Response,Use of Force,and Dispositional Outcomes: An Exploratory Analysis of the Seattle Police Department’s Crisis Intervention Training on Confirmed Crisis Events

Abstract

Considering the increased responsibility placed on law enforcement for handling crisis encounters, many departments have implemented crisis intervention training to reduce use of force through de-escalation and increase diversion/referral to community services. Despite the wellintentioned nature of such programming, evidence to date has been both elusive and inconclusive. The purpose of the current study is to investigate the effectiveness of the Seattle Police Department’s crisis intervention program in meeting these intended goals. Utilizing a covariate matching approach, the research questions compare confirmed crisis event responses with and without a crisis-certified officer to assess potential differences in crisis event response profiles, police use of force during crisis encounters, and dispositional outcomes resulting from crisis encounters. The results highlight differences in crisis response likelihood, use of force, and dispositional outcomes between CIT-certified and noncertified officers. Implications for future research, practice, and policy are discussed.

Keywords

crisis encounters crisis intervention teams police use of force mental health diversion

Police increasingly handle mental/behavioral health crises due in large part to deinstitutionalization and limited community services (Jachimowski & Cooper, 2020). As such, officers are often termed “street corner psychiatrists” amid rising encounters with persons with mental illness (PMI) (Koziarski et al., 2022; Livingston, 2016). Although mental illness is often a predominant facilitator or exacerbator of crisis-related calls for service, the scope of crisis encounters expands to other often co-occurring behavioral/situational crises including: substance abuse, risks of suicide, domestic cases, homelessness, intellectual/developmental disorders, and various other instances of (non)serious crime/violations (Balfour & Zeller, 2023; Watson et al., 2019). Without specialized training, these encounters more frequently involve use of force or arrest, with PMI disproportionately affected (Laniyonu & Goff, 2021; Morabito et al., 2017; Rohrer, 2021; Rossler & Terrill, 2017).

Since interactions with police use of force increase the risk of injury or death for both law enforcement and involved citizens, addressing the causal mechanisms behind crisis events and law enforcement responses is paramount. The implementation of crisis intervention teams (CIT), originating in Memphis in 1988 and colloquially termed the “Memphis Model,” came after a fatal shooting involving an individual experiencing a mental health crisis. This model and other recent CIT programs were developed to provide specialized training, tailored dispatch, and prioritization for crisis calls (Compton et al., 2008; Rohrer, 2021; Usher et al., 2019; Watson et al., 2021). CIT programs aim to de-escalate safely, reduce use of force and arrests where appropriate, and divert PMI to services rather than punitive sanctions. Initial evaluations of CIT programs have shown improved officer preparedness, knowledge, and attitudes (Compton et al., 2008; Pelfrey & Young, 2020; Watson & Compton, 2019), but evidence on force/arrest reductions is more mixed and inconclusive (McLean et al., 2022; Rogers et al., 2019; Taheri, 2016; Todd & Chauhan, 2021; Watson & Compton, 2019).

Considering that CIT were implemented with the primary objectives of better equipping and deploying patrol to crisis response, using less force, and increasing referrals and decreasing arrests (Morabito et al., 2017), the lack of conclusive evidence on the effectiveness of CIT programming represents a significant gap within the literature. If CIT is to continue to be implemented across various police jurisdictions, it is crucial to obtain more robust and conclusive empirical evidence as to whether there is adequate potential to reduce police use of force during crisis encounters through CIT informed approaches. Furthermore, the ways in which CIT is utilized for crisis situations, especially pertaining to specific types of crisis events and strategies for CIT officer response, is seldom explored. Therefore, the goal of the current study is to execute an exploratory analysis of crisis encounters within the Seattle Police Department to answer the following research questions:

(1) Are there differences in the type of crisis encounters that receive a CIT-certified officer response first compared to crisis events without a CIT-certified officer as first response?¹

(2) Does the presence/absence of CIT officers during crisis encounters affect the frequency in which force is utilized by police officers?

