Mamdani’s new department can work if done right.
In recent weeks, New York Mayor Zohran Mamdani has begun to staff up his Office of Community Safety, charged with implementing “a whole-of-government” approach to challenges ranging from gun violence prevention and mental health and behavioral crisis response to hate crime prevention, victim services and subway safety. It’s an important start to creating a broader ecosystem of community safety that can both produce better outcomes and be more cost-effective than public safety approaches that rely more heavily on traditional enforcement. Although the mayor’s approach is sensible, evidence-supported and generally popular, it faces stiff headwinds. The difficulty lies partially in the inherent complications of implementing a new approach to an issue as complex as community safety in a city as large as New York. But the even larger challenge is the set of assumptions and misapprehensions —not to mention vested interests — supporting the status quo. For New York City’s new approach to realize its potential and to endure, the Office of Community Safety will have to minimize missteps (and be afforded some grace to course-correct when the inevitable one occurs), and a critical mass of New Yorkers will need to recognize (as millions across the country do already) that being “serious about safety” requires rejecting tired “tough on crime” tactics that over-rely on police and instead adopting what research and experience tell us works to promote safety and build vibrant communities.
There is no shortage of evidence- and experience-informed guidance for jurisdictions seeking to create a broader community safety ecosystem, including guidance tailored specifically to New York City. These expanded safety ecosystems often center around “community crisis response” (or “alternative response”), an umbrella term given to programs that operate alongside 911, police, fire and EMS as a sort of “fifth branch” of first response. Community crisis response can include not only licensed clinicians responding to mental health or behavioral health crises, but also support by peers who have had similar experiences and conflict mediation programs that can handle garden-variety neighborhood disputes. Street outreach programs, in which social service workers go to people in need rather than waiting for those people to come to them, are also part of this ecosystem. Still another part of the community safety ecosystem focuses on community violence intervention, a range of strategies used to prevent violence, including hiring community members who work to interrupt cycles of retaliation and de-escalate neighborhood conflict, including with victims of gun violence who are still in the hospital. Other community violence intervention programs may also seek to identify the small number of individuals responsible for gun violence in a community and focus deterrent efforts — and incentives to desist from violence — on that group; or provide skills training and support to help often traumatized kids better regulate their emotions; or implement victim support and trauma recovery programs that, in addition to providing critical relief, recognize and break the link between being hurt and hurting other people.
The lessons being learned from this growing ecosystem are too varied and robust to recount in full here, but it is worth focusing on some key takeaways specific to community crisis response. One of those lessons is that it can be safe and cost-effective for cities to send community crisis responders — whether licensed mental health professionals, mediators, or trained peer advocates, rather than police to a broad range of calls that don’t require handcuffs, a badge and a gun to resolve. In recent months, Greensboro, NC, Columbus, OH and Chicago, IL, have joined groundbreaking cities like Albuquerque, NM, Durham, NC and hundreds of others across the nation, in demonstrating how.
You’ve probably observed a situation in which a person seemed to need help from someone, but it didn’t seem like that someone needed to be a law enforcement officer. Maybe you have observed an individual obviously in the midst of some sort of mental health or behavioral crisis on the subway, or been concerned about someone who appeared to have nowhere to go and had fallen asleep in the corner booth at a coffee shop on a cold day. Or maybe it was a different kind of problem — a conflict with that new neighbor who won’t stop practicing the drums late at night, or that other neighbor who won’t keep her dog out of your flowers. Or different still, maybe you are the parent of a child living with a serious mental illness who is fine most of the time, but not always.
In addition to being familiar with incidents like these, you likely know that, on relatively rare occasions, they end tragically when police respond. But you, like most people, might be surprised to learn that up to 20% of the time police spend on calls they are dispatched to is spent on behavioral health-related concerns, making up a significant portion of their overall workload, as well as taking up more time on a per-call basis.
It is true that some calls need a police response. But it is not the case, as some have argued, that it almost always makes sense to send the police, either on their own or as part of co-response teams that might include mental health professionals, social workers and others. The portion of calls that can be safely and effectively handled by community crisis responders instead of police will vary by jurisdiction, but research and experience make clear it’s substantial. One study showed that nearly half of police call types, comprising up to a quarter of police service time, could be safely handled by non-police responders. Another study estimated that up to 20% of police dispatch time is spent responding to behavioral health-related concerns, ultimately concluding that “the true unmet need for alternative response — and the deaths alternative response could avert — in U.S. cities is far greater than previously thought.”
Two of the nation’s leading community crisis response programs demonstrate how it is possible to free up a significant portion of police time and yield better outcomes. Durham’s HEART program saved Durham police officers more than 5,500 hours in its first two years of operation. This included calls like a response to a “barricaded suspect” scene where responders talked a distraught man into leaving the home and getting the help he needed — with no injuries or even force used. More routinely, HEART responds to individuals in mental health crises or without a place to go and, unlike police, make what are known as “warm handoffs” to available services. Responding to similar calls, Albuquerque’s Community Safety Department freed up over 31,000 hours for police and fire, responding to more than 42,000 calls in 2025 alone.
How do programs like these safely respond, instead of police, to such a large portion of community need? In the most well-developed programs, community crisis responders are deeply integrated into the first response architecture, where risk assessment is part of the dispatch decision. They can be directly dispatched by 911, are GPS-tracked and carry radios with emergency buttons in case they need to ask for EMS or police backup. Just like EMTs and firefighters, they are trained and prepared to safely respond to uncertain situations.
