CAHOOTS in Denver: Experimenting with Police Alternatives

A particular problem for the police is that they are often the first to respond to all types of emergency calls, not just those related to an obvious crime. In some cases, there might be a more efficient alternative?

Newspaper in autumn 2021 Journal of Economic PerspectivesMonica Bell argues that academic research on policing has focused heavily on the number of police officers and police deployment methods, but much less on alternatives to policing or alternative ways to reduce crime.“Next Generation Police Research: Three Proposals”. 35:4, 29-48). Here is one of the examples she gives:

But there are a number of other community programs or alternatives to traditional policing that remain largely unexplored, even as some of them become models for other jurisdictions around the country. For example, CAHOOTS (Outdoor Crisis Help) started in Eugene, Oregon to send two-person clinical response teams to help people in mental health crises without relying on armed police officers. Although the program has been in existence for more than three decades, in the summer of 2020 it captured nationwide attention and became the model for numerous pilot programs in San Francisco, Denver, Rochester, Toronto and other countries. The Eugene CAHOOTS program is funded and overseen by the police department, but some other newer programs are funded and administered separately from the police. Despite its long duration—even longer than the violence prevention programs mentioned above—CAHOOTS has never been rigorously evaluated. Among other things, there is intense debate about how to measure the leakage rate (Gerety 2020). There is a lack of information and modeling of a police-free crisis response, although it is hoped that this will change as more cities use these approaches.

Well, now there is at least one proof of a CAHOOTS-style program. Thomas Dee and James Pine write about Denver’s experience in “Community Response to Mental Health Crises and Substance Abuse Reduced Crime”. (Scientific Advances, June 8, 2022 8:23 AM). Here is some background (citations omitted):

The Support Team Assistance Response (STAR) program in Denver provides a mobile crisis response for community members experiencing issues related to mental health, depression, poverty, homelessness, and/or substance abuse. The STAR response consists of two medical professionals (i.e. a psychiatrist and a paramedic in a specially equipped van) who provide rapid on-site assistance to individuals in crisis and refer them for further appropriate treatment, including requesting police involvement if necessary . The STAR program design is based on the Crisis Assistance Helping Out The Streets program developed in Eugene, Oregon.

STAR went live on June 1st, 2020 for a set 6-month pilot period. During this period, STAR limited its operations to selected 911 emergency calls in eight specially selected police stations (i.e., out of 36 city stations) where the need for STAR services was expected to be greatest. … [A]All but one of the areas in the STAR pilot service area are also designated by the City as “displacement-vulnerable” areas, rapidly gentrifying urban spaces that are displacing poor and otherwise at-risk residents. …

Emergency call operators directed STAR personnel to the appropriate scenes of incidents that took place at designated police stations and during program hours (Monday to Friday, 10:00 am to 6:00 pm). The identification of emergency calls eligible for STAR services was based on two specific selection criteria. First, the incident had to be identified by at least one of several codes: calls for help, intoxication, suicide streak, welfare checks, obscene exposure, unwanted intrusion, and disposal of syringes. Secondly, no evidence was required to send the STAR van that the incident involved serious criminal activity such as weapons, threats or violence, or serious medical needs. The STAR team also responded to calls from uniformed police to engage with members of the community in crisis and began to work on the ground themselves. During the 6-month pilot period, the STAR team responded to 748 incidents, or almost 6 incidents in an 8-hour shift. Approximately a third of the calls to STAR were made at the request of arriving police, while the remainder involved direct 911 dispatches or STAR teams responding to field surveillance on their own—none of which required the police to be called in for help or to call. response to a crime.

Since the program was deployed only in certain areas and only at certain times and days of the week, it becomes possible to compare trends and patterns, considering both patterns within the sites and in comparison with adjacent territories. (Those who want details on how these “difference-in-difference” comparisons are made can view the article itself.) In particular, the authors looked at some low-level crimes that appeared to be correlated with the types of situations to which she reacted. STAR team, such as disorderly conduct, trespassing, alcohol and drug use. Again, the STAR team was not sent for major crimes or medical emergencies.

The authors note that “the service reduced the number of STAR-related offenses in the treated areas by 34% during the 6 months of the pilot phase. … This impact assessment assumes that the STAR pilot program prevented almost 1,400 criminal offenses at eight participating sites and in 6 months of operation… This program-induced reduction in measured offenses is broadly in line with the scale of STAR operations. Specifically, the STAR team answered 748 calls during our research period. At baseline (i.e., pre-treatment period), each STAR-related incident resulted in an average of 1.4 reported offenses at treated sites. This suggests that we should expect 748 STAR field visits to result directly in a reduction in reported offenses of just over 1,000 (i.e., 748 × 1.4 = 1,047).”

It is interesting to note that some decrease in the number of criminal offenses occurred outside of the 10-6 time frame when the STAR program was actually running, which is consistent with the idea that STAR interventions not only addressed immediate problems but improved the overall situation. Moreover, this approach seems to be potentially cost-effective:

The total cost of the 6-month STAR pilot program was $208,141 (fifty). One useful way to frame these public costs is to note that the corresponding reduction of 1,376 offenses implies a cost of $151 per offense reduced. To put this in perspective, the available estimates (eight) suggests that direct criminal justice costs for a minor criminal offense (such as incarceration and prosecution) average $646 (in 2021 dollars). In other words, the direct cost of having the police provide first aid to people in mental health and substance abuse crises is more than four times higher than the cost associated with a community-based response model. A more complete calculation of the costs and benefits associated with community response models will also include the costs and benefits associated with any medical services provided through emergency response services. For example, police officers may be more likely than community first responders to refer people in crisis to a relatively expensive emergency room or receive no care at all. However, the results presented here indicate that community response models deserve careful consideration as a highly cost-effective way to reduce police interactions with non-violent people in a crisis and instead provide adequate medical care.

Of course, the study has some obvious limitations. For example, don’t extrapolate a program that runs from 10 to 6 on weekdays to something that might be happening late on a Friday or Saturday night. Also, the attention to the number of reported low-level crimes is interesting, but certainly incomplete. Yes, it’s good that people don’t accumulate a list of petty crimes. But was the problem that caused the STAR command to be resolved in a way that would help the people directly involved in it? Did other affected members of the community feel that the problem was solved?

With these and other caveats in mind, the partial results are clearly encouraging. As the authors write:

However, the results presented here indicate that community response models deserve careful consideration as a highly cost-effective way to reduce police interactions with non-violent people in a crisis and instead provide adequate medical care.