Helena A. Addison, MSN, RN, PhD(c)*, Ruth Shefner, MPH, MSW, PhD(c)†, Jennifer Wood, PhD, MA‡, Evan Anderson, JD, PhD*
ABSTRACT
Exclusive reliance on police in the response to mental health crises can result in avoidable injury or missed connections to supportive services. Many cities are experimenting with co-deploying police officers alongside health professionals or deploying teams comprised entirely of civilian health professionals. No studies have explored the perspectives and preferences about these programs among residents in structurally disadvantaged areas where: mental health distress is more common, mental health services are less accessible, and involvement with police is more frequent and fraught. In this survey of residents from two such areas, many respondents suggested that police presence is necessary during the response to mental health crises because of the risk of violence but were simultaneously uncomfortable with police officer involvement. Discomfort with police involvement was especially strong among younger and Black residents. Support for co-deployment was high across all subgroups.
Key Words: Co-deployment, behavioural health emergencies, community perspectives, first responders, structural disadvantage
Police officers respond to a variety of incidents that do not involve crime or immediate threats to public safety (Vermeer et al., 2020). Many of these incidents involve a community member with unmet mental health needs. Some estimates suggest that 6% to 7% of all police interactions involve a person with a diagnosed mental illness (Morabito et al., 2018), and that police are involved in the pathway to care for as much as one-third of people with mental illness in the United States (Watson et al., 2021). Given their skill sets and available resources (Anderson & Burris, 2017), some of these interactions result in avoidable harm. People with untreated mental illness are 16 times more likely to be killed in a police shooting than other community members (Fuller et al., 2015; Rohrer, 2021). Police officers also may lack the knowledge and time to link people in distress with appropriately supportive resources, especially where there are other pressing demands, such as responding to serious violent crime (Shefner et al., 2023).
In the past few decades, cities have been experimenting with two strategies to respond more safely and effectively to acute mental health distress in the community (Beck et al., 2020; Watson et al., 2019). One is enhanced training for police officers such as the Crisis Intervention Training (CIT), developed in the 1980s, which provides training for police officers on behavioural health, strategies for de-escalation, and information about local health services (Rogers et al., 2019). The other is the deployment of response teams that rely less or not at all on police involvement but are dispatched through 911/emergency calling systems. This includes the co-deployment of police officers alongside mental health professionals (Shapiro et al., 2015) and deployment of fully civilian response teams, such as the longstanding CAHOOTS model, which pairs a nurse or an emergency medical technician (EMT) with an experienced crisis worker (El-Sabawi & Carroll, 2021).
Extensive research documents the benefits of CIT, including improved officer knowledge of mental illness, reduced use of force, and increased diversion to community psychiatric services rather than jails (Kane et al., 2018; Rogers et al., 2019; Taheri, 2016; Watson et al., 2017). A smaller but growing body of evidence describes the role of new response models in decreased involuntary psychiatric hospitalization (Puntis et al., 2018), reduced injury (Lamanna et al., 2018), increased linkages to psychiatric services (Shapiro et al., 2015), and shortened officer time on scene (Kisely et al., 2010). Some qualitative evidence also suggests that community members perceive less police-centred response teams to be better at de-escalation and less threatening (Boscarato et al., 2014; Puntis et al., 2018).
However, generalizing about the effects of these new models is complicated because they operate in different forms and contexts (Marcus & Stergiopoulos, 2022). Most studies evaluating programs have been conducted in settings— unlike the United States—where access to behavioural health care and high-quality safety-net housing are high and where rates of gun possession and serious violence are relatively low. Few, if any, studies explore the experiences and perspectives of community members living in areas marginalized by high rates of poverty and serious crime, where rates of mental distress are higher, due to disproportionate exposure to poverty, violence, and other trauma (Walker & Diforio, 1997), and involvement with police is greater and riskier (Gaston et al., 2021; Leslie et al., 2022). The present study examines preferences for police involvement in mental health crises in two areas of Philadelphia where there is concentrated economic disadvantage and where unmet social and health needs are extensive.