And this preparation is working. In Durham, N.C., where the police department reports a violent crime rate comparable to New York’s, the HEART program requested police backup in only .02% of the over 45,000 calls it handled over the past four years, and responders reported feeling safe 99% of the time. Albuquerque’s Police Department reports violent crime rates almost double that of New York’s, but Albuquerque Community Safety responders also called for police assistance less than 1% of the time. This makes sense when you think about it; both firefighters and EMTs respond to other health-related calls without needing police to clear the scene first; experience and research indicate that most behavioral health crises can be treated the same.
Ultimately, even the most patient, well-trained, well-suited police officer still arrives wearing multiple weapons, a badge and handcuffs — any of which can on their own escalate a situation dealing with person in crisis.
In addition to allowing police to focus on policing issues, community crisis responders can prevent outcomes that are routinely subpar and occasionally tragic. It is important to recognize that bad police outcomes are generally not the result of bad police. Most police officers are well-meaning and many are quite skilled at de-escalation. It’s simply that we ask police to respond to calls that — predictably — would be better handled as non-enforcement issues by non-enforcement responders. We demand that police routinely handle a wide variety of behavioral health and other noncriminal issues, including substance use and homelessness, even where the ideal resolution is not a jail cell, but rather a connection to services. Unlike community crisis responders, it is not the job of police to know and make those connections. And the threat of enforcement and force carried by a police officer, whether explicitly named or not, can contribute to escalation that makes that connection much more difficult to make — and sometimes results in tragedy. On the rare occasions when police backup does become necessary, community crisis responders can be patched directly into the 911 computer-aided-dispatch systems just like firefighters or EMTs.
Community crisis response teams can include licensed clinicians with advanced degrees; peer advocates who have similar life experiences battling mental health challenges, addiction or homelessness; and other people with the right backgrounds and temperament for the job. Regardless of their background, community crisis responders typically receive weeks or months of initial and in-service training that prioritizes recognition of mental illness and substance use crises, conflict resolution, motivational interviewing, a knowledge of community resources, cultural sensitivity and a breadth of de-escalation skills.
Given this different mandate and orientation, community crisis teams are more likely to get people help that avoids future contact with the criminal legal system, rather than funneling them into it. An evaluation of Denver’s STAR program found a statistically significant reduction in arrests in the year post-contact compared to a police-only response. Analyses of Durham’s HEART program have found that it reduces crime reports, arrests and response times. Mobile crisis teams — an intervention closely related to community crisis response — appear to lower rates of arrest and hospitalizations and increase informal resolutions, connections to services and provision of follow-up services, as compared to both law enforcement-only responses and co-responses made up of law enforcement paired with a clinician. A multi-site study found that the likelihood of arrest in the year post-contact was 45.2% lower for mobile crisis teams.
Additionally, community crisis responders are not armed, meaning there is virtually no risk they will unnecessarily kill a person in crisis or turn a call about a minor crime into a wrenching tragedy. People with mental illness account for between a fifth and a third (or more) of those killed by law enforcement. To be clear, compared to the number of interactions police have with persons in mental health or behavioral crisis, police killings of people in crisis are relatively rare. But when they happen, they are not only incalculably tragic to the loved ones of the person killed, but they undermine the legitimacy of policing writ large. Even setting aside the use of deadly force by law enforcement, police are 1.4 to 4.5 times more likely to use force against people with mental illness, and people with mental illness report feeling vulnerable and fearful of police.
A police response also fuels the overcriminalization of mental illness. 25% of all people with mental illness are arrested at some point in their lifetimes. One study summarized the problem: “Arrest may be the most convenient, efficient and pragmatic disposition option, even when officers recognize that it will not best serve the needs of the [person with mental illness].” More colloquially, when your only tool is a hammer, every problem looks like a nail.
It’s unrealistic to hope we can train police into being something they are not. Police have a different mandate — one that includes enforcing the law and apprehending criminals, using violent coercion, consistent with the law, as deemed appropriate. Police training thus prioritizes quick control and the projection of authority, all as part of how officers are taught to ensure scene safety in criminal situations. This runs directly counter to what someone in crisis needs: an approach that slows a situation down, includes the person in decision-making, and creates space to de-escalate so the person can regain their own sense of control. Ultimately, even the most patient, well-trained, well-suited police officer still arrives wearing multiple weapons, a badge and handcuffs — any of which can on their own escalate a situation dealing with person in crisis. Co-responder programs, in which police and clinicians team up to respond to calls, can mitigate some but not all of of these inherent disadvantages of a police response, and they have been found to be less effective than a non-police community crisis response. Similarly, Crisis Intervention Training teaches police about mental illness and how to better approach such situations, but research and experience indicate it is no substitute for a community crisis response approach.
While community crisis responders are proving effective at reducing arrests, freeing up police resources and successfully resolving crises in neighborhoods across the country, there is plenty of work to be done before the approach embodied by the Office of Community Safety, and ultimately a new Department of Community Safety, can work at scale in New York City. Alongside the challenges of implementing new training, technology and equipment to make this approach work is the even harder challenge of dismantling decades of misapprehensions about what the police should do given the widespread belief that they’re the best answer to public safety-related challenges. But New York City has done it before. There was a time when NYPD officers were the City’s sanitation workers. Most would agree that reassigning those duties to other public servants made sense. If New York’s new Office of Community Safety is allowed to live up to the vision Mayor Mamdani outlined in his campaign, it might someday be looked back on as an equally sensible advance in public safety that transformed the public safety landscape in New York City, to the benefit of everyone who lives, works or visits.
