Half of people released from S.F. jail before trial were accused of a new crime while free, according to four-year study

Roughly half of people charged with crimes and released from jail before their trials in San Francisco in recent years failed to show up for court, and a similar share were accused of committing a new crime while free, a new study found.

More than 1 in 6 defendants allegedly committed a new violent offense, according to the findings from May 2016 to December 2019 published by the California Policy Lab, based at UC Berkeley and UCLA.

The factors behind the statistics are complex, experts and advocates said, and present challenges for the city’s effort to reduce the number of low-risk people in jail before they’re convicted of a crime and get them the support they may need to better their lives.

The data doesn’t include many of the lowest risk defendants who get a citation reminding them to show up at court and are released immediately. Numbers could also be driven in part by homelessness and addiction that can fuel crimes and hinder people from showing up for court. One in three people in San Francisco jails were unhoused and nearly 3 out of 4 had a history of substance use in recent counts.

“Nobody can look at this report and say we’re doing great,” said Supervisor Catherine Stefani, a former prosecutor who has voiced skepticism about previously reported success rates of people released from jail before trial. “It validates the experience that people in San Francisco are feeling when they’re concerned about crime.”

Stefani plans to introduce legislation to require more extensive reporting of court appearance and reoffense rates, including crimes committed outside the city, for the duration of a defendant’s case. She also wants to implement the report’s recommendation for more intervention to help people keep their court dates.

The report comes at a pivotal moment as criminal justice reform sweeps parts of the country and some counties have moved to free more pretrial defendants, countering inequities in policing and in the bail system that left more poor people and people of color behind bars.

Assertions that San Francisco is too lenient with people accused of crimes has spurred two recall efforts against progressive District Attorney Chesa Boudin, though the new report’s data predates his time in office. Boudin, who ended cash bail for all criminal defendants in the city, and his supporters say the nation’s reliance on incarceration has proven to be a failure.

Historically, courts have relied on the bail system to release people charged with crimes, who can also be freed with electronic monitoring or to rehabilitative programs such as substance use treatment. People arrested for minor crimes are often let go with a citation that instructs them to show up for a court date.

San Francisco has sought to lead in pushing reform. In 2016, the city implemented a new nationally standardized algorithm tool to predict whether a defendant might reoffend or run if released from jail. The goal was to avoid using bail, which many criticized as penalizing poor people, and limit bias in judges’ decisions about who should be released.

Since 2018, courts have released more defendants to the oversight of the nonprofit San Francisco Pretrial Diversion Project, which has provided services for more than four decades. The nonprofit oversaw about two-thirds of the cases reviewed in the recent report and currently contracts with the Sheriff’s Department.

The organization has shared higher success rates than those in the report, but the numbers aren’t directly comparable because the nonprofit limits its tracking to the time clients are in its programs, which averages 90 to 120 days, and, unlike the UC report, does not track newly committed crimes outside San Francisco.

The recent report analyzed three different measures of success: how many defendants released from jail before their trial showed up for court; how many were not accused of reoffending; and how many didn’t allegedly commit new violent crimes. Counties want high rates of all three.

San Francisco’s rates were “substantially lower than the local and national validated rates,” according to the California Policy Lab study, which was released on July 1. The author pointed out that comparisons to other states are problematic due to differing policies and reporting standards.

The data doesn’t include many defendants charged with low-level crimes who just receive a citation, report author Johanna Lacoe said. In counties outside of California, those defendants might be assessed, skewing data comparisons.

San Francisco mirrored Los Angeles’ statistics, except for the rate of people accused of committing new violent crimes, which was 18% in San Francisco, compared with 9% in Los Angeles.

Separately, the California Policy Lab report evaluated the accuracy of the algorithm used to guide court decisions on releasing defendants. The tool uses factors like a defendant’s age and previous record to calculate a risk score and recommend release or incarceration. Judges followed recommendations in about 70% of cases in the report.

The report concluded the tool was a “fair to good” predictor of flight and repeat offenses. But it found some indications of race and gender bias.

Black and Latino people were more likely to show up for court than white counterparts who had been assigned the same risk of flight, the report found. Women were less likely than men with similar risk scores to commit new crimes.

Socioeconomic factors may create bias in who’s recommended for release, Lacoe said. For instance, Black individuals might have more serious criminal histories as a result of where they live and how they’re treated by police, and that would be reflected in their risk score.

San Francisco public defender Mano Raju criticized the tool, saying that “we need to worry about meeting people’s material needs, not reduce people to risk scores.”

Paul Briley, the regional chapter coordinator for the nonprofit Legal Services for Prisoners with Children, who was previously incarcerated, attributed disparities to overpolicing of the Black community.

“Rather than look at the people as if they are the problem, we should look at the problems that people face,” Briley said.

Stefani’s legislation would require courts to report what happens in cases where judges don’t follow the risk assessment tool, and it would adopt the report’s recommendation to address bias in the tool.

The researchers looked at 9,881 cases of people charged by the District Attorney’s Office and released from jail before trial. Some people were involved in more than one case. Forty-nine percent did not make it to their court appearances. Fifty-five percent were arrested for committing a new crime, the report said.

Among clients in San Francisco Pretrial Diversion Project’s programs in the average year between 2016 and 2019, 63% showed up for court, 16% allegedly committed a new crime and about 5% were accused of committing a new violent crime, California Policy Lab reported.

Given that the nonprofit represented a majority of the cases in the recent report, Stefani questioned the figures. “We and the public have been led to believe they have had extremely high appearance and safety rates, and that is just not true,” she said.

Stefani voted against the nonprofit’s two most recent contracts — $15.9 million for the past four years and $18.8 million for the next three years.

David Mauroff, CEO of the San Francisco Pretrial Diversion Project, said the nationally accredited nonprofit’s data is accurate and its reporting standards were set by a group including the district attorney, the public defender and the probation department. His organization hasn’t included crimes allegedly committed by defendants released before their trial outside of San Francisco in its tallies because it had limited access to that information, he said.

“Our No. 1 priority is to keep people safe and support our clients,” he said.

Rachel Marshall, a spokeswoman for the District Attorney’s Office, said the nonprofit’s clients have had “tremendous success.”

Mauroff acknowledged that San Francisco releases more defendants than many other jurisdictions, raising the chances they’ll reoffend. He stressed that the city has stricter standards for gun-related offenses.

He and others said San Francisco could help people show up for court and not reoffend by connecting them with housing, treatment and support to apply for jobs and schooling.

“Jail is not a place where people are rehabilitated,” Mauroff said.

Mallory Moench is a San Francisco Chronicle staff writer. Email: [email protected] Twitter:@mallorymoench

Please support Catherine Stefani for Supervisor with a donation today!