Early this month, judges in five Michigan district courts started using a new tool to make bond decisions.
It’s part of a new pilot program spearheaded by the Michigan Supreme Court. The judges will use a standard risk assessment to help make those bond decisions.
Participating district courts are Flint, Hamtramck, Clinton Township, Mount Pleasant, and Escanaba.
Michigan Supreme Court spokesperson John Nevin says it’s a year-long pilot that aims to give judges more information.
“The idea here is to figure out, can we stop the practice of locking up defendants who are low-risk, who really aren’t going to be a threat to society or anybody in their community, and let them go home?” said Nevin.
According to the Supreme Court, more than half of Michigan’s jail inmates are pre-trial defendants. Many are there for non-violent offenses.
The flip side of that, Nevin says, is to free up space for truly risky and violent offenders, including those who are serving post-trial jail sentences
The risk assessment includes factors like whether defendants have failed to show up in court before, and drug use. Nevin says the main idea is to keep low-risk, pre-trial defendants out of jail just because they can’t afford to make bail. “And then once we have a solid test that’s working, we want to expand it to the rest of the state,” he said.
“No Michigan residents should be sitting in jail just because they can’t afford to pay their bail,” said Chief Justice Bridget McCormack. “Our goal is to help judges make bond decisions that protect rights, enhance public safety, strengthen communities, and save money.”
But some criminal justice system activists say that while the effort might be well-intentioned, it could backfire.
Amanda Alexander, head of the Detroit Justice Center, questions the use of risk assessments to make bond decisions.
“In other states that have implemented these tools, such as Kentucky, there's no evidence that they have reduced pretrial jail populations or racial disparities in pretrial detention,” Alexander said.
“While cloaked in scientific objectivity, these algorithms rely on factors that reflect the racial bias of the criminal legal system. They include data on prior arrests and prior instances when a defendant failed to appear for court. They don't tell us why someone missed their court date, or what could have helped them get there. Research has shown that data about prior arrests, charges, and failures to appear tell us more about the behavior of police and prosecutors than about the individuals appearing before a judge.”