Respondents were recruited in collaboration with the Institute for Survey Research (ISR) at Temple University, which has extensive expertise exploring the opinions of “hard-to-reach” populations in Philadelphia. ISR maintains a panel of community members who have opted to participate in ongoing research, called BeHeardPhilly (BHP), which is a broadly representative sample of Philadelphia residents. Participation in this study was limited to BHP residents living in four police districts with high rates of serious crime. These police districts were chosen because they offer a pre-booking diversion program (Anderson et al. 2022), which was the subject of another set of survey items. At survey deployment, 1,443 members of BHP were eligible. Invitations and reminders to participate in the survey were distributed by telephone call, e-mail, and text message, and ten $30 gift cards were provided as a raffle incentive. The response rate to the survey was 20%, resulting in 293 respondents. Table I provides demographic information for the respondents.
TABLE I Demographic information of respondents
The survey asked respondents to provide their perspectives about the involvement of police in responding to people in acute mental health distress. The items were developed in part based on findings from 10 focus groups with over 50 Philadelphia Police Officers conducted a few months earlier (Shefner et al., 2023). Those officers suggested that their involvement in the response to mental health crises is often inappropriate or even harmful but suggested that their presence is practically necessary given the high risk of violence in such incidents. Officers also suggested that community members often embellish the risk of violence in 911 calls to get a more certain and rapid emergency response. Preliminary measures exploring these propositions and associated concepts were created and refined through iterative review by academic and practice-oriented colleagues. ISR also conducted internal survey pretesting before deployment.
Respondents were first asked to identify how many times they have summoned emergency assistance by calling 911, using a 5-point scale (never, 1–5, 6–20, 21–50, 51 or more). They were next asked to identify their agreement, on a 4-point Likert scale (strongly agree, agree, disagree, strongly disagree), with the statement that Philadelphia police respond quickly to 911 calls. Respondents then were provided with the following statement “Some people claim that some community members who call 911 sometimes mention that there is a gun or other weapon— even if there isn’t one—to get a faster police response” and asked to identify whether this happens never, once in a while, or all the time. They indicated whether they have ever done so in the next question (yes/no). Finally, respondents were asked whether they perceive their neighbourhood as safe.
Respondents were asked next to indicate their level of agreement (strongly agree, agree, disagree, strongly disagree) with six statements about police involvement in mental health crises. One item explored the perceived adequacy of police training for mental health emergency response. Three items asked respondents to imagine that a loved one was having a mental health crisis and then to identify (1) whether they believed that a police officer is necessary to preserve safety, (2) whether they would feel comfortable with a police officer present, and (3) whether they would want a police officer involved in the emergency response. Two items also assessed respondent comfort, in the same situation, with receiving assistance from a co-deployment team and from a civilian-only response team. Two additional questions assessed agreement with the statements that the city should hire more police and mental health professionals.
To examine the relationships between demographic characteristics and perspectives about mental health responses, chi-square analyses were conducted after collapsing the 4-point agreement scale into agree (strongly agree, agree) and disagree (disagree, strongly disagree) and dichotomizing income and age by median splits. To explore whether the number of previous 911 calls exhibited any dose-response relationship to perspectives, Cochran-Armitage Trend tests were conducted after collapsing the 5-point scale (never, 1–5, 6–20, 21–50, 51 or more) into three levels (0–5, 6–20, 21+). Logistic regression models explored whether the three preference dimensions (need, comfort, want) for police involvement varied based on race, gender, age, income, and feeling safe in the neighbourhood. In these analyses, race was dichotomized as Black and non-Black, gender was dichotomized as male and non-male. Predictor variables were included individually and then collectively to determine unadjusted and adjusted odds ratios (ORs). P values were reported with .05 as the threshold of statistical significance.
Table I provides data about the demographic characteristics and the 911 experiences of respondents. Most respondents (72.7%) identified as female. Of the respondents surveyed, 43.7% identified as Black, 38.5% identified as White, and 16.4% identified as another race (including Asian, Native American, or Pacific Islander) or two or more races; most respondents (71.3%) were non-Hispanic. About half of the respondents were over the age of 55 years. A large percentage of respondents had completed high school (24.2%), or some college or an undergraduate degree (52.2%), and more than half (57.6%) reported an income of less than $35,000 annually. Most respondents had either never called 911 for emergency assistance (28.5%) or had called between 1 and 5 times (53.9%).
