Clevelanders speak about crisis response, what works, and how to improve it

Authors’ note: This report centers the experiences of people who are directly impacted by the current system for response to behavioral and mental health crises — those who have the most at stake in decisions about that system’s future. Grounding our research in communities closest to the issue strengthens our research by incorporating the unique expertise of people with lived experience.

Conducting research like this requires love, commitment, and selflessness from everyone involved. We can only do this work successfully through genuine communication and authentic relationship building. We can only do it ethically through grassroots and community-based efforts.

That's why we collaborated with grassroots organizations who are, day in and day out, serving and advocating for communities of people who are too often ignored, pushed aside, and abused by those in power. These partners facilitated sincere, culturally competent dialogue with survey participants. This work would be impossible without them.

We are grateful to the following organizations — their leaders, staffs, and volunteers — for their expertise, dedication, time, and resources:

• The LGBT Center of Greater Cleveland
• Friendly Inn Settlement, Inc.
• Magnolia Clubhouse
• Northeast Ohio Coalition for the Homeless
• Stella Maris Cleveland
• University Settlement

Thank you for your continued support and partnership. We look forward to many more opportunities to serve and advocate for our communities together.

For almost five years, a coalition of directly impacted community members, clinicians, advocates, researchers, and policy experts has explored what care response could look like in Cleveland and Cuyahoga County.

The conversation about care response here has included this two-part study looking at community members’ experiences with first responders and crisis response in general, as well as several other key publications.

In August 2022, Policy Matters Ohio collaborated with the Center for Community Solutions and the Mental Health & Addiction Advocacy Coalition on a policy brief describing care response and ways to sustainably fund such a program.[1]

In October 2022, Policy Matters Ohio and REACH published the first part of this research study, “Talk to me like a regular person, not a criminal.” The report focused on the homeless community, featuring research gathered in partnership with the Northeast Ohio Coalition for the Homeless and its Homeless Congress.[2] Among other public writings, our two organizations co-published a May 2023 op-ed in cleveland.com effectively laying out the benefits care response could bring to our community.[3]

In 2023, R Strategy Group consultant and former OMHAS medical director Dr. Mark Hurst presented the ADAMHS Board of Cuyahoga County with a report outlining major program components and planning considerations for a care response pilot program.[4]

We wrote two more recent op-eds, one about the Shaker Heights co-response program that is expanding to other suburbs and seems to be moving toward a version of care response,[5] and one drawing on what we’ve learned from Cincinnati’s Alternative Response to Crisis program and the need for community engagement.[6]

Along the way, community members and advocates have written op-eds, shared petitions, hosted teach-ins, run statewide roundtables, and given public comment to city and county elected officials.

This report relies on the same research methodology as the phase 1 report. Taken together, they provide a comprehensive look at the opinions, needs, and experiences of the community members most likely to use a care response service in Cuyahoga County.

Our methodology contained a mix of quantitative and qualitative questions, based on a research study designed by Portland State University for Portland Street Response, a non-police crisis response program in Portland, Oregon.[7] This methodology leverages strong relationships between direct service providers and community organizations and the community members they serve, using representatives from those organizations as survey administrators so that survey participants feel more comfortable and empowered to share the reality of their experiences.

The survey results revealed no single, simple takeaway, except that no two crises are alike. To provide the best possible care, dispatchers and first responders need a diverse array of options. The city and county should expand those options by expanding care response, investing in its success and, most importantly, creating opportunities for community members to provide input that gets turned into action.

Our goal was to hear from our most vulnerable community members, neighbors, and family members — those who frequently experience crisis, make use of crisis-response services, or live in underserved and overpoliced neighborhoods. The survey demographics reflect this goal, not a random sample.

A total of 580 survey participants contributed to this report. Most participants identified as either Black (46%) or white (35%). Eighty-eight percent said they were 18 to 64 years old. Forty percent of participants identified as members of the LGBTQ+ community and 52% said they had experienced homelessness. The table below details the survey population:

The survey asked about people’s experiences calling 911 or receiving assistance from any first responder for any crisis. About three in five survey participants provided answers to this survey based on their own experience, while others relayed experiences that had occurred to others in their lives.