(3) How does the presence/absence of CIT officers during crisis encounters affect the outcomes (e.g., diversion to services versus arrests) of such encounters?

In addition to better understanding crisis event responses, use of force frequency, and dispositional outcomes, the proposed findings have implications for law enforcement policies related to future CIT programming implementation. Results can inform response strategies by improving how crisis events are identified and prioritized for CIT response and provide insight in reducing critical outcomes like use of force and the presumed reliance on arrest dispositions. Together, these improvements could help increase perceptions of procedural justice and improve relationships between the police and community (Merenda et al., 2021).

Literature Review

Crisis Event Profiles and Deployment Strategies

Crisis encounters often involve PMI but extend to co-occurring issues (i.e. substance abuse, homelessness, etc.), complicating responses (Lamb & Weinberger, 2020; Watson et al., 2021). Prevalence estimates vary (∼1% of calls, which is likely underestimated) due to underreporting and/or misclassification (Koziarski et al., 2022; Livingston, 2016). As ‘responders of last resort,’ police are uniquely positioned to influence subsequent outcomes of crisis encounters for PMI (Segal et al., 2018). However, crisis encounters often expand to other forms of “crisis” beyond those that exclusively pertain to mental illness or psychiatric disorders, increasing the contextual complexity of each encounter. To fill this gap, police require additional skills and training to avoid unnecessary negative outcomes like escalated use of force, added criminal charges, and threats to public and personal safety.

Different classifications of crisis encounters and related calls for service may be associated with underlying mental/psychiatric health needs, yet each varies in terms of nature, scope, and severity. For example, research on crisis-related calls for service indicates that “crisis” calls can range from incidents involving individuals in diagnosed mental/behavioral crisis who do not require police assistance due to a lack of criminal/law-violating behavior to those who may pose a substantial danger to themselves or others (Hendy et al., 2022). By extension, it is known that PMI often suffer from co-occurring disorders such as substance (ab)use (Iqbal et al., 2019) and other situational factors such as homelessness, veteran status, joblessness, low socioeconomic status, and family/domestic disturbances (Draine et al., 2002); the combination of which further increases the likelihood for repeated encounters with law enforcement (Willis et al., 2021).

Even with chronic underreporting and/or misclassification, crisis encounters represent a significant problem for law enforcement officers and PMI, both historically and in the present (Tucker et al., 2008). In a review of one of the first extensive evaluations of crisis encounters conducted by the American Bar Foundation (Brakel & South, 1968), Matthews (1970) concluded that PMI experiencing crises represents a medical problem; one that law enforcement officers are ill-equipped to handle due to a lack of training and resources. This trend has continued in recent history, with many law enforcement officers reporting high levels of frustration in their encounters with PMI, manifesting from a lack of training and the unavailability of community mental health resources (Wells & Schafer, 2006). In addition to the strain on law enforcement officers, individuals experiencing crises tend to suffer secondary negative consequences during encounters with the police, most commonly with increased use of force and probability of arrest. Other contextual factors found to increase the likelihood of police use of force in crisis-related calls for service include subject suicidality, family/domestic disturbances, presence of firearms or other weapons, substance-related impairment, threats of violence toward others, mental health treatment non-adherence, and apparent symptoms of psychosis (Wood et al., 2024).

Police Use of Force

Systematic reviews of the literature pertaining to police use of force suggest that it is an elusive concept, which manifests in differential conceptualizations and operationalizations across both law enforcement and research alike (Di Nota et al., 2021; Hollis, 2018; Klahm et al., 2014; Terrill & Paoline, 2013). Further complicating matters is the thin line of discretion separating use of force as a lawful extralegal power and official mandate of police (Bittner, 1970), with the unauthorized exercise of this same force being conceptually defined as unlawful or violent. This narrow conceptual difference has led to a more limited definition of police use of force that primarily focuses on instances in which police engage in “physical” acts of force. This general definition of force is also importantly distinct from other physical force actions that are perceived as overly severe and/or fatal (Klahm et al., 2014); which may be due to the increased visibility of (and subsequent public concern for) violent use of force incidents that have been shared or publicized in the news or social media (Hollis, 2018). Klahm and colleagues (2014) also argue that these restricted definitions fail to consider nonviolent coercive behaviors commonly utilized by police officers to achieve law enforcement goals. Thus, debates around use of force definitions hinge on outlining within-policy use of force, excessive force, and or violent/illegal forms of force (Adams & Alpert, 2023).