Table II presents participant experiences with and perceptions about 911 services and neighbourhood safety across subgroups defined by race and number of times summoning 911 assistance. Disagreement with the statement that Philadelphia Police respond quickly to 911 calls was high overall (65.6%) and was positively correlated with having called 911 previously (Chi-square p=.002; Cochrane Armitage Trend Test=0.027). Table II also presents respondents’ perspectives on and experiences of call embellishment. There was a positive and graded relationship between the perception that embellishment occurs and previous times calling 911 (Cochrane Armitage Trend Test=0.005), with agreement ranging from 62% among respondents with five or fewer lifetime calls to 93% among respondents with more than 20 previous 911 calls. In contrast, the 5% of respondents who reported suggesting themselves that there was a gun or other weapon present to get a faster response were not disproportionately distributed across the three groups (Chi-square p=.544). No statistically significant differences were observed on these items between different racial subgroups. Just over half (54.0%) of respondents reported feeling safe in their neighbourhood.
TABLE II Experiences with 911 systems in emergency response, by race and previous 911 calls
Table III presents respondent perspectives on the involvement of police in mental health crises. Although only one-third of respondents (32.0%) believe that police are adequately trained to help people who are having a mental health crisis, three-quarters (76.6%) agreed that police are needed during a mental health crisis for safety reasons. White respondents were more likely to endorse both perspectives although the differences were small and statistically insignificant. In contrast, vastly fewer Black respondents (43.3%) felt comfortable with police involvement in a family member’s mental health crisis compared with white respondents (71.7%; p<.001). This disparity tracked differences in wanting police officers involved between Black (60.3%) and White respondents (80.0%, p=.004). Notably, gender played a role in wanting police among Black respondents as 83% of Black men want police involvement compared with only 55% of Black women (data not shown). The difference among White men and women was smaller but followed the same pattern, at 84% vs 80%, respectively (data not shown). A large majority of all respondents reported being comfortable calling a co-deployment team (92.9%) and a large proportion were comfortable calling a civilian-only crisis team too (76.9%). However, only 55.3% were comfortable with a conventional police-only response (data not shown). Most respondents believed that Philadelphia should hire more police officers (81.0%) and nearly all respondents believed the city should hire more mental health professionals (98.6%).
TABLE III Perspectives on police and alternative responses to mental health crises, by race and previous 911 calls
Table IV presents the findings of logistic regression models exploring the relationship between sociodemographic predictors (race, gender, age, income, and feeling safe in their community) and the perspectives of respondents on the three preference dimensions for police involvement in mental health responses. Black respondents were 50% less likely to believe police were necessary (adjusted OR 0.51, p=.047) and older respondents were 96% more likely to believe police were necessary (adjusted OR 1.96, p=.052). Black respondents were similarly less comfortable (adjusted OR 0.45, p=0.006) and older residents were similarly more comfortable with police involvement (adjusted OR 2.21, p=.009). Respondents who reported feeling safe in their community were also more likely to report being comfortable calling the police (adjusted OR 1.85, p=.034). Finally, Black respondents were less likely to want a police officer involved (adjusted OR 0.51, p=.037), while feeling safe in the community (adjusted OR 3.12, p<.001) and being male were associated with higher odds of wanting a police officer present (adjusted OR 3.99, p=.002).
TABLE IV Predictors of needing, being comfortable with, and wanting police presence during a mental health crisis
This study is subject to important limitations. The readability of the survey instrument was examined closely by the research team and ISR staff. However, misclassification bias is possible to the extent that any questions were misunderstood. The survey questions are abstract, and it is possible that respondents might respond differently to questions that are anchored in more specific scenarios. Respondents likewise might envision what constitutes a “mental health crisis” very differently. There is assumed variability across the sample with respect to relevant lived experiences and social vulnerabilities. Some respondents—or their loved ones—may have experienced involvement of police during a mental health-related event that informed answers to the survey in a way that would differ from those answering in the abstract. Finally, the sample appears to be older, wealthier, and more highly educated, with a higher proportion of women, than the population from which it is drawn despite the general representativeness of the BHP panel and the relatively high response rate. The resulting selection bias underscores the difficulty of engaging populations in research who have been marginalized for generations. Our findings should be understood as reflecting only a portion of the perspectives in these high need areas; engaging younger, poorer, and less educated community members in similar surveys is an important priority for future research.
