Voices in the criminal justice reform movement largely agree that a lack of money to post bail should not be the determining factor in whether a defendant is jailed before trial. However, some voices, especially in the civil rights community, are equally critical of an alternative approach: the use of pretrial risk assessments that use data on a person’s criminal history. 

Pretrial risk assessments have grown in popularity as tools for assessing whether a given defendant poses a risk to public safety if left free before trial. Part of this shift is the result of policy advocacy backed by Arnold Ventures, a leading criminal justice reform funder in recent years. But while Arnold has backed such tools as part of a larger push against what it calls “unjust pretrial detention,” the area of risk assessment has become controversial. Some civil rights advocates argue that these instruments reflect existing inequities in the nation’s criminal justice system and may even exacerbate them.

Now, thanks to support from another top criminal justice funder, that viewpoint is backed by stronger research.

Evaluating the Record

A new report funded by the John D. and Catherine T. MacArthur Foundation examines the weaknesses of pretrial risk assessment tools, and argues that their use should largely be abandoned and replaced by alternative forms of pretrial justice. The paper, “Civil Rights and Pretrial Risk Assessment Instruments,” was authored by Washington, D.C.-based nonprofit Upturn Inc., one of the recipients of funding under the MacArthur Foundation’s Safety and Justice Challenge.

MacArthur launched the Safety and Justice Challenge in 2015, aiming to safely reduce mass incacreration, as well as reduce racial and ethnic disparities in jails and prisons. Between 2015 and 2017, the funder lavished more than $100 million to support efforts by local communities to develop and implement solutions designed to reduce jail populations. And the money has kept flowing. Thanks to a new set of grants announced this week, MacArthur’s total commitments—which have engaged 51 cities and counties—now stands at $217 million.

Early evaluations suggest the work of the initiative is bearing fruit. A 2019 evaluation by RTI International found that average jail populations and racial/ethnic disparities declined in sites that received Challenge funding.

The University of Virginia (UVA) School of Law was one of many recipients of grant funding under the Safety and Justice Challenge. A grant of $800,000 from MacArthur enabled the law school to launch a collaborative project focused on the state of risk assessment instruments in the criminal justice system. Upturn Inc., the author of the report discussed here, was one of UVA’s project partners. Upturn is a technology policy organization interested in fairness and equity in the use of quantitative decision making tools.

A “Legitimacy Problem”

In the Upturn report, authors David G. Robinson and Logan Koepke cite what they call an “inherent legitimacy problem” in pretrial risk assessments: the fact that today’s criminal justice system, characterized by mass incarceration, racially inequitable laws, and differential treatment of people of color, is the very system that provides the data that inform such assessments. “There is, justifiably, distrust that tools developed on data reflecting this racially inequitable system will avoid perpetuating these patterns, let alone advance substantial reform.”

The report also criticized proponents’ claims that pretrial assessment tools are well-calibrated for different racial and ethnic groups, and thus race-neutral. Civil rights advocates acknowledge efforts at calibration, but insist that calibration cannot make an instrument free of racial bias. Because police arrest people of color at higher rates than white people, a well-calibrated tool could still classify more African Americans as high risk than whites. 

Further, the implementation of risk assessment instruments often occurs concurrently with other changes to pretrial justice, it is difficult to determine empirically if these tools drive reductions in incarceration or reduce racial disparities. The evidence from existing research to date is not conclusive.

In an increasingly data-driven society, the authors’ findings may seem surprising at one level. Rather than use the data gathered by these tools to make critical decisions about pretrial incarceration, the authors suggest that jurisdictions should largely abandon the use of such instruments.

Charting Another Path

So, what is Upturn’s alternative?

The report argues for ending the use of pretrial detention for suspects arrested for nonviolent offenses and minor misdemeanors. The practice of incarcerating such persons before trial keeps an estimated 450,000 behind bars, according to some estimates. Pretrial detention, the authors contend, should be “limited to a narrow set of serious charges.” Further, people who face such charges should have a thorough court hearing before being detained for any period of time exceeding a few days.

It will be interesting to see how this debate plays out over time within the criminal justice reform community and among funders. Arnold Ventures has a lot at stake in this area, having given millions to develop and promote its Public Safety Assessment tool. Clearly, though, there is still no consensus in the policymaker, advocate, or funder community on the efficacy of these tools.

Ending cash bail and eliminating the use of quantitative risk assessment tools may be politically difficult to achieve, a point Upturn concedes in its report. However, it is important to consider all  perspectives when trying to achieve the important goals of reducing mass incarceration, as well as curb racial and ethnic disparities in correctional systems. Alternatives to these tools exist, and policymakers should consider their use. 

Share with cohorts