Early in the morning of July 12, 2012, a man had just left the gay Badlands nightclub in San Francisco's Castro district when Antoine Dilworth approached him, the man testified months later in a San Francisco Superior Court hearing.
After they had chatted for a bit, the victim, who was 27 at the time, reluctantly agreed to give Dilworth, 30, a ride. The man, who the Bay Area Reporter isn't naming because he's a crime victim, drove Dilworth to a spot near Beck's Motor Lodge a few blocks away.
As he tried to drop off Dilworth, he testified, Dilworth stabbed him in the neck.
The man said Dilworth had a "mean, scary glare," and he described the terror he'd felt as he struggled to get out of his seatbelt, and Dilworth stabbed him again, this time seriously injuring his thumb.
The victim finally managed to get out of the car. Dilworth drove away but was soon stopped and arrested.
It wasn't the first time Dilworth, who's gay, had been accused of carjacking, and he had a lengthy trail of previous crimes, according to Dilworth, who said in a recent interview that his full legal name is Antoine Joulent Tooks-Dilworth, and court records.
Asked why he couldn't stay out of trouble growing up, Dilworth said, "I didn't have a good support system. I didn't have that foundation I needed back then ... there was nobody I could talk to and nobody I could turn to. I didn't have anything. It was just me."
Like Dilworth, many of the people accused of crimes in San Francisco have been arrested before. Local authorities tout substance abuse and mental health treatment, batterers' intervention, and other programs designed to reduce recidivism, but statistics often aren't available to help determine how effective those efforts are.
Like many officials interviewed for this story, Judge Bruce Chan, the supervising judge of San Francisco Superior Court's criminal division, said just locking people up isn't the answer.
"If you just throw them in jail and do nothing, it doesn't change anything in the long-term," he said. Whatever the approach, he said, "We will never eradicate crime completely. ... We're not magicians here."
Chan said many hotly debated topics in San Francisco are related, from health care to the minimum wage, and "this is the proving ground right here" for those issues.
"We really are the emergency room for all of society's problems," he said.
Chan also said the programs offered here, on which millions of dollars, including money from grants, has been spent, have put San Francisco ahead of other jurisdictions.
Beginning in October 2011, state Assembly Bill 109, which was designed to reduce prison overcrowding, recidivism, and other problems, transferred the responsibility of many people who had been "convicted of lower-level offenses from the California Department of Corrections and Rehabilitation to county jails and probation departments," a January 2014 report presented by the Community Corrections Partnership Executive Committee noted. (The process is known as realignment.)
One challenge has been that there isn't a consensus of what recidivism means, and there's no single rate that everyone points to.
The Board of State and Community Corrections has been holding hearings in different cities on a proposal to define recidivism "as a conviction of a new crime committed within three years of release from custody or committed within three years of placement on supervision for a previous criminal conviction."
The hearings are part of an effort to get consistent data from the state's 58 counties. In 2013, Governor Jerry Brown signed Assembly Bill 1050, which requires the board to develop definitions of recidivism and other terms "so that the state can provide a standard that counties can use to measure the effectiveness of their evidence-based rehabilitative practices," according to a board news release.
'Over and over'
Some of the people who see the effects of recidivism most are police.
San Francisco Police Inspector Len Broberg, the investigating officer in the Dilworth case and a gay man, said, "It gets so frustrating for cops, because we see the same people over and over and over again. Whether it's robberies or whether it's domestic violence ... you keep seeing the same faces."
Public Defender Jeff Adachi, whose attorneys represent some of the city's poorest residents when they're accused of crimes, acknowledged some of the difficulties in dealing with defendants.
"It really is a person-by-person struggle in getting people back on their feet," he said, adding, "the worst thing in the world as a public defender" is seeing someone you've helped "walk through the door again."
However, asked about the frustration people feel when defendants with criminal histories have been given repeated chances and end up being accused of more offenses, Adachi said, "We don't punish people for committing future crimes."
Assortment of programs
The city offers many programs to help people succeed.
One idea officials stress is the importance of addressing crimes by young people in order to keep youth from developing a criminal lifestyle.