Notwithstanding these limitations, our study provides some important insights to guide efforts at realizing a healthier response to mental health crises. The Philadelphia residents in our sample were overwhelmingly supportive of a co-deployment model for responding to mental health crises in the community. This support was conceptually consistent with their beliefs that police officers do not have adequate training to respond to people having a mental health crisis but that police officers are needed at such incidents given their potential to turn violent. Perspectives on police involvement, however, were far from uniform. Black respondents were significantly less likely than White respondents to feel comfortable with or want police officer involvement in the response to a family member’s mental health crisis. In our adjusted analyses, Black respondents were also significantly less likely to agree that police officers are needed in mental health responses to address potential safety concerns, although the absolute proportion of Black respondents questioning the need for police involvement was still substantially less than half. These findings are consistent with a large body of evidence documenting higher levels of mistrust and dissatisfaction with police among Black community members (Lai & Zhao, 2010; MacDonald & Stokes, 2006; Schuck et al., 2008).
Additionally, respondents’ overall preferences for police involvement in mental health crises may reflect stigmatized views about mental health, and specifically perceptions that individuals living with mental illness or experiencing a mental health crisis are likely to become violent. Previous research has documented that perceptions on these issues vary significantly between racial and gender subgroups (Anglin, Link & Phelan, 2006; Whaley, 1997), which might account for some of the observed patterning. Understanding how stigma shapes perspective on the need for police involvement, how those views vary between subgroups, and how to counteract them are important research priorities. Prior literature also aligns with our finding that younger residents are less comfortable in interactions with police, and with our finding that feeling safe in the community correlates with being comfortable with and wanting police involved in the response to a family member’s mental health crisis (Bolger & Bolger, 2019; Lai & Zhao, 2010).
Despite high face validity and growing evidence supporting the benefits of civilian and co-deployment response models, implementation has been limited in scale in the sense that such teams are only deployed in response to a small proportion of emergency services calls. In addition to concerns about budgeting and staffing (Carroll et al, 2021), there are concerns about sending these alternate response teams to situations that may involve ongoing or potential violence. This concern is exacerbated by the difficulty of assessing the characteristics of an incident through the 911 call-taking system, which is compounded by persistent disinvestment in emergency response systems. Few residents felt that the police responded quickly to 911 calls, most believed the city should hire more police officers, and nearly all agreed that the city should hire more mental health professionals. Relatedly, respondents provided the first documented evidence of call embellishment described anecdotally in focus groups of Philadelphia police officers (Shefner et al., 2023). More than half of the respondents believe that people embellish calls either “once in a while” or “all the time,” and about 5% of respondents reported that they themselves had suggested in a 911 call that there was a gun or other weapon to get a faster police response. It is hardly surprising that community members value speed and certainty in emergency responses and will manipulate existing systems to extract the most utility out of finite available resources. Our findings suggest that the denominator for potential alternate responses may be larger than observed in previous research (Lum et al., 2022) but only if those alternate models have the capacity to meet community expectations in terms of speed and quality of the response.
Call embellishment is one dimension of police mobilization (Wood & Anderson, 2023), and it is likely that police mobilization dynamics depend on contextual factors in Philadelphia, including but not limited to general deficits in the emergency response system. Another dimension of police mobilization is the collection of norms surrounding what people expect, and what they think they need, to address the crises they are experiencing. Changing emergency response practices—including through deployment of new response models—will succeed to the extent that initiatives reshape norms around enlisting the services. Underserved communities have learned to rely on the 911 system and on police in addressing a wide range of social and health problems because alternatives have been missing for generations (Watson et al., 2021). Philadelphia is part of a welcome national effort to implement a continuum of services to support people in distress with interventions better suited to their needs, but the need for structural interventions that address the social determinants of mental health—such as poverty, inequality, and institutional racism—must not be forgotten. There was considerable tension in respondent perspectives on the involvement of police in mental health crises, with many residents viewing them as necessary but not comfortable with or wanting their presence. This reflects a sad but understandable concession to the longstanding absence of services other than police. Resolving that tension will require not just more supportive services but concerted efforts to demonstrate—empirically but, perhaps more importantly, socially—how those services can better support people in crisis while preventing the vulnerabilities that result in acute distress in the first place.