The nuance surrounding definitional ambiguity for the use of force is not new. In fact, over 50 years ago Bittner (1967) studied factors that influence how police assess use of force in emergency apprehensions of “mentally ill persons” citing a formal recognition of the risk to life, physical health, property and order in public places, people, and suspect (p.279). Although Bittner (1967) offers a myriad of response options for police officers when encountering what we would term today as encounters with PMI, he admits that this “psychiatric first aid” is not a formal strategy and often requires the more serious use of force to diffuse the situation (p.288). Since then, the subjectivity associated with the meaning and application of police use of force (e.g., soft/hard, justified/unjustified, and lethal/non-lethal force, verbal commands/physical detaining methods) has led to substantial barriers in estimating accurate recording of use of force incidents (Garner et al., 2018). Nonetheless, Garner and colleagues (1995) are credited for proposing one of the first formal definitions for police use of force; a definition which is continuously used in the present. According to Garner and colleagues (1995), police use of force can be summarily defined as a law enforcement officer’s intentional use of threats, attempts, or purposeful infliction of physical harm onto any member of the community.

Despite the challenges presented in defining police use of force, progress is being made in more accurately determining force prevalence rates (Garner et al., 2018). Recent research has been motivated by public concern pertaining to the frequency of publicized police use of force incidents, especially against vulnerable populations such as those with mental illness (MacLean et al., 2022). Again, prevalence approximations of police use of force against PMI and other individuals in crisis vary, but it has been empirically demonstrated that these individuals are overrepresented as victims of police use of force compared to the general population (Laniyonu & Goff, 2021; Morabito et al., 2017; Rohrer, 2021; Rossler & Terrill, 2017). For example, some estimates indicate that PMI are approximately twelve times more likely to be subjected to police use of force than individuals without mental illness, and approximately ten times more likely to suffer physical injuries because of police use of force (Laniyonu & Goff, 2021). In addition, research suggests that PMI are more likely to experience higher levels (or more severe) of force compared to individuals without mental illness (Rossler & Terrill, 2017).

To reduce police use of force against PMI and other individuals in crisis, several reform efforts have been made regarding policing practices (Rohrer, 2021). Some examples of such policing reforms include the establishment of community-oriented policing approaches, police partnerships with mental health professionals/agencies, and crisis intervention teams/training (Rohrer, 2021). Although each of these reformed approaches to handling crisis encounters have been explored by various law enforcement agencies, crisis intervention teams and training are regarded as the most popular approach to addressing the issue of police use of force during crisis events (Watson & Compton, 2019).

Crisis Intervention Teams and Associated Training

Crisis intervention teams (CIT) refer to programs in which law enforcement officers that have received specialized mental health training are dispatched (by emergency call-takers who are also CIT-trained) to handle crisis contacts and calls for service involving PMI and other individuals experiencing crisis (City of Memphis Police Services, n.d). The original Memphis model of CIT is rooted in substantial mental health training for officers (40 hr total) who volunteer to participate in the program, as well as the use of an established mental health facility located within the community, which must accept all patients diverted by law enforcement (Rogers et al., 2019). Despite variations in CIT implementation and procedures across practicing jurisdictions, it generally represents an interdisciplinary approach surrounding police responses to individuals experiencing crisis in which officers make use of specialized mental health and crisis resolution training to de-escalate crisis situations while simultaneously linking PMI to community mental health services (Pelfrey & Young, 2020). Some of the overall objectives of CIT include improving safety for both subject and officer during crisis encounters, developing more appropriate police responses to individuals in crisis/reducing police use of force, and increasing diversion of PMI to community mental health treatment as opposed to criminal justice system involvement (Watson & Compton, 2019).