Adachi's office has developed the three-year-old Legal Educational Advocacy Program, or LEAP, which works with young clients, their families, and school staffs "to create the linkages and supports necessary for a permanent exit from the juvenile justice system," according to his office.
In the just over two years after it started, 90 clients had successfully completed court probation, and fewer than 13 percent of them had reoffended six to 12 months after exiting the program, according to Adachi's office. The agency doesn't have information on whether those youth reoffended as adults.
Officials also point to San Francisco's collaborative court system as key to helping people start to live crime-free. The system includes Behavioral Health Court, which addresses the needs of defendants with mental illness, including those who also have substance use disorders; Drug Court, a felony court that provides what the court calls "intensive supervision case management" for non-violent offenders with "substantial substance abuse problems;" and the Community Justice Center, which offers defendants in some of the city's poorest neighborhoods the opportunity to access social service referrals for substance abuse and other needs.
Comprehensive data on the courts' outcomes, however, can be hard to find.
Ann Donlan, a Superior Court spokeswoman, wasn't able to say how many people who'd completed the Drug Court had been re-arrested.
According to statistics Donlan provided, in 2011, 68 out of the 510 people, or about 13 percent, the Drug Court served graduated. In 2012, 71 out of 358 people, or approximately 20 percent, graduated.
"The court's collaborative justice programs are designed to treat the highest-risk and highest-need clients," Donlan said in an email. "... Despite our best efforts, not everyone succeeds. But we do strive to help them achieve sobriety and drug-free lives."
She also talked about why data on re-arrests isn't readily available.
"Each collaborative court has a database that we use as a case management system essentially for progress reports," said Donlan. "We do not track re-arrests. Tracking re-arrests is time-consuming and goes beyond our staffing and current resources." Undertaking recidivism studies such as an upcoming study by an outside group for the Community Justice Center involves seeking Department of Justice data, she said. "It is a complex process and costly."
Donlan provided a copy of a study related to Behavioral Health Court that was published online in 2010 by the American Medical Association and concluded, in part, "Mental health courts meet the public safety objectives of lowering post-treatment arrest rates and days of incarceration." The study examined mental health courts in San Francisco and three other counties.
San Francisco Police Chief Greg Suhr praised the work of programs like the Behavioral Health and Drug Courts, calling them "terrific." Like Chan and others, Suhr also said locking people up isn't always the answer.
"I think that we have a very holistic approach, and rightfully so, to restorative justice," he said. "We don't want to jail our way out of something."
However, he said, "there are some people" who offend multiple times and are "criminals by nature. To that end, I don't think that however well-designed and well-intended the program or approach is to try and get those folks to not be who they're determined to be," sometimes those people "have to be incarcerated just to give the neighborhoods they like to do their crimes in a break."
Assessing individual needs
Adult Probation Chief Wendy Still expresses pride in her agency's work. She said there's been a decline of almost 2,000 people in the probation population in recent years to 4,697 people.
Still doesn't believe in a one-size-fits-all approach. She described assessing each individual's risks and needs as being critical to the probation department's efforts. Still points to the collaborative courts and the Community Assessment and Service Center, in which her agency is a partner, as examples of efforts that help keep defendants on track.
The center, which opened in June 2013, provides on-site probation supervision and services including therapy, education, and vocational training.
In 2013, Still said, 85 percent of people on probation in San Francisco successfully completed it that year.
Like other agencies, the probation department can only do limited analysis of program outcomes, but the office is getting a new data system, and Still said her staff would aggregate data for people who complete probation and are then convicted of a new crime going one, two, and three years out.
She said the data would be related to convictions rather than arrests because unless someone's been convicted or pleaded to a charge, "they haven't necessarily participated in a crime."
Many of the people Still's office oversees are probationers who've been convicted in domestic violence cases and take part in state-mandated 52-week batterer intervention programs.
"We're absolutely tracking who's being arrested" on an individual basis, she said, but the current data system doesn't allow for aggregating the data to see how many people who've participated in the 52-week programs have been re-arrested.
Adachi, the public defender, said the domestic violence programs are "well-intended," but "there's no evidence to show" that they're "an effective deterrent to future violence."
Asked about how confident he is in programs such as the batterer interventions when there's little data in areas such as re-arrests, District Attorney George Gascón said, "I'm not at all."
