« “MAGA Bomber” Cesar Sayoc sentenced to 20 years in prison despite LWOP guideline range | Main | Notable Govs make the case for pressing forward with additional criminal justice reforms »

August 5, 2019

Are pretrial risk assessment algorithms really part of "socialist agendas that are sweeping this country"?

The question in the title of this post is prompted by this curious new Fox News commentary authored by US Senator John Kennedy under the headline "Bail, bond decisions are being made today with algorithms -- That puts your safety at risk."  Here are excerpts:

Jurisdictions across the U.S. are snapping up algorithms as tools to help judges make bail and bond decisions. They’re being sold as race- and gender-neutral assessments that allow judges to use science in determining whether someone will behave if released from jail pending trial.

Really, they’re a dangerous collision of the poorly vetted cost cuts and socialist agendas that are sweeping this country.

The algorithms scare me because they’re being implemented for the same reason as the early release programs that are getting people killed.  The goal isn’t to protect public safety.  It’s to empty jail cells and release dangerous criminals on their own recognizance.

As a member of the Senate Judiciary Committee, I’m concerned about the recklessness of public policy that endangers people’s lives, especially in minority communities, where crime often is such a scourge.  These algorithms -- called pretrial assessment release tools -- are the equivalent of using a Magic 8 ball in courtrooms.  The results are disastrous to communities and great for criminals.

In my home state of Louisiana, New Orleans decided a few years ago to reduce the jail population. City officials started using a pretrial assessment release tool that was available for free from a nonprofit founded by a former hedge fund manager who became a billionaire through risky investments that turned into gold.

Do you know what happens when you allow a hedge fund manager to restructure your criminal justice system? You get a model that’s fraught with risk.

The new tool comes into play when someone is arrested on a felony charge, such as robbery or rape. The tool comes up with a score of one to five based on the defendant’s age, criminal history and several other factors. A “one” is considered a low risk to public safety. A “five” is considered justification for maximum supervision.

You would think that a risk level of “one” would be limited to people who jaywalk or shoplift. You would be wrong. In practice, a “five” apparently is reserved for people who kill busloads of nuns.  Ordinary thugs get a “one” as long as they promise that they’ll spend all their time in church and attend every court appearance.  They don’t have to regularly check in with a court officer or even call once a month....

The Metropolitan Crime Commission found that 37.6% of the people arrested for violent felonies in New Orleans during the third and fourth quarters of 2018 received the lowest risk level of “one.”  That included more than 32% of the people arrested for homicide and 36.5% of the people arrested for rape.

Algorithms diminish public safety in this country.  They ask us to pretend that lengthy arrest records and violent crimes don’t matter. They ask police to scoop up the bad guys only for the courts to immediately release them.  They turn us into a bad joke.

The use of risk assessment algorithms, whether pretrial or at sentencing or in the prison system, is an important modern criminal justice development that justifies much scrutiny and can be criticized on many grounds. But this commentary by Senator Kennedy reads a bit like a parody.

For starters, one of the main reasons risk assessments are appealing is because judicial decision-making without the help of data can itself often seem a lot like "Magic 8 ball" decision-making.  Moreover, all sound risk-assessment tools factor in arrest records and violent crimes, so they cannot properly be attacked for pretending that these past acts "don’t matter."  And, most amusingly, I cannot  quite fathom how efforts to make criminal justice decisions based on useful and relevant data amounts to part of "socialist agendas." 

I would welcome Senator Kennedy encouraging the Senate Judiciary Committee to hold hearings about the pros and cons of using risk assessment algorithms in modern criminal justice systems.  But, since he suggests giving judges more information is part of "socialist agendas that are sweeping this country," I worry he might think informing Senators more about these matters also somehow has mysterious sinister socialist undertones.

August 5, 2019 at 07:07 PM | Permalink

Comments

Good Lord he's a demagogue. Him and Cotton. What does he get out of this?

I sure hope getting rid of Trump eventually rids the GOP of the special kind of stupid they're indulging in right now.

Posted by: Fat Bastard | Aug 5, 2019 10:11:20 PM

What a complete moron.

Posted by: whatever | Aug 6, 2019 11:02:38 AM

In my opinion: the Grand Jury should have a bigger role in the Court system, as long as they are educated on Constitutional Law, know the rights of Americans, and are not handpicked by Judge or Prosecutor.

Posted by: LC in Texas | Aug 6, 2019 12:54:54 PM

As someone who self-identifies as a socialist I find in sad. "Socialist" has become a slur in the same way that "pedophile" has become a slur. It doesn't even matter if it is true or not, it us just a token of dislike. I dislike pretrial risk assessments too, precisely because they are NOT socialistic.

Posted by: Daniel | Aug 6, 2019 3:10:58 PM

It is stunning that about one-third of those charged with violent crimes (murder and rape) in New Orleans are assigned the lowest risk assessment rating and then released. A rating of "one" should be reserved for those charged with non-violent misdemeanors. What data are available to determine the reliability of these assessments -- including the percentage of those who commit other violent crimes while back on the streets or fail to appear for hearings?

Posted by: MIke | Aug 11, 2019 1:39:08 PM

The question here should be prospective. Do the defendants with the lowest risk levels in fact go on to commit the smallest number of crimes.

Posted by: William C Jockusch | Aug 12, 2019 7:02:04 PM

Post a comment

In the body of your email, please indicate if you are a professor, student, prosecutor, defense attorney, etc. so I can gain a sense of who is reading my blog. Thank you, DAB