To encourage these outcomes, CIT training emphasizes the use of de-escalation techniques, in which CIT officers attempt to peacefully resolve crisis encounters without having to resort to the use of physical force (Watson & Compton, 2019). In a recent scoping review investigating the effectiveness of de-escalation by police officers, van Lith and colleagues (2025) found that measures of effectiveness for de-escalation techniques vary. More specifically, the authors found that “de-escalation techniques encompass officers’ skills and knowledge in communication style, exchanging information, empathy and understanding, self-regulation….and positioning and safety” (p.638). Thus, measuring effectiveness for de-escalation training in officer-citizen encounters is often a product of both training and officer personality, experience, and prior skills and training. In summary, the CIT model represents a collaborative approach for the handling of crisis-related calls for service in which crisis events are dispatched to officers with specialized resolution and de-escalation training, with the intended outcomes consisting of improved safety for both officers and subjects through a reduction in police use of force, as well as diversion of subjects away from further criminal justice system involvement and toward community-based treatment programming.

Effectiveness of CIT Programs

Considering that CIT (or at least some variation of them) are widely employed by police departments across the country (Watson & Compton, 2019), it is necessary to empirically evaluate whether CIT are effective in achieving established goals. To date, most empirical evaluations on the effectiveness of CIT have focused on officer-level outcomes (Pelfrey & Young, 2020). Research suggests that the CIT approach is effective in increasing officer feelings of preparedness for crisis encounters, improving officer knowledge surrounding PMI in crisis and the symptoms they may exhibit, and enhancing officer attitudes (e.g., less stigmatization) toward PMI (Compton et al., 2008; Pelfrey & Young, 2020; Watson & Compton, 2019). Improved officer-level outcomes appear to be the most salient finding within the pre-existing literature (Rogers et al., 2019), however, other objectives and outcomes also warrant continued inquiry when evaluating the effectiveness of CIT.

Prior research is inconclusive in whether CIT implementation and utilization effectively reduce the frequency and levels of police use of force during crisis events (Rogers et al., 2019; Watson & Compton, 2019). Very few empirical examinations of the CIT approach have been able to demonstrate reductions in police use of force. In a 2014 comparison of CIT-trained officers versus non-CIT-trained officers, Compton and colleagues(2014) found that CIT-trained officers exhibit lower levels (or less severe forms of) force than those who are not trained in CIT, but there were no differences or emergent patterns in overall use of force across the two samples. More recently, Willis and colleagues (2023) found similar evidence that CIT-trained officers use force against PMI at a lower rate than officers without CIT training. However, a larger body of empirical investigations into the effects of CIT on police use of force have produced mixed findings regarding both the level and frequency of force used by CIT officers, highlighting a substantial gap in the literature surrounding the efficacy of CIT and police use of force (McLean et al., 2022; Rogers et al., 2019; Taheri, 2016; Watson & Compton, 2019).

Another goal of CIT is to divert individuals in crisis to community mental health services rather than the criminal justice system. In the last decade, evaluation studies examining the success of diversion in CIT have been less than conclusive, suggesting mixed results. Some research has found, not surprisingly, that CIT implementation effectively increases instances of individuals in crisis being linked to community mental health treatment (Compton et al., 2014). More surprisingly however, this tends to occur only within jurisdictions with streamlined mental health intake facilities maintaining a no-refusal policy for patients dropped off by law enforcement officers (Rogers et al., 2019). While the existence of such a facility may be feasible in large urban areas, it is less feasible (and indeed rarer) in rural communities (Pelfrey & Young, 2020; Rogers et al., 2019). The goal of diversion may also be examined through analyses of arrest rates. Initial CIT research has found limited reductions in arrests of PMI and other individuals in crisis (Taheri, 2016), suggesting that diversionary community mental health services may not be utilized to their fullest potential (Foss, 2019).