Like others, Gascón has had to grapple with limited data capabilities in his office. He said tradition and funding have been challenges. Money is more often devoted to "things that will lead to higher levels of incarceration," he said, and he's had trouble getting funds for things that aren't necessarily related to locking up people.
This year, his office asked Mayor Ed Lee for about $500,000 to fund several analysts' positions. Among other duties, those people would look at the impact of the DA's staff's work on recidivism and what efforts reduce the likelihood of reoffending.
"All those requests were denied," said Gascón. Lee's spokespeople didn't respond to an emailed request for comment.
One effort Gascón's made to reduce recidivism is the creation of an alternative sentencing planner position. Among other tasks, Gasc�"n said, the planner looks at a defendant's "overall readiness" to "start turning their life around" and works to determine appropriate placement, which may include incarceration.
The San Francisco Sheriff's Department has also been working for years to cut recidivism. Current Sheriff Ross Mirkarimi's office was able to provide more comprehensive statistics than most other agencies.
With the 2003 founding of the Five Keys Charter High School, the department was the first such agency in the country to have a high school built in, Mirkarimi recently pointed out. Last Thursday, dozens of students were recognized at graduation ceremonies held at the Hall of Justice.
From 2011 through the most recent point available in 2014, 230 people graduated Five Keys in San Francisco while incarcerated, according to the Sheriff's Department.
Between 2011 and 2013, the recidivism rate for people who completed the program was 28 percent "with a re-arrest of one year out on a new charge," Mirkarimi's office said in response to a request for statistics.
In a recent interview, Mirkarimi talked about the importance of being able to analyze data.
"For as fore thinking and sophisticated as San Francisco is, we are unsophisticated" when it comes to criminal justice data management "overall in the city," he said, but his department is working to correct that.
Such effort is necessary because the criminal justice system statewide and "particularly in San Francisco is evolving in fascinating ways, and if we want to make sense out of these changes" that "seem to be working, then we need a better metric system to really and honestly evaluate our impact," said Mirkarimi.
There shouldn't be any more of the "throwing spaghetti against the wall" approach to see what works, he said.
While data available on the batterer intervention programs is limited, Mirkarimi has firsthand experience with one of them. In March 2012, he pleaded guilty to a misdemeanor charge of false imprisonment that stemmed from an incident in which he bruised the arm of his wife, Eliana Lopez. Lopez disputed the charges.
As part of his sentence, Mirkarimi had to attend one of the domestic violence programs.
"I found it extremely helpful," he said. "It required me to work hard, and I value the experience."
Beverly Upton, executive director of the San Francisco Domestic Violence Consortium, expressed some support for the interventions.
"Batterer's intervention programs can be successful, but there are many variables" to consider, including whether the person wants to change, Upton said in an email.
She offered another gauge besides re-arrests.
"San Francisco went 44 months without a domestic violence homicide that we know of," she said. "That's unheard of in any major city in the U.S. So we're heading in the right direction. That said, I fully support a study on recidivism."
Asked via email about the importance of having data that show how many defendants who complete substance abuse, mental health, and other programs designed to reduce recidivism are subsequently re-arrested or convicted in connection with new offenses, Dan Lawson, executive director and chief of University Police at the University of San Francisco, responded, "It is not definitive but it is important data that needs to be considered for the obvious reasons. Evidence based science can help us better understand human behavior and programs/treatment that can best influence that behavior."
A third of his life in custody
Dilworth, who earlier this year pleaded guilty to the 2012 carjacking as an aggravated assault charge against him in the case was dismissed, has been in custody since his arrest. In a recent jailhouse interview, he said he was 13 when he ran away from home, started using methamphetamine, and got arrested for the first time, for buying and receiving stolen property. He estimated he's been arrested 26 times, half of those as a juvenile, and spent almost a third of his life in custody.
In Dilworth's previous carjacking case, he eventually pleaded guilty to auto theft in exchange for other charges being dismissed, according to court records, and in March 2006, he was sentenced to more than two years in state prison. Dilworth estimated he'd spent an additional year and a half in prison altogether for violating parole.