Many cities are experimenting with co-deploying police officers alongside health professionals or deploying teams comprised entirely of civilian health professionals. In a survey of residents living in areas marginalized by high rates of unmet social and health needs, there was broad support for co-deploying police officers and behavioural health providers in response to mental health crises. The implementation of these and other alternate response initiatives must account for the role of help-seeking norms and the concerns and experiences of community members who have been historically underserved by local governments and who often harbour substantial concern with police involvement in crisis response.
Funding from Vital Strategies supported the formative design of this work. The funding sponsors had no role in the design of the study; in the collection, analysis, or interpretation of data; in the writing of the manuscript; and in the decision to publish the results. This work was partially supported by the National Institute on Drug Abuse, National Institutes of Health [R.S, grant number T32-DA037801].
The authors have no conflicts of interest to declare.
*University of Pennsylvania School of Nursing, Philadelphia, PA, USA;
†Columbia University, New York, NY, USA;
‡Temple University, Philadelphia, PA, USA.
Anglin, D. M., Link, B. G., & Phelan, J. C. (2006). Racial differences in stigmatizing attitudes toward people with mental illness. Psychiatric Services, 57(6), 857–862.
Anderson, E., & Burris, S. (2017). Policing and public health: Not quite the right analogy. Policing and Society, 27(3), 300–313. https://doi.org/10.1080/10439463.2016.1231188
Anderson, E., Shefner, R., Koppel, R., Megerian, C., & Frasso, R. (2022). Experiences with the Philadelphia police assisted diversion program: A qualitative study. The International journal on drug policy, 100, 103521. https://doi.org/10.1016/j.drugpo.2021.103521
Beck, J., Reuland, M., & Pope, L. (2020). Behavioral health crisis alternatives. Vera Institute of Justice. https://www.vera.org/behavioral-health-crisis-alternatives/case-studytiles
Bolger, M. A., & Bolger, P. C. (2019). Predicting fear of crime: Results from a community survey of a small city. American Journal of Criminal Justice, 44, 334–351.
Boscarato, K., Lee, S., Kroschel, J., Hollander, Y., Brennan, A., & Warren, N. (2014). Consumer experience of formal crisis-response services and preferred methods of crisis intervention. International Journal of Mental Health Nursing, 23(4), 287–295.
Carroll, J. J., El-Sabawi, T., Fichter, D., Pope, L. G., Rafla-Yuan, E., Compton, M. T., & Watson, A. (2021). The workforce for non-police behavioral health crisis response doesn’t exist—we need to create it. Health Affairs Forefront.
El-Sabawi, T., & Carroll, J. J. (2021). A model for defunding: an evidence-based statute for behavioral health crisis response. Temple Law Review, 94, 1.
Fuller, D. A., Lamb, H. R., Biasotti, M., & Snook, J. (2015). Overlooked in the undercounted: The role of mental illness in fatal law enforcement encounters. Treatment Advocacy Center, 1–27.
Gaston, S., Teti, M. J., & Sanchez, M. (2021). Does racial congruence between police agencies and communities reduce racialized police killings of civilians? Criminology & Public Policy, 20(4), 665–690.
Kane, E., Evans, E., & Shokraneh, F. (2018). Effectiveness of current policing-related mental health interventions: A systematic review. Criminal Behaviour and Mental Health, 28(2), 108–119.
Kisely, S., Campbell, L. A., Peddle, S., Hare, S., Pyche, M., Spicer, D., & Moore, B. (2010). A controlled before-and-after evaluation of a mobile crisis partnership between mental health and police services in Nova Scotia. The Canadian Journal of Psychiatry, 55(10), 662–668.
Lai, Y.-L., & Zhao, J. S. (2010). The impact of race/ethnicity, neighborhood context, and police/citizen interaction on residents’ attitudes toward the police. Journal of Criminal Justice, 38(4), 685–692.