Although progress has been made in evaluating the effectiveness of CIT programs, it is also clear that the results of such evaluations remain mixed and inconclusive. These gaps are especially notable across outcomes pertaining to CIT officer deployment, use of force, and final dispositions for crisis events (Watson & Compton, 2019). Several factors inhibit rigorous evaluation of CIT programs, including small sample sizes (due to the relative infrequency of police use of force and subsequent deaths compared to general trends in police activities); variations in the implementation of CIT programming across departments/jurisdictions; variations in the definitions, measurements, and documentation of incidents involving police use of force; limited research consisting of randomized controlled trials and case-control designs; and overreliance on self-report and qualitative data (Cross et al., 2014; Rogers et al., 2019; Watson & Compton, 2019).

The Current Study

The current study seeks to explore three critical gaps in CIT research. First, significant variations in CIT implementation across departments and jurisdictions (Pelfrey & Young, 2020) suggests the need to better understand whether discernable “profiles” and “strategies” exist for crisis events that garner CIT responses. Second, inconclusive findings pertaining to the effects of CIT on reducing police use of force against individuals in crisis (McLean et al., 2022; Rogers et al., 2019; Taheri, 2016; Watson & Compton, 2019) are indicative of the need for more research designed to test whether CIT alters the likelihood of force being used during crisis events. Third, mixed findings surrounding the impacts of CIT on arrest rates (Foss, 2019; Taheri, 2016; Todd & Chauhan, 2021) are indicative of another gap in the literature regarding whether CIT effectively improves diversion of individuals in crisis to community treatment services. Considering that both diversion to community treatment programming and reducing police use of force against individuals in crisis are cited as primary goals of CIT (Usher et al., 2019), a lack of empirical research to support each of these aspects is a cause for concern; especially since CIT is already widely used by police departments across the country (Watson & Compton, 2019). Therefore, the current study will seek to address the gaps in the literature through an exploratory and descriptive examination of the following research questions:

(1) Are there differences in the type of crisis encounters that receive a CIT-certified officer response first compared to crisis events without a CIT-certified officer as first response?

(2) Does the presence/absence of CIT officers during crisis encounters affect the frequency in which force is utilized by police officers?

(3) How does the presence/absence of CIT officers during crisis encounters affect the outcomes (e.g., diversion to services versus arrests) of such encounters?

The current study uses a matched case-control design to determine whether Seattle’s CIT model differentially deploys CIT certified officers, is effective in reducing police use of force during crisis encounters, and effectively increases diversionary outcomes as opposed to arrest. The matching design uses a propensity score matching technique to compare crisis events with and without CIT-certified officers across deployment strategy, use of force, and dispositional outcome. The findings and contributions from the exploratory study are best positioned to indicate if there is an observable relationship between CIT officer deployment and presence on several unique outcomes of interest.

Methodology

Study Setting

The current study analyzes crisis encounters included in the Seattle Police Department’s (SPD) crisis contact dataset (City of Seattle, 2017). The sample of crisis contacts examined in the current analysis occur in the SPD’s five individually delineated precincts (North, East, South, West, and Southwest) serviced by the department. Seattle is the 15th most populated city in the United States and has experienced steady growth of approximately 20% in the last decade according to Census Bureau estimates. Recent years have seen the city embroiled in several police-community controversies including protest events following the George Floyd killing in 2020 that led to continued unrest, police and governmental turnover, and elevated crime levels (Piza & Connealy, 2022). Despite these challenges, Seattle has been notably touted as a progressive law enforcement entity in terms of modern responses like CIT (Rosenthal, 2021).

