Law students are taught that there are four factors a judge should consider when a criminal defendant is sentenced: specific deterrence, general deterrence, rehabilitation and retribution. The Commonwealth of Pennsylvania is now proposing addition of a fifth factor: future criminality.
Buckle your seatbelts if you care about your liberty: roughriders wearing robes will soon by locking folks up based on assessments of what they think folks are likely to do next.
You are no doubt familiar with the concept of future criminality if you saw the 2002 film “Minority Report,” directed by Steven Spielberg. The film is set in Northern Virginia/Washington, D.C., in the year 2054. Police rely on psychics called “precogs” to predict when someone is going to commit a crime. The person is arrested based on an assessment of what they will do.
The film was a dark fantasy most civil libertarians hoped would never become a reality. Rather than rely on psychics, Pennsylvania proposes reliance on statisticians engaged in the complex art of “risk assessment.”
First, a word about the current lay of the land in the criminal courts.
We do not punish folks in the United States based on forecasts of what they will do in the future. The criminal law is retrospective. The state must prove beyond a reasonable doubt that a person committed a precisely defined offense before locking them up.
When folks are locked up, a sentencing judge is supposed to consider the deterrent effect of the sentence. Specific deterrence is intended to send a message to the offender himself: don’t do this again. General deterrence is supposed to signal others similar to the convicted to avoid doing what the convict has done.
Deterrence is a nice theory, but it generally fails in practice. Decades into the war on drugs, clogged prisons have deterred few; and, when it comes to murder, I’ve yet to meet the man or woman accused of the crime who tells me that they were thinking of what others did at the homicidal moment.
Rehabilitation is a goal honored in theory, but ignored in practice. The prisons remain overcrowded, understaffed, and chock full of folks who are mentally ill. We no longer institutionalize folks suffering from mental illness — we wait for them to cross a line, perhaps inevitably so, and then warehouse them.
Only punishment remains a viable goal of the system. And we do a good job of that. Consider the sentence imposed on Ricardo Myers in New Haven this week. The judge slapped him with 47 years in prison after a jury found him guilty of two counts of first-degree assault for a shooting, noting that the crimes were both “inexplicable” and “out of character” for the defendant. All at once, the defendant, 23, was reduced to a caricature of his worst moment.
Don’t get me wrong, I do believe that there are people who are a danger to society, and I believe such folks should be locked up for a good long time. But every defendant deserves a second chance — no one is the sum of their worst moment. At the 20-year mark, each should be evaluated for release back to the community. Only the United States imprisons so many for so long; we are the laughing stock of the world.
So what of this brave new world Pennsylvania contemplates? Into what navel will judges gaze to make predictions of future criminality?
Let me deliver the good news first: Pennsylvania is not proposing that uncharged folks be swept off the street to answer for acts they have not yet committed. As proposed, the new regime would simply permit judges to base sentences for committed crimes to include an assessment of how likely a person was to offend again.
Judges make these sorts of decisions on an informal basis all the time when they set bond or bail for those accused of a crime. Among the factors the Eighth Amendment to the Constitution permits a judge to consider is danger to the community. And judges typically have broad discretion in imposing sentence on a convicted person, bound only to sentence within the range lawmakers specify for a crime.
What alarms about Pennsylvania proposal is the method used to assess risk. Statistical data will be used to forecast dangerousness based on the characteristics of the accused — a common enough practice among insurance actuaries deciding how much to charge for an insurance policy, but a dangerous doctrine when used in a criminal justice system based on respect for individual persons.
There are reportedly 60 or so risk assessment tools on the market now. These attempt to score risk based on such things as family criminal background and criminal history to employment. These tools are currently used by prison, probation and parole officials to help determine a person’s custodial status. The Connecticut bail commissioners use a Byzantine formula to make bail recommendations.
But statistical evidence does not predict outcomes. At most, it recognizes patterns in large sets of data. Philosophers are quick to point out that, at best, statistics correlates facts: it does not account for causation, and it cannot predict individual conduct.
Hence, the horror of the Pennsylvania approach to sentencing.
It is paradoxical that at the very time this nation begins a serious debate about mass incarceration and second chances, Pennsylvania intends to adopt a course that will hold selected prisoners for longer periods of time. Whence comes this instinct to lock people up and throw away the key?
I repeat now, and will say it daily to whomsoever will listen: No one is the sum of their worst moment. Treating people as mere incidents to their statistical correlations is the very worst sort of way to evaluate people. It doesn’t work with regard to race, to gender, to sexual orientation. It won’t work with prisoners.
Shame on the Commonwealth of Pennsylvania.