This is part 3 of a multipart series. Part 1 Part 2
Data from the California Policy Lab’s July 2020 study, which showed that participants in diversion programs continued to offend, must have surprised and disappointed its authors and supporting institutions because, in June 2021, without any additional historical data, CPL repurposed data from the 2020 study and released a working paper entitled “The Impact of Felony Diversion in San Francisco.”
Unlike the original study, which used actual data and a straightforward methodology, the 2021 paper was like an Olive Garden never-ending word salad, stumping even some experts. Our next step was asking a data scientist to review both studies. His conclusion was the authors resorted to using a “non-standard 2SLS methodology without justification” to explain away the June 2020 report, but “could not explain away all the problems identified in that original study.”
For example, on page 3 of the 2021 paper, the authors admit: “Regarding measures of future criminal involvement, simple comparisons of means as well as mean differences that regression adjust for observed personal and case characteristics generally show that those who are diverted are more likely to be arrested in the future, more likely to be arrested in the future for felonies, and more likely to experience new convictions.”
At page 25: “We see even larger estimated effects of diversion on cumulative arrests through the second year, with the diverted seven percentage points more likely to be arrested and eight percentage points more likely to experience a felony arrest relative to those who are not diverted.”
“Those who are diverted are more likely to be arrested in the future, more likely to be arrested in the future for felonies, and more likely to experience new convictions.”
In a May 5, 2022 article SFGate reporter Eric Ting cites the 2021 working paper in support of Chesa Boudin’s unprecedented expansion of diversion programs. “Boudin’s office argues that diversion leads to lower rates of recidivism than conviction and incarceration. A California Policy Lab report from 2021 found that individuals going through diversion were 10 percent less likely to be arrested for drug possession and 28 percent less likely to be arrested for drug sales after one year than were individuals convicted and incarcerated,” Ting writes. It appears Boudin cherry-picked the most impressive sounding data for Ting, but both apparently ignored the rest of the study. The chart Ting drew his numbers from shows individuals diverted were 5 percent more likely to be arrested for crimes against a person after year one and 19 percent more likely to be arrested for crimes against a person after year two than individuals convicted and incarcerated. (“Crimes against a person” include assaults, robberies, and other violent offenses.) But more important, these figures from the 2021 working paper are only “estimates” using a nonstandard methodology, and these estimates do not match up to the actual results reported in the 2020 study, which cites failures across the board for felony diversion programs through the same study period (see Table 11 above). The study also does not explain the high numbers of new arrests while defendants were in the programs.
Despite these findings from the 2021 paper, Boudin dramatically increased diversion referrals for crimes against a person: assault cases diverted by Boudin increased by 338 percent versus his predecessors, and robbery cases diverted by Boudin increased 244 percent versus his predecessors. If Boudin was relying on the 2021 paper, which shows diverted participants had a 19 percent increase in arrests after two years, he shouldn’t have been sending violent crimes to diversion at all. And the failure rates in both studies should have indicated the need to pause and fix the problems, not exponentially increase referrals.
‘If it was up to me, I’d kill everything that stands.’
There is no better example of failed diversion than the Patrick Thompson story. Thompson had been charged with three separate cases in 2017: misdemeanor contempt of court order, felony assault with a deadly weapon other than a firearm and battery, and felony battery with serious bodily injury. A judge declared him incompetent, and he was transferred to Napa State Hospital for the mentally ill. After his release, he was accepted into San Francisco’s Mental Health Diversion program, which the San Francisco Public Defender’s Office said he “successfully completed” in August 2020. A court transcript shows that the DA’s office under Boudin didn’t object to Thompson’s release, despite the fact he had been arrested five months prior — while still in the diversion program — for threatening a female police officer with an axe.
According to a police incident report, on Tuesday, April 14, 2020, at 4:28 a.m., two uniformed officers, one male and one female, were entering a marked patrol SUV parked in an alley off of Eddy Street after completing a call for service. “I heard a male voice yell something unintelligible; I looked up and saw a large male walking toward me at a brisk pace,” the female officer wrote. “I was momentarily blinded by a security light on the building and then saw the male closing the distance between us while holding a large axe in his right hand.”
She drew her gun, ordering the man to stop walking and drop the axe; he slowed but initially did not comply. The male officer drew his gun and ordered the suspect to drop the axe, upon which he did. The officers then ordered the man to move away from the axe, but he refused, shouting, “Shoot me, you motherfucker!”
Two additional officers arrived on scene, one carrying an extended range impact weapon, or ERIW, that shoots bean bag rounds. When the officer racked the ERIW the noise startled the suspect, at which point he put his hands on his head. As another officer seized the axe from the sidewalk the suspect turned to him and said, “If it was up to me, I’d kill everything that stands.” He then disclosed he was “a mental patient.”
Reached by phone, one of the officers on scene that morning in 2020 said, “10 years ago we would have booked this guy on a 148 for felony threats. Now we think, ‘Will the district attorney and the public defender even try this guy?’”
On May 4, 2021, after he was deemed a “successful” diversion graduate, Thompson randomly stabbed two women, ages 63 and an 84, multiple times.
San Francisco police officers responded to an incident at 4th and Stockton Streets just before 5 p.m. where two victims were suffering from multiple stab wounds. Both were transported to the hospital, the 84-year-old with life-threatening injuries. Meanwhile, police were alerted to a photo taken by a witness, and they instantly recognized the man from prior incidents. Just two hours later they arrested Thompson in the 600 block of Eddy Street. He was booked on two charges of attempted murder and elder abuse.
Later that evening, I received a photo anonymously. “This is an X-ray of the knife still in the 84-year-old,” the message read. “He was so aggressive the handle broke off, and it punctured her lung.”
In part 4: Former San Francisco District Attorney Chesa Boudin sent hundreds of violent offenders to diversion and never tracked a single one.
THE STUDIES:
July 2020, California Policy Lab, Alternatives to Prosecution: San Francisco’s Collaborative Courts and Pretrial Diversion
June 2021, California Policy Lab, The Impact of Felony Diversion in San Francisco