Lamanna, D., Shapiro, G. K., Kirst, M., Matheson, F. I., Nakhost, A., & Stergiopoulos, V. (2018). Co-responding police–mental health programmes: Service user experiences and outcomes in a large urban centre. International Journal of Mental Health Nursing, 27(2), 891–900.
Leslie, T. F., Frankenfeld, C. L., & Hattery, A. J. (2022). Differentiating Black and Hispanic: outcome differences of segregated communities and police shootings in the USA, 2015–2020. Injury Epidemiology, 9(1), 1–11.
Lum, C., Koper, C. S., & Wu, X. (2022). Can we really defund the police? A nine-agency study of police response to calls for service. Police Quarterly, 25(3), 255–280.
MacDonald, J., & Stokes, R. J. (2006). Race, social capital, and trust in the police. Urban Affairs Review, 41(3), 358–375.
Marcus, N., & Stergiopoulos, V. (2022). Re-examining mental health crisis intervention: A rapid review comparing outcomes across police, co-responder, and non-police models. Health & Social Care in the Community, 30(5), 1665–1679.
Morabito, M. S., Savage, J., Sneider, L., & Wallace, K. (2018). Police response to people with mental illnesses in a major US city: The Boston experience with the co-responder model. Victims & Offenders, 13(8), 1093–1105.
Puntis, S., Perfect, D., Kirubarajan, A., Bolton, S., Davies, F., Hayes, A., Harriss, E., & Molodynski, A. (2018). A systematic review of co-responder models of police mental health “street” triage. BMC Psychiatry, 18(1), 1–11.
Rogers, M. S., McNiel, D. E., & Binder, R. L. (2019). Effectiveness of police crisis intervention training programs. The Journal of the American Academy of Psychiatry and the Law, 47(4), 414–421.
Rohrer, A. J. (2021). Law enforcement and persons with mental illness: Responding responsibly. Journal of Police and Criminal Psychology, 36(2), 342–349.
Schuck, A. M., Rosenbaum, D. P., & Hawkins, D. F. (2008). The influence of race/ethnicity, social class, and neighborhood context on residents’ attitudes toward the police. Police Quarterly, 11(4), 496–519.
Shapiro, G. K., Cusi, A., Kirst, M., O’Campo, P., Nakhost, A., & Stergiopoulos, V. (2015). Co-responding police-mental health programs: A review. Administration and Policy in Mental Health and Mental Health Services Research, 42, 606–620.
Shefner, R. T., Koppel, R., Wood, J., & Anderson, E. D. (2023). Co-deployment is an answer, but what are the questions? Insights from officer focus groups in Philadelphia. Police Practice and Research, 1–7.
Taheri, S. A. (2016). Do crisis intervention teams reduce arrests and improve officer safety? A systematic review and meta-analysis. Criminal Justice Policy Review, 27(1), 76–96.
Vermeer, M. J. D., Woods, D., & Jackson, B. A. (2020). Would law enforcement leaders support defunding the police? Probably—If communities ask police to solve fewer problems. JSTOR.
Walker, E. F., & Diforio, D. (1997). Schizophrenia: A neural diathesis-stress model. Psychological Review, 104(4), 667.
Watson, A. C., Compton, M. T., & Draine, J. N. (2017). The crisis intervention team (CIT) model: An evidence-based policing practice? Behavioral Sciences & the Law, 35(5–6), 431–441. https://doi.org/10.1002/bsl.2304
Watson, A. C., Compton, M. T., & Pope, L. G. (2019). Crisis response services for people with mental illnesses or intellectual and developmental disabilities: A review of the literature on police-based and other first response models. Vera Institute of Justice.
Watson, A. C., Pope, L. G., & Compton, M. T. (2021). Police reform from the perspective of mental health services and professionals: Our role in social change. Psychiatric Services, 72(9), 1085–1087.
Whaley, A. L. (1997). Ethnic and racial differences in perceptions of dangerousness of persons with mental illness. Psychiatric Services (Washington, DC), 48(10), 1328–1330.
Wood, J. D., & Anderson, E. (2023). Triaging Mental Health Emergencies: Lessons From Philadelphia. Law and Contemporary Problems, 86(1), 29–53.
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/. For commercial re-use, please contact sales@sgpublishing.ca.
Journal of CSWB, VOLUME 8, NUMBER 3, September 2023