When asked who responded to their crisis, participants reported interactions with police or other law enforcement officers, firefighters, EMTs/paramedics, and others. Many reported that two or even all three types of these responders came. Police were the most common, with 334 participants reporting that police came, either alone or with other responders.

EMTs/paramedics, also alone or with others, were a close second, with 324 participants saying they responded. A firefighter response, with others or alone, was listed by 138 participants. Because respondents were able to choose more than one, the total number of answers is greater than the number of survey participants. Forty didn’t respond and seven gave other responses, including a friend, a social worker and a nurse in the building.

Participants described a wide range of crisis situations and needs. Figure 1 shows that they reported more than 240 crises related to physical health, 188 related to mental or behavioral health, 139 related to substance use or withdrawal, and 89 related to a lack of stable housing or homelessness. Participants were able to choose more than one answer and shared stories of crises related to domestic violence, fires, car accidents, and varied threats to safety. Fifty-eight shared that their experiences in crisis were solely related to mental or behavioral health.

Figure 1


As shown in Figure 2, when asked who called for help in the crisis situations described, the most common responses were that the participant called (221) or a friend or family member (136).

Figure 2

Participants were asked to assess their most-recent interactions with first responders. A little less than half (48%) said they were positive or very positive. Almost a third (31%) rated them neutral. More than one in five (21%) assessed them negatively or very negatively.

Figure 3

Two open-ended questions reveal a more complex picture about how participants evaluate first responders. We asked participants:

• What did the first responder(s) do that made the interaction feel positive or negative?
• What made you feel safe or unsafe in this interaction?

Of the 177 participants who rated their interactions as neutral, 39% (69) provided an open-ended response that we judged to be wholly or partially negative based on the experiences participants shared and the language they used to describe them, and 38% (68) gave a response that we found wholly or partially positive; the rest were neutral or left blank. Of those who rated the interactions as positive or very positive, 12% provided responses that described negative experiences, while 19% of participants who rated them negative or very negative provided responses that suggested positive experiences.

Figure 4

Sample quotes from participants who rated interactions as neutral but gave negative responses to open-ended questions:

• “I told them multiple times that I am trans…. When the officer patted me down [they] continuously grabbed my crotch and other places saying ‘something’s not right.’”
• “When I asked for a ride to a shelter they wouldn’t help me. Told me I should walk because I ‘could stand to lose some weight.’”
• “Wouldn’t help me find a shelter…. Just yelled at me to leave.”
• “They didn’t seem to genuinely care and were moving slow.”

Sample quotes from participants who rated interactions as neutral but gave positive responses to open-ended questions:

• “Helped me find a shelter that would take me with AIDS/HIV.”
• “I felt safe because their tone was respectful.”
• “They came fast and in a hurry to the scene.”
• “The dispatcher lady stayed on the phone with me the whole time until the EMT got there.”

We asked participants how safe different types of first responders make them feel. Questions focused on police/law enforcement, EMT/paramedic, social worker, mental health professional, and a peer with similar experiences. While care response programs by definition do not involve police, we felt it was important to include law enforcement in this part of the survey. Most participants indicated prior interactions with law enforcement, but their prior experiences with other response types are likely limited in crisis situations, given that police and EMTs/paramedics are currently the only first responders in Cleveland from this list. (We did not include firefighters in this part of the survey.)

For this set of questions, we asked participants if they agreed with the following statements when these types of personnel respond to a nonviolent crisis situation:

• A police or law enforcement officer makes me feel safe.
• An EMT/paramedic makes me feel safe.
• A social worker makes me feel safe.
• A mental health professional makes me feel safe.
• A peer with similar experience makes me feel safe.

Answers to this set of questions illustrate complex interactions and relationships with first responders. Community members who perceived previous interactions with social workers as negative — such as a Child Protective Services social worker disrupting parental custody — may be less likely to rate a crisis response from a social worker as safe. Lack of knowledge may be a factor as well: Crisis response by a peer with similar experience is uncommon outside the substance use recovery community; many survey participants could only answer that question hypothetically.

This diversity of experience and perceived safety is especially clear in feelings about police officers: Just over one-third of participants (36%) said police officers made them feel safe, just under a third (31%) disagreed, and a third (33%) were neutral.

