Pretrial risk assessment tools are increasingly popular amongst court decision-makers and elected officials. Why is that?

Proponents of using pretrial RATs market them as objective and able to limit the rampant racial, ethnic, class, and gender biases that can manifest in judges’ and magistrates’ decision-making; as a replacement for money bail; and as a tool to reduce jail populations and mass incarceration.

Sold as “Objective”

For decades, researchers, criminologists, court administrators, and some judges have seen pretrial RATs as accurate and reliable predictors for pretrial “failure.”1Pretrial Justice Institute: Pretrial Risk Assessment: Science Provides Guidance on Assessing Defendants

Developers of RATs may sell them as being:

Some pretrial RATs may seem potentially less racially biased than a judge or magistrate on their own, since they are standardized and computerized. Most risk assessments don’t include explicit questions around the race of an accused person.  

Some argue that RATs bring “objective” data analytics11Anne Milgram: Why smart statistics are the key to fighting crime, TED Talk into the legal system, a system which otherwise relies on judges’ instincts, judgement and experience. Their assumption is that the predictions RATs make are fairer than the unstructured judgement of magistrates or judges. 

They argue that using such analysis “keeps communities safe” while allowing jurisdictions to increase “fairness” and release more accused or convicted people.12Anne Milgram: Why smart statistics are the key to fighting crime, TED Talk

Judges on their own are certainly already biased. One study, for instance, found that judges in Miami and Philadelphia were biased against Black defendants and relied on inaccurate stereotypes about dangerousness to make release decisions.13David Arnold, Will Dobbie, Crystal S Yang: Racial Bias in Bail Decisions, The Quarterly Journal of Economics

Our own interviews with various pretrial services agencies across the country showed that many are enthusiastic and hopeful about using risk assessment tools because they supposedly standardize a subjective pretrial process and seem like they will inject objectivity  and science into what they know is a deeply flawed system.

Several interviewees spoke to the biased decision-making they have seen from judges and magistrates, and felt that though they know RATs are not perfect, they serve as a balance to judges’ and magistrates’ otherwise unchecked power over pretrial outcomes.

Replace Money Bail

Risk assessments are also sold by developers an essential piece of the puzzle for jurisdictions working to limit or end their use of money bail14Emmeline Clien: Here’s how to help end cash bail, The Nation in pretrial decision-making systems.

In an environment where pretrial decision-makers often incarcerate people pretrial using money bail, RATs are meant to evaluate the risk of accused people being released based on what jurisdictions “care about” — such as an accused person’s supposed flight risk or predicted dangerousness to the community. Developers sell RATs as having the power to make judicial decisions easier and fairer, supposedly resulting in less racism and more transparency in pretrial decisions.15Matt Henry: Risk Assessment: Explained, The Appeal

We know that using money bail and pretrial incarceration are both unjust. Money bail especially creates a system of incarceration based on wealth and criminalizes poverty.16ACLU: Bail Reform

People of color and poor people are hit the hardest as they are often unable to pay bail, meaning they have to either stay in jail before trial, take a chance on the bail bond industry, or plead guilty to their charges17Stef W. Kight: Ending Cash Bail, Axios  in order to come home to their families.

In an environment where elected officials and decision-makers worry about accused people failing to appear or getting arrested again if released pretrial, RATs might seem like a good solution to get judges and magistrates to let more people go. It is the fear of violence, and the fear of flight from prosecution, that drives systems to enshrine risk assessment; even though rearrest and violence rates and failure to appear rates are actually quite low in most jurisdictions.

Using the term “failure to appear” in predicting risk makes forgetting or being unable to attend court seem far more intentional and malicious than it actually is.18Ethan Corey and Puck Lo: The ‘Failure to Appear’ Fallacy, The Appeal In reality, when people don’t appear in court, it is usually due to the systemic impacts of poverty.

In a national study of thousands of criminal legal system records, researchers found that only 16% of all defendants released pretrial were rearrested and only 1.9% were rearrested for a violent crime.19Shima Baradaran and Frank McIntyre: Predicting Violence, Texas Law Review

For failure to appear, data from a 2009 study show that 83% of felony defendants released pretrial came to all court dates, and only 3% of released felony defendants did not come to court at all after a year.20Lauryn P. Gouldin: Defining Flight Risk, The University of Chicago Law Review

Reduce Mass Incarceration

Elected officials and judicial decision-makers often describe RATs as the answer to a lot of the problems in the criminal legal system,21Angèle Christin: Algorithms in practice: Comparing web journalism and criminal justice, Big Data & Societyespecially mass incarceration and the overcrowding of city and county jails. 

The issue of reducing overcrowding is particularly pressing because the massive increase in jail populations over the last two decades is largely due to pretrial detention.22Wendy Sawyer and Peter Wagner: Mass Incarceration: The Whole Pie 2019, Prison Policy Initiative

Proponents of RATs therefore market RATs as a way to decrease jail populations.23Phillip Knoxx and Peter Keifer: The Risks and Rewards of Risk Assessments, National Center for State Courts Our interviewees expressed concerns about jail overcrowding, and many felt risk assessment tools would or already do help reduce the number of those detained. 

In some places, jurisdictions have embedded pretrial RATs into their decision-making systems and have, at the same time, seen some reductions in their jail populations. However, there isn’t clear evidence that the use of the tool has been the cause of or a direct part of that decarceral trend in those jurisdictions as the use of RATs is often accompanied by other non-algorithmic reform plans with explicit focus on decarceration.24John L. Koepke and David G. Robinson: Danger Ahead: Risk Assessment and the Future of Bail Reform, Washington Law Review

In other places, it is very difficult to attribute any changes in jail size to the risk assessment itself.

See our section on the impact of RATs to read more about instances where using a risk assessment has not led to smaller jail populations or decarceration.