A new report from the Pretrial Justice Institute (PJI) says that despite years of advocating for the use of Risk Assessment Instruments (RAI), research and data indicate the tools are dangerous and have many unintended consequences.
Their main consequence, according to the study, entitled “The Case Against Pretrial Risk Assessment Instruments,” is that overall, RAIs cannot do what they’re meant to: they cannot accurately predict whether an individual detained to face charges represents a risk to the community if he is released before trial.
The assessments are inherently flawed, perpetuating structural racism due to the biased data the tools are based on, resulting in a misleading assessment of risk, said PJI, a nonprofit advocacy group.
“RAIs simply add a veneer of scientific objectivity and mathematical precision to what are really very weak guesses about the future,” the report authors explain.
There is a wide variety of RAIs currently in use around the country. Most are defined as “actuarial tools,” meaning they do not give the court specific individualized information about the person facing charges.
Instead, using a plethora of data collected on people who committed crimes similar to the individual in question, the RAIs create categories and show the user the statistical likelihood of certain kinds of behavior before the disposition of their current case.
The tools assess, for example, the risk of reoffending, or the likelihood that the person will show up for a scheduled trial date.
While there is a vast amount of data backing the systems, it’s not individualized enough, resulting in the wrong call potentially being made, the report authors maintain.
This is in part because RAIs do not distinguish between people who have the means to flee the area, and people who don’t have the resources, such as community members without access to transportation or are experiencing housing instability.
Because the tools fail to distinguish this, someone may be incorrectly deemed a “flight risk,” thus hurting their ability to be freed pre-trial.
The same sentiment could be expressed for RAIs and their failure to reliably predict the likelihood that someone will commit a violent crime.
In Washington, DC, for example, 88 percent of all people arrested are released before trial, and 89 percent of them remain arrest-free.
Of those who are released, only 1 percent are rearrested for a violent crime — proving that assessment tools that prioritize violent crime recidivism are not reliable at identifying a second “statistically rare event,” since 13 percent of reoffenders commit nonviolent crimes before trial, the authors explain.
On the flip side of these numbers, The Public Safety Assessment, developed by Arnold Ventures, “contains a flag suggesting someone is at a higher risk for being arrested for what they label a ‘New Violent Criminal Activity.’” However, over 9 times out of 10, individuals who receive that flag on their assessment are not arrested for a violent offense while on pretrial release.
To make matters worse, the PJI report found that many RAIs even combine the statistical risk of someone missing court and the risk of any re-arrest into a single factor of “pretrial failure” — further muddying the accuracy of the assessments, the report details.
Biased Data Lead to Biased Outcomes
“Pretrial justice requires a radical reconstruction, which prioritizes racial equity in the presumption of innocence and pretrial liberty, and commits to long-term racially equitable solutions,” said the report.
A study cited by the PJI found that of 175,000 people arrested in New York City for similar crimes carrying similar sentences, found that “23 percent of Black defendants would have been classified as high-risk and flagged for detention under most RAIs, compared with 17 percent of Hispanic defendants, and only 10 percent of white defendants.
This bias stems from the crime data available which are used by the tools to gauge risk. That complicates the picture since studies suggest that policing is inherently biased against minorities and people of low socioeconomic status.
In other words, biased data in the risk assessment tools leads to biased outcomes for individuals, the authors explain.
“It is time to put away RAIs and forge an approach that does not perpetuate racial inequality, court involvement, debt or poverty, or create barriers to pretrial liberty and the presumption of innocence,” the report concludes.
“And we must prioritize helping people succeed—from assistance with court appointments to connecting people to support services — while addressing the needs of all people victimized by crime.”
The full report can be accessed here.