Majorities said other types of first responders made them feel safe, including 63% for EMTs/paramedics, 56% for social workers, 66% for mental health professionals and 64% for peer responders. Fewer than 10% of participants disagreed with the statement that these types of responders made them feel safe, with the exception of social workers, with 14% saying social workers didn’t make them feel safe.

Figure 5

Below are sample responses to the open-ended question of how first responders (of any type) made participants feel safe:

• “They made me feel safe because they made sure I got what I needed.”
• “They made me safe by how they talked to me.”
• “They made me feel safe because they showed up fast.”
• “They treated me with respect which made me feel safe.”

Participants were then asked what a non-police care response team could do to make them feel safer. Figure 6 shows the most popular responses: providing proper identification, including wearing badges or uniforms; promising confidentiality to people receiving crisis care, including not checking for warrants; and arriving in a non-police vehicle. Fifteen participants said being respectful, six said nothing can be done to make them feel safer, and three said reducing response times. Participants were allowed to provide more than one answer.

Figure 6

Next, the survey asked participants to consider what supplies, information, or services first responders should be able to provide. Figure 7 shows the most common responses. In addition to those listed in the chart, 55 said “connections to community resource centers.” Participants were allowed to provide more than one answer.

Figure 7

Participants were asked to choose from a list of training areas they considered most important for care response teams, with the opportunity to select more than one and to add their own answers. The top answers are listed in Figure 8. Notably, some participants chose only one response, and their top two choices were “mental health awareness” (36) and “listening” (30).

Figure 8

Because 58% of participants (334) reported responses involving police, either alone or with other types of responders, we were especially interested in how officers influenced people’s sense of safety. Figure 9 breaks down open-ended responses to the question, “In general, does police presence in emergencies or crisis situations make you feel safer or less safe?” Half of all participants made statements to the effect that police presence made them feel safe or safer; 22% indicated it made them feel less safe or unsafe and 21% said it depended on the situation, or that it made them feel safer in some ways and less safe in others. The remainder provided no response or a response that was unclear.

Overall, white and Black survey participants answered this question similarly. Our analysis found more variation between participants who identified as LGBTQ+ and those who did not. Forty-one percent of survey participants who identified as LGBTQ+ reported feeling safe or safer due to police presence, compared to 55% of those who did not identify as LGBTQ+. Twenty-four percent of LGBTQ+ participants reported feeling unsafe or less safe, compared to 20% of non-LGBTQ+ participants. Similar percentages – 22% who identified with the LGBTQ+ community and 20% of those who did not – qualified their responses by saying it depended on the situation, they felt both safe and unsafe, or a similar response.

Survey participants who said they had experienced homelessness were less likely to report feeling safe or safer (47%) with police presence than those who had not experienced homelessness (54%) and were more likely to report feeling less safe or unsafe (27%) with police presence than their counterparts (17%). Of those who had experienced homelessness, 17% said it depended on context, or that they felt some combination of safe and unsafe, compared to 25% of their counterparts.

A higher percentage of female participants (57%) reported feeling safe or safer with police presence than men (49%), and both cisgender males and females were much more likely to report feeling safe or safer with police presence than nonbinary (29%) and transgender participants (22%). Transgender participants were much more likely to report feeling unsafe or less safe (46%) with police presence compared to nonbinary people (25%), cis men (24%) and cis women (17%). Participants who said it depended on the situation or that they felt both safe and unsafe in this analysis ranged from 19% to 29%.

Figure 9

While most of the answers were just one or two words (safe, unsafe, etc.), longer responses included:

• “Depends on all ethnicities involved. Black on Black crisis I might feel safe, Black on white crisis I feel less safe.”
• “I usually feel safe if they are not racist cops.”
• “Less safe - when in active addiction, they were not sympathetic.”
• “Less safe in general, very unsafe if they have a gun.”
• “Makes me feel safe for me because I know how to stay out the way.”
• “Safe, police are there to help.”
• “They make me feel safer just being around and roaming the city.”
• “Yes, because they have the proper tools and training to protect me.”
• “They are clueless in a lot of situations.”
• “Safer as a white woman, but concerned for less privileged groups and treatment they'll receive.”
• “In general, more safe. At times, less safe. Depends on the situation and the police who are there.”

