Illustration by Tom White
Solutions_Cops_Clinicians_illo
This article first appeared in the Solutions newsletter. Click here to learn more about the newsletter and sign up.
Here are three calls that officers of the St. Louis Metropolitan Police Department responded to over the past year:
A suicidal woman “locked inside her vehicle with a loaded gun.”
A man “who would cause commotion in stores and would refuse to leave areas—attempting to get into fights…”
An “extremely escalated” woman who was “hearing voices” and “cycling from crying to yelling.”
In all these cases, the officers didn’t respond alone: Each arrived with a “purple shirt”—the nickname given to social workers from Behavioral Health Response, a nonprofit contracted by the city to participate in its crisis response unit, or CRU. (The verbiage quoted above comes from an internal report that the nonprofit recently gave me.)
Variants of the CRU program, which some at City Hall refer to as “cops and clinicians,” exist in other cities such as Eugene, Oregon, Houston, and Denver. The logic is this: Folks in crisis need help more than they need jail time, while police want to divert resources to more serious incidents. Here in St. Louis, the SLMPD is overstretched as it is. It’s budgeted for 1,275 officers but has only about 950, according to the St. Louis Post-Dispatch. The 911 call center is also understaffed.
CRU kicked off in early 2021. After the first year, BHR did an evaluation with help from a data service donated by Mastercard. The evaluators found, among other things, that CRU had saved 2,000 hours of police and EMS-worker time and had averted 750 hospitalizations.
After the second year, I was curious how things had progressed, so I asked for and received the most recent data. It contains signals that things are moving in the right direction, though some caveats are in order.
CULTURE SHIFT
When CRU began, officers were “wary,” recalls Leonard Day, the SLMPD lieutenant who commands it from the police side. They were unsure, he says, whether it was “one of those fly-by-night programs that come and go.” But many have since come to see it as helpful. “The culture has really changed,” Day says, “and officers are really requesting CRU on a lot of calls for service that fit that criteria.”
Meanwhile, adds Eli Horner, associate clinical director at BHR, the police department’s call takers and dispatchers have gotten better at recognizing over the phone which situations could be handled by CRU.
The program data supports these impressions.
Courtesy of Behavioral Health Response
BHR - Top Presenting Problems
By far the most common situation that CRU handles is categorized as crisis intervention (represented in the bar graph above by the acronym “CIT”). If the incident is nonviolent—that is, there’s no weapon, the person isn’t attacking anyone, and has no history of violence—the cop-and-clinician duos can show up alone, without the scene having to be stabilized by district officers. Already this year, CRU has responded to hundreds more of these nonviolent CIT calls than they did in all of last year. What that means, Horner says, is that district officers are increasingly freed up to respond to more serious calls.
You might notice in that bar graph that CRU has responded to far fewer domestic violence calls than it did last year. Horner says this is a healthy sign of adjustment: The unit concluded that the help already being delivered by the SLMPD’s domestic abuse response team made CRU’s presence on those scenes feel duplicative. They’ve therefore spent more time on cases of “intimate partner violence” (which differs from domestic disputes in that the parties don’t live together).
And how do the individuals fare afterward? The most common scenario, according to the data, is that they accept behavioral-health resources. Sometimes this means the purple shirts help the individuals get set up with new services or reconnect to a clinic or therapist they already know; sometimes it means that BHR is able to resolve the matter on-site. From January to early August of this year, “resources offered and accepted” was the outcome in 2,528 cases, compared to 476 cases in which the individuals declined.
LESS HOSPITALIZATION
And about those outcomes: In nearly every case that CRU handles, the person whom the team interacts with avoids jail. But program data also suggests that CRU is getting better at helping people avoid hospitalization.
Courtesy of Behavioral Health Response
BHR - Jail & Hospital Diversions
If you look on the left of this bar graph, you’ll see that when CRU started, in 2021, their hospital diversion was at 77 percent. In plain English, that means that roughly one out of four cases ended with the person being hospitalized. Today, one out of 10 cases ends that way.
“That would be a significant improvement,” says Chris Sullivan, a professor who chairs the Department of Criminology and Criminal Justice at the University of Missouri–St. Louis. But he urges caution in judging the effectiveness of CRU in this specific aspect. The counterfactual isn’t known, he points out; maybe the numbers would’ve looked similar without the existence of CRU. Still, Sullivan says, even randomized controlled trails aren’t “ironclad,” and these numbers “look like a step in the right direction.”
EXPANSION
Wil Pinkney, Jr., who oversees CRU from City Hall as the director of the Office of Violence Prevention, says that the program is scheduled to expand in two ways. First, clinicians will be embedded in the 911 call centers. There, they’ll help decide whether to dispatch CRU to scenes and try to help callers resolve crises over the phone. Second, some of BHR’s purple shirts will start responding to nonviolent CIT calls without any armed police presence at all. (Denver has tried this with its STAR program, and in the first half of 2022, it responded to 2,837 calls for service and never had to call for backup because of a safety issue.)
Pinkney says he hopes to have an empirical evaluation of CRU, possibly conducted by Sullivan and his colleagues at UMSL, within the next couple years. In the meantime, he says, the data contains promising signals, as do the anecdotes that he’s heard. “We obviously need to have data because we live in a world where we want to make sure we’re being good stewards of the city’s money,” Pinkney says. “But at the same time, this work is about more than numbers. We want to make sure we’re telling the story of people and how we’re changing a mindset.”
I’ll relay one story told by BHR. (Granted, the nonprofit is describing its own work and therefore not a neutral observer, but I’ll relay it anyway.) The story is about the woman mentioned above who’d felt suicidal and locked herself in her vehicle with a loaded gun. According to a written summary of the incident, CRU showed up. They learned from the woman’s partner that she’d attempted to pull the trigger but the safety had been on. They found the woman sitting on some steps next to her friend. She was “crying,” “upset,” and “overwhelmed.” CRU stayed for two hours. They helped her to calm down. She shared her personal strengths, interests, triggers, and supports. She also agreed to create a safety plan so that in any subsequent crisis, she’d have a playbook to follow. At the end, the woman reportedly said, “No, I don’t have any questions right now, but I want to tell you that our conversation we had has been the best conversation I have had with someone…” She did not end up in the hospital, or in jail.
