Adding to the warring choruses about the state of San Francisco and crime, Supervisor Catherine Stefani appears to have taken her case to the Chronicle, which rolled with a fairly sensational headline about how half of all suspects granted pretrial release go on to reoffend.

The Chronicle's piece today, headlined "Half of people released from jail in S.F. before trial allegedly reoffended. Why it's complicated", draws its data primarily from a California Policy Lab report that it does not provide a link to. It appears to be this report, which uses San Francisco as a study case to validate the nationally used Public Safety Assessment (PSA) tool, which guides pretrial release decisions based on a risk-assessment algorithm.* But Stefani appears to want to highlight differences in the data that the California Policy Lab (CPL) — a research institute based out of both UC Berkeley and UCLA — found regarding felony suspects versus data that's reported by the nonprofit San Francisco Pretrial Diversion Project, which has been providing services to the Sheriff's Department for four decades.

Also, even though this conversation seems to point back to arguments about our progressive DA Chesa Boudin and his desire to incarcerate fewer people, the data we're talking about all comes from before his tenure, from May 2016 to December 2019.

Here's the point that it seems like Stefani wants to make via the Chronicle: The CPL finds around half of suspects granted pretrial release during that 3.5-year period went on to reoffend, and around half also failed to show up in court. Also, around 1 in 6 went on to commit a new violent offense, the data suggests. This is a less shiny picture that the Pretrial Diversion Project's own data has shown, which can be explained by a number of factors the Chronicle goes over — for example, the CPL's data includes offenses that were committed outside of San Francisco, and outside of the Pretrial Diversion Project's 90- to 120-day purview while clients are in their programs.

The Pretrial Diversion Project says that this other data is often incomplete or hard to access, and therefore it limits its data to offenses in San Francisco and within that three- to four-month period.

To Stefani, this isn't enough. "Nobody can look at this report and say we’re doing great,” she tells the Chronicle. "It validates the experience that people in San Francisco are feeling when they’re concerned about crime."

She adds, "We and the public have been led to believe they have had extremely high appearance and safety rates, and that is just not true."

It should be noted that Stefani doesn't seem to be a fan of pretrial diversion, or of this nonprofit — she's voted against approval of their two most recent contracts with the city, which now amount to around $6.3 million per year.

The District Attorney's Office says that the Pretrial Diversion Project has had "tremendous success." And David Mauroff, CEO of the nonprofit, tells the paper, "Our No. 1 priority is to keep people safe and support our clients."

Two recent examples of crimes that are likely driving the politics here are the suspect in the early May stabbing of two elderly Asian women on Market Street, Patrick Thompson, who last year had successfully completed SF's mental health diversion program following a 2017 stabbing arrest; and Daniel Cauich, the suspect in the mid-June stabbing of a 94-year-old Asian woman in the Tenderloin, who had just been released from jail for a string of robberies last fall, and who was granted pretrial release after a subsequent May arrest. DA Boudin had requested Cauich remain detained, but the judge granted release anyway, with an ankle monitor and a mandate to seek substance abuse treatment.

Cauich also spent three years in jail on a murder charge which was dismissed on a technicality in early 2019.

Related: Breed Says '10 Groups' Responsible for 1,000 Burglaries and Car Break-Ins Each Month, and Arrests Have Been Made

*This article has been corrected to link to the correct CPL report being referenced.