Member Directory
Member Directory

The Dangers of “Big Data” in the Criminal Justice System

Following statistics blindly will lead to unintended consequences. In the criminal justice system it could result in gross miscarriages of justice.

By Mark M. O’Mara, a Partner, and Shawn Vincent, Communications Director, of the O’Mara Law Group in Orlando, FL.

Attorney General Eric Holder issued a warning about the dangers of using risk assessments based on “big data” statistics in criminal sentencing decisions.

In an address before a standing-room only crowd at this year’s annual meeting of the National Association of Criminal Defense Lawyers, Holder said, “By basing sentencing decisions on static factors and immutable characteristics -- like the defendant’s education level, socioeconomic background or neighborhood -- they may exacerbate unwarranted and unjust disparities that are already far too common in our criminal justice system and in our society.”

Sometimes “big data” is wrong

“Big Data” is the art and science of aggregating information from massive data sources and producing actionable results. It’s how Netflix can guess what movie you want to watch, and it’s how Google can guess what you’re searching for before you type it in. And now in states like Pennsylvania and Virginia, big data is being used to determine how long a sentence should be, or whether a defendant should await trial behind bars or out on bond.

The problem is that sometimes Netflix suggests bad movies, and sometimes Google misjudges what we’re looking for. Sometimes “big data” is wrong, and in the criminal justice system, it could mean the wrong people go to jail, for too long, and for the wrong reasons -- and that is intolerable.

But it doesn’t have to be that way.

Holder also mentioned during his address that a balance between “big data” and judicial discretion might serve the interests of justice.  This is where legendary chess master Garry Kasparov can teach legal advocates how to approach “big data.”

Advanced chess

In 1997, IBM’s Deep Blue supercomputer beat Kasparov in a highly publicized chess match.  It marked a milestone in computing, and it caused Kasparov to reassess the game of chess.  Humans and computers approach chess in different ways, Kasparov determined, and while a supercomputer can beat a chess master when it comes to calculating infinite possibilities, it cannot match a human’s intuition and creativity.

Kasparov concluded that a human playing with a computer partner could be superior to grandmasters and supercomputers playing on their own.  He called this hybrid model “advanced chess,” and in 2005, an online tournament was staged where any combination of humans and computers could compete on a team.  Masters and supercomputers competed, and in the end the winners were armchair chess enthusiasts working with a PC and off-the-shelf chess software.

We’ve seen big data -- quantification -- disrupt industries before.  The book and movie “Moneyball” shows how statistical analysis changed baseball. We’ve seen computer trading take over the financial markets, and we’ve seen people like Nate Silver predict elections outcomes with unsettling accuracy.

Inevitably, however, a blind reliance on statistical data can lead to a mass disaster that, in hindsight, should have been obvious.  The financial meltdown of 2008 provides a perfect example of what can happen when the quantification of an industry goes too far. These days, most industries -- professional sports, finance, and politics, just to name a few -- have learned from the mistakes and evolved to a hybrid model that marries human experience and intuition with the benefits of “big data” statistical analysis.

Heed the lessons

Now that we see the criminal justice system beginning to embrace “big data,” we should heed the lessons from the industries previously disrupted by the “big data” revolution.  We know that following statistics blindly will lead to unintended consequences. In the criminal justice system it could result in gross miscarriages of justice, and it could amplify existing inequities.

But imagine a judge armed with statistics that indicate the defendant before her is about to receive a sentence that is disproportionately aggressive compared to the national average for similar charges? (I'm concerned that most big data analyses might suggest harsher sentencing schemes.) And what if she had data that showed African Americans were disproportionately recipients of the aggressive sentences? Would it give her a chance to evaluate if there was any inadvertent bias at play in her courtroom?

  • What if we had data to show that pretrial diversions that require clean criminal records favor white defendants?
  • What if we had data to show that a 10% reduction in prison sentences do not negatively affect crime?
  • What if a defense attorney could cite “big data” to show the benefits of counseling over incarceration?

The combination of “big data” and zealous advocacy has the possibility to correct inequalities in our justice system, and to lead to more compassionate and effective sentences for our citizens accused.  But for that to be the case, it means that that attorneys, the zealous advocates, must understand the perils and benefits of statistical analysis, and to insist the “big data” be used only to further the interests of justice and equality. We just need to be careful with what big data is going to show.

Shawn Vincent is the Communications Director at O’Mara Law Group in Orlando, Florida. He is a writer, a technology advocate and a media liaison. He has an ongoing interest in the relationship between message and medium.

Find an Attorney

The National Trial Lawyers stands as an esteemed, exclusive association comprising top-tier trial attorneys nationwide. Whether you require a Civil Plaintiff Lawyer or a Criminal Defense Lawyer in your state, our network ensures access to premier legal expertise tailored to your needs.
Find Attorney

Read More Legal News

© Copyright 2022, All Rights Reserved | National Trial Lawyers
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram