Mayor Eric Adams’ senior advisor on the topic has some thoughts on the incoming mayor’s plans for a new Department of Community Safety.
In his plan to heal a public mental health system ravaged by decades of neglect, New York City Mayor-elect Zohran Mamdani seeks worthy and achievable goals: shifting more of the burden of crisis response from police to mental health professionals, streamlining a hopelessly complex and often redundant array of community-based services, and removing the barriers that everyday New Yorkers face in accessing basic mental health care.
It is immaterial whether these are precisely the goals I would prioritize in an alternate universe where Mayor Eric Adams won a second term. The voters chose a different vision, and we should all hope it succeeds.
For that to happen, the mayor-elect will need to demonstrate adaptability. Informed by what I have learned over three-plus years in City Hall and two decades in mental health advocacy, I believe the shiny new vehicle Mamdani has designed to deliver on his ambitions — a $1.1 billion city agency to be known as the Department of Community Safety (DCS) — should not make it out of prototype.
In support of creating DCS, the Mamdani platform promises to revolutionize crisis response and fill the gaps in community-based care, rightly bemoaning the “patchwork of programs” presently targeted to the city’s severely mentally ill.
But as details emerge, it becomes clear that DCS as envisioned would add even more illogic to our maze of community-based services, pull us off the path to progress on crisis response and re-open a pernicious service gap we have only just begun to fill.
Piling on the patchwork
From press coverage, one might have the impression that DCS would be laser-focused on crisis response — shouldering the burden of answering mental health-related 911 calls that now falls mainly on the city’s police department. While that is certainly part of what Mamdani has in mind for the new agency (and more on that below), it represents just one of many major assignments to be stuffed into DCS’ portfolio.
DCS would also become the central command for efforts to conduct outreach to people who appear to be unsheltered and/or under the influence of substances in public areas. Unlike crisis response, outreach work is largely about building rapport and trust with familiar faces over time and learning as much as possible about what led each person to the street in hopes of eventually persuading them to accept shelter and services. These efforts are currently overseen primarily by the city’s Department of Homeless Services (with the Department of Health and Mental Hygiene pitching in on substance use outreach), mostly through contracts with community-based nonprofits.
DCS’ authority would not end there. The Mamdani platform positions the new agency at the forefront of “preventative and ongoing care” for mental illness of all kinds, with oversight over a massive new “Community Mental Health Navigator” program, an expanded network of peer-run “clubhouses,” an initiative to establish new “crisis residences,” various mental health services for youth and a plan to scale up important outpatient support programs like Intensive Mobile Treatment (IMT) and Assertive Community Treatment (ACT).
Regardless of the merits of each of these proposed new programs and expansions of existing ones, there is no question that they currently fall within the purview of the city’s Department of Health and Mental Hygiene (DOHMH), which is tasked under state law with developing and managing the city’s “local services plan.”
Far from streamlining all these functions and services, adding DCS to the mix would make the patchwork exponentially patchier.
So long as we have a Department of Homeless Services, it remains the only logical agency to oversee street outreach. Many clients shuttle frequently between the shelters and safe havens overseen by DHS and the street. It makes no sense for the tracking of their cases to be passed back and forth constantly between two agencies.
Similarly, unless the mayor-elect intends to pull oversight of all mental health services away from DOHMH (essentially, slicing it down to a mere “DOH”), it would add unprecedented layers of complexity and probably flout state law to splinter accountability for the city’s local services plan, leaving some major mental health initiatives at DOHMH with others, such as clubhouses and crisis residences, overseen elsewhere. The interconnectedness of these programs runs deep.
Answering crisis calls without police
One program that does need an organizational shakeup is B-HEARD, created by former Mayor Bill de Blasio in 2021 to provide a fully clinical alternative to police dispatch for lower-acuity 911 mental health crisis calls. But a DCS takeover of B-HEARD would achieve nothing, except to derail a much more sensible reorganization already underway.
Here is what Mamdani gets right: He is correct to argue that sending cops to crisis calls presenting little or no public safety risk needlessly diverts them from fighting crime, and that many such calls would have better outcomes if we sent caregivers alone. He has good cause to be frustrated that after nearly five years, B-HEARD has not come close to fulfilling its promise, failing both to expand citywide and to respond to all appropriate calls within its current footprint.
The question is how best to set B-HEARD on a course to succeed and grow. To date, the program has operated as a partnership between the City’s public hospital network, Health + Hospitals (H+H), which provides mental health counselors and clinicians, and the Fire Department’s EMS Division, which supplies emergency medical personnel and ambulances. Sometimes it proves to be a very good thing that the B-HEARD response includes EMS, but more often than not, the person they respond to does not need transport to a hospital or emergency medical services.
In light of a severe workforce shortage and other challenges to EMS’ capacity to respond to the full range of medical emergencies occurring every minute across the city, this partnership was destined from the outset to be uneasy. It is unreasonable to ask an EMS system already stretched thin and working under the kind of pressure few of us can imagine to contribute even more of its limited resources to support B-HEARD expansion. A better idea is to let EMS get back to its critical core mission and consolidate B-HEARD under one agency equipped to supply everything it needs.
That agency is H+H, which already operates its own diverse vehicle fleet, including ambulettes, and employs medical professionals of all kinds. Indeed, earlier this month, Mayor Adams and H+H CEO Mitchell Katz announced plans to give H+H full responsibility for B-HEARD by next spring. This should be the breakthrough the program needs to start expanding its reach.
The incoming administration would be ill-advised to scuttle those plans out of blind determination to make B-HEARD the centerpiece program of its new agency. Unless DCS were to establish its own ambulance fleet and emergency medical staff, potentially adding years of delay to its launch with additional political and regulatory hurdles, B-HEARD under DCS would be just as reliant on EMS as it is now. The Mamdani platform’s pledge to expand B-HEARD “so that every neighborhood has a team and the 20 neighborhoods with the greatest need have two to three” is utter fantasy with a structure based on keeping EMS in a forced marriage.
I must add a word of caution on B-HEARD: There are significant limits to how much of the total mental health crisis call volume can be handled without police. Before a 911 call can be diverted to B-HEARD, the dispatcher must determine that the person of concern presents no imminent risk of harm to self or others, is not clearly in urgent need of hospital transport, is not immediately suicidal or violent, does not have a weapon and does not require tactical expertise or resources (e.g., a person standing in traffic). These are all situations that require police to ensure safety, and they encompass at least half of all crisis calls. There is a world of difference between striving to maximize the use of a non-police response within the bounds of public safety and believing that police can be mostly erased from the picture.
The mayor-elect’s platform gives reason to worry that he sits in the latter camp. As an example of what is possible in New York City, the platform cites the highly misleading example of Eugene, Oregon’s now-shuttered CAHOOTS program, a forerunner of B-HEARD. It notes that “CAHOOTS teams have resolved almost 20% of all calls coming through Eugene’s police department,” with 99% of calls resolved without a need for police backup.
Reading that casually, one might have the impression that Eugene diverted 20% of its 911 calls to CAHOOTS, which would be astonishing, since the great majority of 911 calls in any city have nothing to do with a mental health crisis. The implication is that nearly all mental health emergencies had been diverted successfully.
But there is a key piece of information omitted from this widely reported claim. In Eugene at the time of this finding, “calls coming through the police department” did not only mean 911 calls. The Eugene Police Department had established a separate hotline for CAHOOTS and encouraged citizens to use it for non-emergency concerns about their own or another’s mental health. This meant that the “calls coming through police” included a great many calls that no responsible citizen would ever have made to 911 in the absence of CAHOOTS. When I asked a Eugene PD data analyst about this a few years ago, he reported that the rate of 911 calls diverted to CAHOOTS, using diversion criteria essentially identical to B-HEARD’s, was in the range of 3% to 8% each year.
The vital role of “co-response”
We must also beware here of a false choice between police-only and police-free crisis response. There is no question that even many crisis calls presenting a risk of violence would benefit from the inclusion of a clinician and/or peer counselor who is skilled in the art of calming a person down and gently steering them into treatment. There are many situations where police can secure a scene and safely pass the baton to a clinical partner to engage with the person in crisis. This model is known as “co-response,” and many experts consider it the gold standard in responding to situations too dangerous to handle without police.
We do not practice co-response nearly enough in New York City. And yet we may be on the verge of shutting it down entirely. The single most disturbing sentence in the Mamdani mental health platform is this bullet in a summary of DCS’s goals:
“End co-response teams to hand calls off to crisis and outreach teams that are better positioned to address people’s needs.”
Let’s consider what it really means to “end co-response.” The NYPD will (as it must) always insist on sending police to a scene where a person is brandishing a weapon or screaming aggressively at passersby. If we “end co-response,” we foreclose the possibility that police will have the benefit of clinical assistance in these situations.
To keep this in perspective, the NYPD (lamentably) does not currently dispatch co-response teams in response to 911 calls. The department’s small “Co-response Unit” is deployed only in response to certain calls received through 988 or 311 requesting a wellness check for a housed person. These calls are always answered by nurses dispatched by DOHMH. Co-response Unit officers will join the nurse if a record search on the person reveals a history of violence. Is this really what the mayor-elect hopes to end? I imagine the DOHMH nurses dispatched to these calls (and their union) might have thoughts on that.
In the context of outreach — the practice of responding not to calls for help, but to unsheltered folks encountered in the course of canvassing public spaces — New York City under Mayor Adams has made major commitments to co-response. The state-funded SCOUT program, launched as a pilot in the Fall of 2023, teams DHS nurses with officers from the MTA’s internal police department. The subway system is also covered by the city’s own PATH program, launched in the Summer of 2024, teaming DHS’ nurses and outreach specialists with NYPD officers.
Together, SCOUT and PATH typically deploy 17 co-response teams in the subways over a 24-hour cycle. Their mission is to patrol stations and trains in search of people who appear to be unsheltered, engage those they encounter in conversation, assess their needs and endeavor to connect each person with the appropriate level of care. In most cases, it will ultimately be up to the person to accept what the team offers (usually a bed in one of DHS’ low-barrier “safe havens” or a trip to the hospital to address a purely medical need).
Sometimes, the SCOUT or PATH nurse comes across a person who exhibits symptoms of mental illness, presents a danger to themselves or others, and is unwilling to accept voluntary transport for a full psychiatric evaluation. When this happens, the nurses have authority under state law to direct that the person be involuntarily transported by police and EMS to a hospital, which could lead to an involuntary admission for inpatient care if hospital doctors concur with the nurse’s assessment. Under city policy, these individuals are always taken to hospitals operated by H+H and closely tracked to ensure quality of care and adequate discharge planning. If the person is not already connected to an intensive outpatient treatment team, every effort is made to enroll them with one.
These co-response teams do extraordinary work. They find people in the most appalling conditions and give them a fighting chance to gain a foothold on treatment, housing and stability. For a myriad of frustrating reasons, it doesn’t usually turn out like a Hollywood redemption story, but the benefits of whatever medical and psychiatric treatment people receive are surely better than the old policy of benign neglect under the excuse of respecting their “rights” to self-direction. Over two years, the SCOUT and PATH teams have brought nearly 800 people for involuntary hospital evaluation, resulting in over 10,000 nights of hospital care.
It bears repeating here that the troubling bullet point in the mayor-elect’s platform resolves to “end co-response” not just in the crisis call context, but in outreach as well. This suggests that the subway-roaming teams of SCOUT and PATH are directly in the crosshairs. (As a state-funded program using a state-run police agency, SCOUT may be in less peril than PATH. But if the next administration truly wants no part of co-response, it could withdraw DHS and H+H from the collaboration.)
Before doing anything so rash, the mayor-elect should spend some time on a shift with each of these two programs. He should see for himself their compassion and professionalism, take note of how bonded the clinical and law-enforcement members of the teams become through daily partnership, and hear directly from the nurses and outreach workers (most of whom have prior experience with police-free outreach) why officers are so essential to the work they do. It boils down to two irreplaceable police functions.
First, police accompaniment enables the DHS staff to engage with every person in apparent need with a sense of safety, ensuring that no one gets passed by. Without police, it can be scary to start a conversation with someone who appears unstable and unpredictable. Not infrequently, the person whom the outreach worker would reluctantly allow to stumble past them in Penn Station at 3 a.m. is the person most in need of psychiatric assessment.
The second function is logistical. A nurse with mobile crisis certification has authority under state law to direct an involuntary removal, but obviously, the nurse cannot be expected to keep the person on scene until an ambulance arrives, nor force the person onto the ambulance and into the hospital. For the nurse’s authority to have any practical meaning, there must be a police officer on hand at the time the decision is made. Co-response not only ensures that; it ensures that the necessary police work will be handled by officers who have done this many times before and have developed expertise in completing the task with as little physical force and distress to the individual as possible.
Involuntary removal as a form of care
Perhaps the effect of making involuntary removal much more difficult to accomplish is a feature, not a bug, of the mayor-elect’s plan to “end co-response.” Understandably, involuntary removal makes many progressives uncomfortable. In the wrong hands, there is a potential for abuse. And even a perfectly appropriate involuntary removal can be upsetting to witness if the person becomes agitated, as often happens. Progressives are not wrong to prefer that people be brought into treatment voluntarily. Everyone prefers that, and studies confirm that psychiatric treatment is most effective when there is patient buy-in.
Here’s the problem: For many with untreated severe mental illness, the prospects for coming around to accept voluntary treatment are slim to none. The choice isn’t between voluntary and involuntary treatment, but between involuntary treatment and no treatment at all. Their cruel disease renders them too disconnected from reality to understand their own plight. This is not “denial.” As a corollary of their illness, many suffer from anosognosia, a neurological deficit that makes it impossible for them to recognize that they have an illness.
Once anosognosia is understood, the question of individual rights is flipped on its head. Is a person unable to perceive reality truly exercising “choice” when they refuse treatment? Are we really respecting that person’s dignity when we leave them to suffer the horrific consequences of that refusal? Or are we letting ourselves off the hook for failing to do the hard stuff it takes to ensure sustained care?
At a recent Vital City forum, then-candidate Mamdani expressed a general skepticism of involuntary removal, conceding a willingness to use it only as “a last resort.” While I doubt he had fully thought through the implications of that framing, it sounded to me like a retreat to the failed policy of past administrations.
Among the ways we have tried under Mayor Adams to repair the culture of the city’s mental health system is by making a sustained effort to broaden the understanding of what it means to be “dangerous to self” within the meaning of the state’s involuntary removal and hospitalization laws. For decades, a pervasive myth had taken root among community-based providers, police and hospital doctors that a person had to appear imminently dangerous to meet this standard, despite the absence of any such language in the relevant statutes and clear New York case law to the contrary. This led to a wrong-headed consensus that finding danger-to-self required that the person be suicidal, or engaging in outrageously dangerous conduct like running into traffic, or neglecting a medical need that presented a true emergency.
In revised protocols and training materials, we have stressed that “danger-to-self” also applies to those whose mental illness has led to extreme self-neglect creating a risk of serious physical harm in the future, even if such harm is not yet imminent. (Mayor Adams also lobbied Albany over three years for a change in state law to make this explicit, which was achieved in 2025 thanks to Governor Hochul getting behind it.)
This fuller understanding of danger-to-self has been baked into the culture of the SCOUT and PATH programs. In the training of the nurses serving on these teams, we have emphasized the need to look for signs of extreme self-neglect in assessing whether a person displaying mental illness presents a danger, and the nurses have taken it to heart. In most of the removals SCOUT and PATH have conducted, extreme self-neglect, rather than violence, suicidality or outrageous conduct, has been the basis for finding danger. Some of these individuals, who would have been left in the subway if SCOUT and PATH were sticking with the old narrow interpretation of the legal standard, are today in or on track for permanent supportive housing.
And that is why I am so troubled by the mayor-elect setting the boundary at the point of “last resort.” It justifies the old practice of declining a removal if the risk of harm is not imminent, because tomorrow just might be the day this person in acute psychosis agrees to voluntary care. That mindset, however well-intentioned, has been calamitous for New Yorkers with severe mental illness. It has led to needlessly prolonged periods of untreated psychosis, a condition which is toxic to the brain and lowers prospects for recovery. It rests on hope that a clinician and police officers will be on scene at the moment just before disaster strikes. In the real world, that rarely happens. Much more often, the person or someone else is harmed, and/or the person commits a criminal act that lands them in the clutches of the criminal justice system, and we can all agree on the danger they presented in useless retrospect.
It is my fervent hope that Mayor-elect Mamdani will not take us back to the days of involuntary removal only as a “last resort.” I might even dare to hope that he will continue the work in progress to make the system more proactive in rescuing those in obvious psychiatric decline. I would argue that a progressive outlook requires nothing less.
No doubt the Mamdani team has some serious thinking ahead about how to translate their full slate of exuberant campaign pledges into viable policy. That should include a hard look at whether establishing a Department of Community Safety is truly essential to achieving their larger goals for mental health reform. Four years goes by fast. Do they really want to spend a big chunk of it seeking revision of the City Charter and working through all the heavy lifting of creating a major new agency, when they could just get straight to work on enhancing and augmenting mental health services within the existing infrastructure of city government? If they decide that creating DCS is just too important to abandon, they should at least tell us why — something their campaign platform never gets around to.
They should also consider which mental health policy battles are worth fighting, and which are distractions. Does the administration want to spend its limited political capital on “ending co-response” and changing protocols to make involuntary removal a “last resort” — policy shifts likely to be highly unpopular with both the public and the mental health professionals already doing the work? If so, they should explain how they plan to convince those lacking awareness of their severe mental illness to volunteer for treatment, and how they plan to keep clinicians and outreach workers safe in their engagements.
Whatever choices they make, the Mamdani team should see no shame in pivoting away from hastily hatched campaign promises as they come into power and learn more about a deeply complicated issue. I hope I can help them do that during this period of transition.
This is the first in an occasional series of Vital City reflections by current and former public officials on lessons they have learned from their time serving the city.