After words like safe or safer, among the most common words in open-ended responses were compassion, care and empathy — used sometimes to describe their presence and sometimes their absence.

Positive responses included:

• “Positive, they were compassionate and caring.”
• “Greeted me - ‘How can I help you?... Please don't move.’ Let me know they would assist in helping me. They were very empathetic and sympathetic.”
• “Calm demeanor - talked slowly and listened to my answers.”
• “Responded quickly, helped out with next steps to take.”

Negatives responses included:

• “I felt negative about all them because they never come when there [is] a emergency.”
• “The ambulance came and didn’t help the woman who was walking in the middle of the road. They sat there and said, ‘if she gets hit, call us back.’”
• “This was a domestic situation so I felt positive that they showed up in a timely manner. But I felt negative because police downplayed my domestic trauma because they did not recognize the social clues of victims of domestic violence.”
• “Disregarded my issues with mental health and substance abuse issues.”

Throughout the survey, participant perspectives showed that they feel safer, or would feel safer, when the first responders training and abilities are aligned with their needs in a particular crisis situation.

Showing concern and empathy, for example, helped a lot in certain situations, but when there was a violent threat, participants wanted police presence, when there was a medical crisis, they wanted an EMT or paramedic, and when there was a fire, they wanted firefighters.

Two questions asked survey participants specifically about their perspectives on when police involvement is helpful, and when it isn’t:

• In what emergency or crisis situations would you like police to be involved, and why?
• In what emergencies or crisis situations would you not want police involved, and why?

Their responses aligned with the overall takeaway that they needed responders to provide the help they needed in each situation.

The most common responses from participants about when they do want police involved were: emergencies (90) although some qualified their responses with words like “actual”; situations that involve violence (69); when they have concerns about their safety (61), being threatened (37) or harmed (30); and situations involving shootings or guns (43). Many participants mentioned more than one of these words or phrases.

• “Any violent crime because someone can be harmed or killed.”
• “Every crisis situation because they do things ‘by the book’”
• “I would like the police to be involved when someone is being aggressive and I feel unsafe.”
• “Police should be involved in actual emergencies. A lot of them aren’t trained for mental health issues.”
• “When someone feels unsafe the police should be involved even if majority of people don't like the police. Police can save a life if they get involved.”
• “Any if they are open minded.”

The most common responses from participants about when they do not want police involved were: mental health (127); family or domestic situations (45), although some participants said that if the situations were violent, police should be involved; homelessness (26); situations where no violence occurs (23); drug overdoses (19); and suicide (10). Many participants mentioned more than one of these words or phrases.

Responses give to this question included:

• “Physical health, or a mental health crisis that hasn’t escalated to where the police are needed.”
• “In any situation that their presence would cause more trauma or chaos.”
• “During any mental illness situation because they are not trained to deal with them.”
• “Family situations because they get involved in it and it escalates a lot of times.”
• “Mental health because I feel their training is very inadequate and just seeing police can [exacerbate] a situation unnecessarily.”
• “Harassment against homelessness people when they are not doing nothing illegal.”
• “Emergencies involving drug overdoses or other health related emergency calls where the person or persons needing attention could be in further danger legally after medical services are rendered.”
• “When it’s a domestic situation I feel like police add fire to the situation.”
• “Any misunderstanding between individuals that have no violence in the situation.”

The first phase of this research project focused solely on community members experiencing homelessness. This second phase focused on LGBTQ+ and BIPOC community members; people diagnosed with severe mental illnesses or substance use disorders; and those living in historically underserved or overpoliced neighborhoods. Taken together, these participants represent much of the potential service base for a non-police care response team.

Participants across both phases generally supported a complex view of first response, where no single response type is best in all scenarios. They generally reflected a rise in feelings of safety or positivity when a first responder’s experience clearly corresponded to the nature of the crisis, such as firefighters responding to fires, law enforcement responding to threats to safety, and EMTs or paramedics responding to physical health crises. A fourth type of response is needed for crises related to mental and behavioral health, which are distinct from other types of crises in their impacts and outcomes.

