrections and Rehabilitation’s computer program, Parole Violation Decision Making Instrument (PVDMI). It’s a blatant cost-cutting measure touted as a reliable way to predict which inmates are “low risk” and “non-violent," and thus eligible for early release from prison with no parole supervision and no notification to local law enforcement.

PVDMI determined that Zachariah Timothy Lehnen fit that category despite a history of arrests for robbery, domestic violence and drugs. As a result, he was released from prison long before his sentence was to end in November 2013. Back on the streets with no parole supervision, Lehnen is accused of murdering Lucien Bergez, 89, and Erica Evelyn Escobar, 27, on May 3, 2011 in Culver City.

Adam Treanor, president of the Culver City Police Officers’ Association, believes the public deserves a supervised release system that is realistic and makes community safety a top priority. “We need a system,” he said, “that empowers law enforcement officers with the ability to re-incarcerate any and all parolees who view their release from prison not as an opportunity to begin a new life as a productive member of society, but merely as a means to return to the very same criminal, predatory behavior that led to their incarceration in the first place.”

We are in complete agreement with Treanor and support his call to overhaul the fatally flawed system we have today. There will always be a need to supervise parolees just as there will also be a need to run a human reality check on the information a computer program puts out. It is, after all, a matter of life and death, as was proven in the case of Zachariah Timothy Lehnen.

1 comment:

bob walsh said...

It is, IMHO, not very hard to predict future violent behavior. However, predicting future non-violent behavior via a computer program is probably very much like predicting future non-violent behavior with a Magic 8 ball or a Ouija board. Much more a matter of luck than science. Pretending otherwise is essentially dishonest.

Bob Walsh