When lifelong felon Troy McAlister struck and killed two pedestrians, 60-year-old Platt and 27-year-old Hanako Abe, on New Year’s Eve 2020 while fleeing a robbery in a stolen vehicle while high on meth, it set the recall of progressive district attorney Chesa Boudin in motion. Five years and a new district attorney later, the Platt and Abe families are still waiting for justice, and a recent development foreshadows a distinct possibility that justice may never come.
On Oct. 3, after a short hearing before Care Court creator and lead judge Michael Begert, a motion for mental health diversion by the San Francisco Public Defender’s Office in the Troy McAlister case was set for Oct. 28, 2025. Before the hearing, residents, activists, and community leaders assembled on the steps at the Hall of Justice to demand a criminal trial for McAlister.
Five seems to be a significant number in McAlister’s trajectory. After Boudin released him from jail, where he sat for five years awaiting trial on a 2015 burglary, the cavalier criminal was arrested five additional times prior to killing Platt and Abe. With over 100 felonies throughout multiple counties in a criminal career spanning 30 years, McAlister is the poster boy for someone who deserves to spend the rest of his life in prison. Yet, in a statement to the press after the short hearing, Deputy Public Defender Scott Grant, who represents McAlister, said his office wants a “multi-year treatment plan with long-term residential placement and an ankle monitor,” stating that it was “proper and reasonable” to pursue diversion for his client. “Research tells us again and again that intensive, targeted treatment for people’s underlying problems — not incarceration — fosters long-term safety.”
That hasn’t worked in the past for McAlister, who has been through diversion programs in the past. In 2015, then public defender Jeff Adachi wanted to send McAlister to the Salvation Army for drug rehabilitation, but then district attorney George Gascón — known for his progressive stance on criminal justice reform — refused, saying “The defendant has a history of violent felony convictions dating back to 1995,” Gascón, a noted progressive, said, “As a result of the charged offenses, the defendant’s exposure is approximately thirty-five years to life in prison.”
In fact, according to two studies, diversion doesn’t work in general.
In July of 2020, California Policy Lab (CPL) released a entitled “Alternatives to Prosecution: San Francisco’s Collaborative Courts and Pretrial Diversion” to report on the city’s various diversion programs. The study addressed recidivism after diversion, new arrests during diversion, and the success rate of the programs. CPL is a nonprofit funded in part by Arnold Ventures (the foundation of billionaire and former Enron energy trader John Arnold), which also funds San Francisco’s Pretrial Diversion Project (PTD) and supports its goals.
According to two studies, diversion doesn’t work.
Since the reviewing agency and the reviewed programs are both supported by Arnold Ventures, the inherent conflict of interest makes it plausible that CPL might ignore or understate some of the problems they discovered — something to keep in mind when considering their conclusions. The report analyzed diversion outcomes between 2008 and 2018, and the key takeaway was that prior to Chesa Boudin becoming San Francisco’s district attorney in 2020, diversion and the collaborative court programs were already in deep trouble. Even historically more successful PTD programs for low-level offenders were experiencing record failure rates.
Those diverted for felonies had the highest rates of recidivism
According to the study, cases referred to diversion had lower conviction rates than nondiverted cases, but referred individuals had higher rates of “subsequent criminal justice contact.” The lower conviction rates were expected because successful completion of the program results in dismissal of the charges.
What was unexpected is that diverted defendants had higher rates of subsequent criminal justice contact than those who were not diverted at all — the opposite of the intended effect.
As the authors explored potential explanations for the poor results, they noted that San Francisco is unique among most other jurisdictions. A national survey of prosecutor-led diversion found that as of 2017, only half of the programs were accepting individuals charged with felonies, and just one-third accepted individuals with prior felony convictions. Many of San Francisco’s programs serve individuals with multiple felonies and long criminal histories.
While referrals to PTD are similar to Collaborative Court referrals and can be made “at any point during the case process,” most low-level offenders are automatically referred to PTD. When San Francisco’s program started in 1976, only first-time misdemeanor arrestees were eligible. Today, PTD is one of the largest programs but, according to the study, it is also “one of the least intense in terms of program requirements,” with participants often taking just a single educational course or community service assignment.
The study analyzed historical data in programs for felony offenders, including Behavioral Health, Community Support, and Young Adult courts. Table 11 of the study shows a pattern of higher arrest rates for those going through the collaborative courts than those who were not diverted at all. For example, 41 percent who went through behavioral health diversion had at least one new arrest after one year, 42 percent from Community Support Diversion had at least one new arrest, and 38 percent from Young Adult Court had at least one new arrest. These numbers compare unfavorably to the 34 percent figure for those defendants who had not been diverted. Three years after disposition, 52 percent from Behavior Health Court and 46 percent from Community Support Court had at least one new arrest. These numbers also exceed the 45 percent figure for those defendants who were never diverted.
If the purpose of diversion is reform, having higher or identical recidivism rates to those who are not diverted demonstrates that the program isn’t working. The recidivism rates for PTD participants with less serious misdemeanor offenses were better, with 25 percent picking up at least one new arrest after one year compared to 34 percent for those not diverted, and 38 percent picking up at least one new arrest after three years compared to 45 percent for those not diverted. While even low level PTD results were not stellar, at least they were in the right direction. Felony diversions, however, were heading in the wrong direction.