Of particular note: About half of all participants in phase 2 reported that they had experienced homelessness, along with every respondent in phase 1. First responders of all stripes are familiar with the ways that homelessness and housing instability correlate with concerns for health or safety, as homeless community members seek alternative ways to rest, find shelter, manage their health and hygiene, and access food and basic services. A care response team would do well to develop deep fluency in providing for community members experiencing homelessness or housing insecurity, including relationships with the Northeast Ohio Coalition for the Homeless and other local housing resources.

Just like phase 1 participants, those in the second phase reflected a wide range of experiences, interactions, and impressions of first responders. When discussing what type of response made them feel most safe, some participants shared that only a police response made them feel safe, while others shared that only a response that did not include law enforcement would make them feel safe. Discussions of whether police officers made an individual feel safe were also complicated by identity and life experience, with some participants sharing experiences with transit police, border patrol, and even security officers as impacting their impressions of law enforcement.

One point of strong concurrence between participants in both phases was the importance of interpersonal skills like listening, communication, mediation, respect, empathy, and compassion. This emphasis was present in commentary about all current first responders and the services available today, as well as commentary about what is hoped for from possible care response teams. In both groups, some participants chose listening or communication skills as the single most important thing that a first responder should have to make them feel safe, over and above being able to provide Naloxone or offer physical healthcare or many other measures that could be perceived as life-saving.

Access to physical healthcare services, including the presence of paramedics or EMTs, was also generally preferred in most cases by most responses. This may further highlight the growing awareness of the intersection between public health and public safety.

Participants in both phases also indicated increased feelings of safety when first responders identified themselves early and clearly. Both groups selected wearing a uniform or showing a badge or identification as the number one factor that would make them feel safe in an interaction with a first responder.

Strengthen community engagement and participation – The ADAMHS Board of Cuyahoga County and the city of Cleveland, including the city health department and those working on crisis response elements of the city’s consent decree, must work together to establish a meaningful role for community members in strengthening the crisis response system. Listening to advocates and directly impacted people would help ensure that, over time, the new care response program truly meets the needs of the community and reaches the people who most need help. This type of engagement can also help educate residents who rely on crisis care, making sure more people know that this new approach is an option. Ideally, the ADAMHS Board and the city will set up an advisory group alongside any care response pilot or program that can help evaluate and direct the evolution of the care provided, and create opportunities in the planning, implementation, and evaluation of the pilot to listen to the voices of those most impacted.

Provide appropriate, ongoing training – The ADAMHS Board and the city should call on advocacy groups, service providers, and community members to help train care response teams and call-takers, rooting instruction in the experience of directly impacted people. Highlighted in this report are the priorities of community members, such as mental health awareness, listening, trauma-informed care, and conflict resolution.

Begin planning the transition to 911 – While the current care response pilot is set to respond only to calls to 988, the national suicide and crisis hotline, it is critical that work begin immediately to set up the system so appropriate calls to 911 can be transferred to care response teams, either directly or through 988. Successful programs in many other cities, large and small, have 911-988 interoperability, recognizing that most people know about 911 and will likely call that number in crisis, while 988 is still new and not widely known. Furthermore, Cleveland is already investing federal funds, partially matched by local dollars, in a three-year program to upgrade 911 dispatcher training and readiness to help people experiencing mental and behavioral health crises. While the timing may not be perfect, we must take advantage of the opportunity to create a strong, coordinated program.

Establish clear, practical protocols – Since 988 is the only route to care response teams for now, FrontLine Service, the ADAMHS Board, and the city must establish criteria that allow 988 call-takers to quickly determine which cases will be sent to care response teams; which, if any, will be diverted to police response; and which will be handled by phone. Armed police response can no longer be the default for people in crisis: Outcomes will only improve when non-police options are truly available and accessible for everyone.

Acknowledgements
The research team at Policy Matters Ohio and REACH would like to thank all participants in the first and second phases of this report for their willingness to spend time sharing their perspectives and experiences.

REACH and Policy Matters would like to thank our funders: The Char and Chuck Fowler Foundation, The George Gund Foundation and the Cleveland Foundation.

We would also like to thank our colleagues at Policy Matters who have assisted in different ways, as they always do, especially Molly Bryden for helping us analyze and think about the data and findings.

For more information on care response and the groundwork being created in northeast Ohio to support this program, contact Policy Matters Ohio.

Tags