Topics: DATA

Adam Benforado in the New York Times  on using …”lessons from behavioral science to make police and courts more fair…. WHAT would it take to achieve true criminal justice in America?

Imagine that we got rid of all of the cops who cracked racist jokes and prosecutors blinded by a thirst for power. Imagine that we cleansed our courtrooms of lying witnesses and foolish jurors. Imagine that we removed every judge who thought the law should bend to her own personal agenda and every sadistic prison guard.

We would certainly feel just then. But we would be wrong.

We would still have unarmed kids shot in the back and innocent men and women sentenced to death. We would still have unequal treatment, disregarded rights and profound mistreatment.

The reason is simple and almost entirely overlooked: Our legal system is based on an inaccurate model of human behavior. Until recently, we had no way of understanding what was driving people’s thoughts, perceptions and actions in the criminal arena. So, we built our institutions on what we had: untested assumptions about what deceit looks like, how memories work and when punishment is merited.

But we now have tools — from experimental methods and data collection approaches to brain-imaging technologies — that provide an incredible opportunity to establish a new and robust foundation.

Our justice system must be reconstructed upon scientific fact. We can start by acknowledging what the data says about the fundamental flaws in our current legal processes and structures.

Consider the evidence that we treat as nearly unassailable proof of guilt at trial — an unwavering eyewitness, a suspect’s signed confession or a forensic match to the crime scene.

While we charge tens of thousands of people with crimes each year after they are identified in police lineups, research shows that eyewitnesses chose an innocent person roughly one-third of the time. Our memories can fail us because we’re frightened. They can be altered by the word choice of a detective. They can be corrupted by previously seeing someone’s image on a social media site.

Picking out lying suspects from their body language is ineffective. And trying then to gain a confession by exaggerating the strength of the evidence and playing down the seriousness of the offense can encourage people to admit to terrible things they didn’t do.

Even seemingly objective forensic analysis is far from incorruptible. Recent data shows that fingerprint — and even DNA — matches are significantly more likely when the forensic expert is aware that the sample comes from someone the police believe is guilty.

With the aid of psychology, we see there’s a whole host of seemingly extraneous forces influencing behavior and producing systematic distortions. But they remain hidden because they don’t fit into our familiar legal narratives.

We assume that the specific text of the law is critical to whether someone is convicted of rape, but research shows that the details of the criminal code — whether it includes a “force” requirement or excuses a “reasonably mistaken” belief in consent — can be irrelevant. What matters are the backgrounds and identifies of the jurors.

When a black teenager is shot by a police officer, we expect to find a bigot at the trigger.

But studies suggest that implicit bias, rather than explicit racism, is behind many recent tragedies. Indeed, simulator experiments show that the biggest danger posed to young African-American men may not be hate-filled cops, but well-intentioned police officers exposed to pervasive, damaging stereotypes that link the concepts of blackness and violence.

Likewise, Americans have been sold a myth that there are two kinds of judges — umpires and activists — and that being unbiased is a choice that a person makes. But the truth is that all judges are swayed by countless forces beyond their conscious awareness or control. It should have no impact on your case, for instance, whether your parole hearing is scheduled first thing in the morning or right before lunch, but when scientists looked at real parole boards, they found that judges were far more likely to grant petitions at the beginning of the day than they were midmorning.

The choice of where to place the camera in an interrogation room may seem immaterial, yet experiments show that it can affect whether a confession is determined to be coerced. When people watch a recording with the camera behind the detective, they are far more likely to find that the confession was voluntary than when watching the interactions from the perspective of the suspect.

With such challenges to our criminal justice system, what can possibly be done? The good news is that an evidence-based approach also illuminates the path forward.

Once we have clear data that something causes a bias, we can then figure out how to remove that influence. …(More)