Court records say during at least one point in his criminal history, Dilworth was ordered to participate in substance abuse treatment. In the interview, he said he'd gone to rehab, but he hadn't been ready to quit using drugs, and other clients there "were still getting high" on meth and other substances. He left after five days, his probation was revoked, and he was sent back to jail, he said.
Realignment
In a February memo, Still said that San Francisco had served more than 4,400 people under realignment in the last two years. The realignment population represents about 10 percent of the people the probation department is supervising, she said in a recent interview.
Still chairs the Community Corrections Partnership Executive Committee, which in January released the report "Realignment in San Francisco: Two Years in Review."
The January report says, "Direct comparisons to recidivism rates" for the population examined "prior to AB 109 are difficult due to the fundamental differences in the recidivism definitions used."
However, for the period October 2011 through September 2013, the report found ways to make rough comparisons, and said that 51 percent of people under post-release community supervision "and 40 percent of mandatory supervision clients were returned to custody during the first two years of realignment, a drastic reduction from the parole return to custody rate prior to AB 109 of 78 percent."
Post-release community supervision clients are people who were released from state prison October 1, 2011 or later and were serving sentences for "non-serious, non-violent, and non-sex offenses," the report says. Previously, "these individuals would have been released to parole." Mandatory supervision is a court-ordered period in the community under the probation department's supervision.
Broberg, the longtime police inspector, said it's important to get people linked to housing, education, and other assistance, but he questioned anyone who would say recidivism is down.
"They still commit [the crimes]," he said, "they're just not getting caught, that's all."
Suhr, the city's police chief, offered one reason why, and referred to a sharp decline in the city's jail population. As of June 11, there were 1,254 people incarcerated in the county jail, and 1,431 people serving in the sheriff's department alternatives to incarceration or alternatives to sentencing, according to the sheriff's office.
"The jail population is down, and arrests are down, but part of this is because we're down 300 police officers too," said Suhr. He said the force would be back to the city charter-mandated 1,971 officers by 2018.
Broberg said that recidivism rates look lower under AB 109 "because they're not violating them, so they're not sending them back."
Asked what she would say to people who think her department is being soft on the people it supervises, Still said, "We're not. We're addressing the behavior. We have a policy of swift and certain sanctions," and when there's a violation "we address it immediately. ... You can only change behavior if you have a balance of rewards and sanctions."
Still pointed to "The impact of probation and parole violations on arrests in four California cities," a 2013 report prepared by the Council of State Governments Justice Center.
The report showed that of the 81,312 total adult arrests in San Francisco during the period examined, 72,004 of those, or about 89 percent, involved people who weren't active parolees or probationers.
"People said it's all these probationers being arrested," said Still, but "it's not."
Asked whether he's seen less recidivism or more, Chan, the superior court judge, said, "It's dangerous to try to declare victory or defeat. When you're here you don't get the bird's-eye view." In his position, he said, "all you see are people who are failing ... but I'm not at their drug court graduations."
He added, "If you're working inside a busy kitchen, it's always going to seem hot to you. ... I don't know if I'm the right person to ask."
Five more years
Dilworth was expected to be sentenced to five years in prison Tuesday, June 24. He'll get a strike on his record, which would enhance punishment for future cases.
His account of what happened that night differed greatly from the victim's. Among other things, Dilworth, who now suggests he's innocent, said he didn't stab the victim and assumed that the victim had accidentally cut himself on the knife, which he said had been lying open on the car's console.
In custody Dilworth has received anger management and other services. He's been working on getting his high school diploma and expected to graduate last Thursday.
Dilworth, who said he's experienced physical and verbal abuse in custody because he's gay, wants to continue receiving services outside of custody rather than go to prison.
He said prison would "harden me, and I don't want that. I don't want people to continue to look at me like I'm an animal."
Asked about Dilworth's desire not to go to back to prison, Gasc�"n, the district attorney, said carjacking and stabbing are "pretty violent," and he noted Dilworth's criminal history.
"Five years is not an unreasonable sentence given the totality of the circumstances," said Gascón. He added Dilworth "has not shown us at this point that he's not a dangerous person."
"He is dangerous, and he needs some time to reflect on his behavior," Gascón said. He said he's hopeful that Dilworth can be reintegrated into society "at a later date."
Dilworth's victim couldn't be reached for comment.