Tuesday, August 24, 2010

Crime prediction software: an idea whose time has come?

I read a news report today (via Drudge) claiming that software capable of identifying and/or predicting those most likely to commit crimes was to be rolled out soon.

Developed by Richard Berk, a professor at the University of Pennsylvania, the software is already used in Baltimore and Philadelphia to predict which individuals on probation or parole are most likely to murder and to be murdered.

. . .

If the software proves successful, it could influence sentencing recommendations and bail amounts.

"When a person goes on probation or parole they are supervised by an officer. The question that officer has to answer is 'what level of supervision do you provide?'" said Berk.

It used to be that parole officers used the person's criminal record, and their good judgment, to determine that level.

"This research replaces those seat-of-the-pants calculations," he said.

. . .

Beginning several years ago, the researchers assembled a dataset of more than 60,000 various crimes, including homicides. Using an algorithm they developed, they found a subset of people much more likely to commit homicide when paroled or probated. Instead of finding one murderer in 100, the UPenn researchers could identify eight future murderers out of 100.

Berk's software examines roughly two dozen variables, from criminal record to geographic location. The type of crime, and more importantly, the age at which that crime was committed, were two of the most predictive variables.

. . .

Scientifically, Berk's results are "very impressive," said Shawn Bushway, a professor of criminal justice at the State University of New York at Albany who is familiar with Berk's research.

Predicting rare events like murder, even among high-risk individuals, is extremely difficult, said Bushway, and Berk is doing a better job of it than anyone else.

But Berk's scientific answer leaves policymakers with difficult questions, said Bushway. By labeling one group of people as high risk, and monitoring them with increased vigilance, there should be fewer murders, which the potential victims should be happy about.

It also means that those high-risk individuals will be monitored more aggressively. For inmate rights advocates, that is tantamount to harassment, "punishing people who, most likely, will not commit a crime in the future," said Bushway.

There's more at the link. Very interesting reading.

Part of me is very concerned about the implications of this software for civil and human rights. I hate the thought that someone who's reformed, and who may be a very low risk for re-offending, might be highlighted by this software as a major risk, and therefore be subjected to much tighter and more intrusive monitoring than he might otherwise receive. Indeed, he might get so annoyed at the intrusion that he commits another crime almost as a way of 'getting his own back' on the system that he regards as oppressing him. I think this is a very real danger.

On the other hand, I've been a prison chaplain. Routinely we staff saw prisoners released whom we knew - knew, beyond a shadow of doubt - were still extremely dangerous to society. Their conduct behind bars, their attitude . . . everything about them screamed a warning to us: but we couldn't keep them incarcerated. They'd done their time, and no matter how sure we were that they still represented a danger to others, we weren't allowed to keep them behind bars.

If this computer system can somehow be used to independently identify such people, and the knowledge of corrections staff can be linked to its predictions, the combination might be a very effective tool to ensure that likely re-offenders are monitored so closely that their recidivism becomes almost impossible. Will this represent an invasion of their civil liberties? Yes, I'm afraid it will. Can it be justified? On strictly Constitutional grounds, no, I daresay it can't. Is it therefore wrong to employ such means? Only the courts can answer that . . . but I'd like to point out that in many states, a convicted felon automatically loses certain rights (e.g. the right to vote). Given such restrictions, should a felony conviction also carry with it the loss of the right to privacy, at least as far as likely re-offenders are concerned? I can't answer that . . . but I can recall seeing hardened offenders let go, and later reading about their arrest for committing the most violent, vicious and horrific crimes after their release.

What say you, readers? Is there any circumstance in which the use of such predictive software can be allowed to trump, or at least limit, one's human and Constitutional rights? If so, what circumstance(s)? If not, what alternative can you suggest to minimize the risk of re-offense? Let's hear your views in Comments.



Popgun said...

I'd have to say - NO, predictive software should not be used in this way. Talk about a witch hunt!

I am against hate crime legislation for the same reason. With those laws, your punishment goes up if the jury believes you think in a certain way. How do they know what you are thinking?

The inside of your head is the last bastion of freedom. No law should apply based on what you are thinking - only on your actions. Because you can think one way and yet choose not to act in that way.


Rich said...

A societal good may come at too high a price.

Two quotes come to mind.

Quis custodiet ipsos custodes?

The road to hell is paved with good intentions.

skidmark said...

As a former correctional bureaucrat, I can see this being utilized regardless of the civil rghts implications. What's worse, I see it beng used on the front end, where judges look for information that will assist them in determining the "appropriate" sentence to impose.

Those whose profile suggests they are "more positive" for committing crimes will receive longer sentences than those who are "less positive". But at the end of the day, it's all still guesswork. Even though it was "common knowledge" that spouse-murderers were the least likely to recidivate, there were those few who managed to do so.

Thinking that this will reduce the crapshhot that is the criminal justice system is about as well-founded as relying on phrenology - which, by the way, worked sooo well, dinnit?

stay safe.

Anonymous said...

My first reaction is: violent sex offenders. My second reaction is: um, this could be abused so easily that I really don't think it is worth it.

Anonymous said...

1. What system are they setting up to track the accuracy of these predictions? Without feedback, it is just guessing.

2. Just looking at parolees, my alarm bells aren't raised much, but I have a sinking feeling they would see this system as successful and apply it to all sorts of things. Wouldn't you just love court judges running all their decisions by a computer to tell them how to rule or how to impose sentences? Criminal Justice meets the modern grade school bureaucracy.


Diamond Mair said...

Our Progeny is dating a parolee/probationer {sorry, I'm not sure of the distinction} - when still a young boy, he was encouraged by his "father" to fight/do drugs/drink - one night, while under the influence of alcohol AND drugs, he became {justifiably} angry at someone, but went WAY overboard - the other man survived, but D. went to prison - he's now in a building trades training program {top of his class}, he's not drinking/doing drugs {he inadvertently popped 'positive' when he took some Nyquil for a cold, went back to jail for 2 weeks} - he's a GOOD KID, who made a horrendous misjudgment, is paying his debt to society {here in Texas, HE pays for all drug tests, plus other remuneration} - a program such as this would dog him for the rest of his life - sorry, unless they're able to fine-tune it where people such as D., who commit ONE violent crime, then their conscience kicks in, and they 'do the right thing', I'm against it ..................

Semper Fi'

Stingray said...

"Is it therefore wrong to employ such means?"

Absolutely and 100% it is wrong to employ this software. If it is rolled out for use on released prisoners, the odds that it will eventually be expanded to monitor people who have never committed a crime in their lives is close enough to 1 to make me want to arrange an accident involving certain hard drives and magnets. And maybe a fire in the backup center.

Whoops, I think I just pinged on someone's beta copy...

Dr. Feelgood said...


Wayne Conrad said...

I am against the idea of parole at all. If a man cannot be trusted with all of his rights, he should be behind bars. If he is not behind bars, he should have the full exercise of all of his rights.

I dislike parole for economic reasons: Parole makes it cheaper for government to restrict the rights of its citizens. You don't have to put every such citizen in prison, just some of them. The rest can be out on parole, rights restricted. But if government could only restrict the rights of inmates, it would have to be much picker about who got such treatment and who did not. Government would, out of economic necessity, have to let go of some of the non-crimes it uses to indoctrinate people into the prison system and turn them into permanent semi-citizens.