Interaction with Bail - Mapping Pretrial Risk

In some cases, pretrial RATs work alongside systems that preserve the use of money bail in pretrial decision-making. In other cases, tools are implemented in system changes that mostly remove the use of money bail from pretrial decision-making.

For example, as of December 2019, New Jersey,¹ California,² and New York³ are in the process of eliminating money bail and relying instead on pretrial RATs to decide who is detained and who is released pretrial.

Instead of having a bail schedule that mandates certain bond amounts for certain charges, these jurisdictions rely on RATs to predict risk, which then supposedly help judges and magistrates “accurately” decide who should stay in jail pretrial and what conditions of release look like for those who can be “safely” released.

Many proponents of RATs claim that they are an effective substitution for bail and a necessary part of the process of removing money bail from pretrial decision-making.⁴ In a number of our interviews, judicial officers noted that RATs were a component used in their jurisdictions as a part of recommending lower bails or recommending release without bail.

However, in some places, the use of a RAT instead of or alongside bail may actually be a part of increasing detention.

Baltimore, Maryland, has had various risk assessments in place for over 50 years. However, recent efforts to reduce the use of money bail lead to a 31% increase in the average daily jail population⁵ from March 2017 to March 2018. Rather than reducing bail amounts or reliance on bail, judges were simply holding more people without bail at all.

Lancaster County, Nebraska currently assesses those held in jail for three days, and the county has been discussing ending money bail. However, the Department of Corrections Director expressed concerns that getting rid of bail might not lead to reductions in the jail population⁶ because if the tool they use, the VPRAI, is applied earlier in the process and is used to make detention recommendations instead of bail, some of those currently out of jail on a bond would be detained based on their VPRAI scores.

In many jurisdictions, pretrial RATs can be a part of deciding whether or not someone receives a financial bond, or the amount of that bond, just like a bail schedule. The predicted risk level may be connected with a particular recommendation regarding bail.

In some of the jurisdictions we interviewed, the use of RATs has led to more people with low scores being released on their own recognizance, meaning they are released without having to pay a bond. For instance, the interviewee from the Michigan Supreme Court PSA pilot stated that a large number of low PSA scores meant that they were issuing more personal recognizance bonds than in the past.⁷

In other places, lower scores lead to a lower bond amount. Interviewees from Yakima County, Washington,⁸ and Nashville, Tennessee,⁹ stated that the bail amounts have gone down under their tool — in Nashville’s case, the interviewees attributed this to the fact that under their tool, bail amounts were less subject to judicial bias.

Several interviewees said that their RATs impacted both bond amounts and how many people were released without having to pay bail. The interviewee from Saline County, Kansas, stated that under their locally-developed pretrial tool, most individuals eligible for pretrial services have lower bonds or receive recognizance bonds.¹⁰ The interviewee from Alachua County, Florida, said that judges look at scores for guidance — low scores may result in release without bail or mean that someone receives a lower bond amount than those who receive higher scores. They did state that most people still get bail.¹¹

Some jurisdictions implement RATs only if the accused person cannot afford their bail or if they are are seeking a bail review process. The interviewee from Anne Arundel County, Maryland, stated that the county uses their tool for those appearing for bail review to help determine supervision level recommendations; most of those in the pretrial program are held without bail before their bail review.¹²

According to our interview with Armstrong County, Pennsylvania, the county uses the ORAS-PAT as part of a pretrial investigation for those who are seeking a bond reduction.¹³

In other places, the RATs were only used on bailable charges, meaning charges that were eligible for release on a money bail. The interviewee from Knox County, Tennessee, said that Knox only uses their locally created tool for those with charges eligible for a bond.¹⁴

In some of our interviews, however, the use of tools had no clear connection to bail at all.

Many jurisdictions stated that RATs had no impact on the amount or frequency of bail assigned, or that they did not track this data. For example, at the time of our interview, Hennepin County, Minnesota, said that it is unclear how their tool impacts money bail.¹⁵