Seattle Police Department CIT

According to Seattle Police (2018), SPD’s approach to CIT can be characterized as a voluntary program in which SPD officers agree to participate in the CIT certification process. The Seattle Police Department (n.d.) asserts that at least 60% of SPD officers are currently CIT-certified. The CIT certification process entails the completion of 40 hrs of crisis intervention and de-escalation training, followed by eight hours of in-service training on an annual basis (Seattle Police, 2018). Examples of topics covered during training include a synopsis of mental health disorders, approaches to identifying symptoms of mental illness, effective communication with individuals experiencing behavioral health crises, and de-escalation techniques (Seattle Police Department, n.d). Upon the completion of training, CIT-certified SPD officers are prioritized for dispatch to crisis contacts/encounters, during which they seek to de-escalate the situation and ideally divert/refer subjects to the appropriate community mental health/treatment services whenever possible (Seattle Police, 2018). In addition to the voluntary CIT certification process, all SPD officers are required to receive eight hours of standard CIT training (albeit less extensive than the 40-hour training for certified officers). For the current study, we will refer to SPD officers who have completed the voluntary 40-hour CIT-certification process as “CIT-certified officers,” and officers who only received the mandated eight-hour standard CIT training as “non-certified officers.” Per department policy, all crisis contacts/encounters must be documented and reported to SPD’s Crisis Response Unit for review and data analysis (Seattle Police Department, n.d).

Data Sources and Variables

In addition to the documentation and reporting requirements for all crisis encounters, SPD provides public access to its crisis contact data (City of Seattle, 2017). These practices provide justification for the current study to utilize SPD’s crisis contact dataset as a purposive sample of SPD crisis contacts. The dataset contains quantitative values for several variables, including the date/time of the call/incident, type of call, disposition of the encounter, whether or not force was used during the encounter, whether or not a “non-certified officer” (receiving only the standard eight hours of CIT training) was requested/dispatched/arrived at the incident or a “CIT-certified officer” (receiving the voluntary 40 hr of training), responding officer information (e.g., gender, race, year of birth, years of experience), veteran status of the party in crisis as gathered, and geographical location of the incident (e.g., which precinct the incident occurred within).

For the purpose of the current study, the independent variables of interest refer to the presence/absence of a “CIT-certified officer” (operationalized as any SPD officer who has completed the eight hours of mandatory annual in-service training and the voluntary 40 hr CIT specific training program), and the presence/absence of a “non-certified officer” (operationalized as sworn SPD officers who have received only the mandated eight hours of standard CIT training) during each crisis encounter/contact included within the dataset. There are three dependent variables of interest including CIT response, use of force, and dispositional analyses. First, the response dependent variable dichotomously measures whether a CIT-certified officer arrives on scene or not for each crisis event included in the dataset. This variable can be leveraged to indicate if certain predictors or case “profiles” reflect a differential strategy in CIT-certified response. Second, the use of force dependent variable dichotomously measures whether police use of force occurred during the event as defined by the Seattle Police Department’s internal policy thresholds which largely center on the use of some form of physical force.² The analyses will test for potential differences in the likelihood of use of force occurring during crisis events across CIT-certified and non-certified responses. Lastly, the study operationalizes a dispositional variable to identify if there are differences in non-arrest outcomes between CIT-certified and non-certified officers following crisis events. A total of 32 unique dispositional outcomes were included within the initial dataset, with each of the individual dispositions being assigned into one of four thematic categories of final outcomes (arrest, emergent detention, referral, and no action).³

Other potentially influential confounding variables were also collected and added to the models as matching covariates. Relevant indicators include the following: age of responding officer (based on year of birth); gender of responding officer (male, female); race/ethnicity of responding officer (African American, American Indian/Alaska Native, Asian, Hispanic/Latino, Native Hawaiian/Other Pacific Islander, White, Two or More Races, Unknown); responding officer years of experience; and geographical location (North, East, South, West, Southwest precincts). The precinct variable was incorporated to compare crisis events within the same geographical and jurisdictional setting to account for potential differences in other sociodemographic indicators across precincts and potential differences in law enforcement leadership and operations within each unique precinct.

Sample

SPD’s crisis contact dataset was initially created and published in October of 2017, and is updated daily (City of Seattle, 2017). At the date/time of download on April 11th, 2025, the full dataset consisted of 98,099 crisis events occurring between the years 2017 and year-to-date 2025. The dataset was prepared around the identified variables of interest producing a dataset with 93,080 crisis events. Crisis events with incomplete data for key variables were excluded from the subsequent analyses. Table 1 below includes key descriptives for the independent, dependent, and confounders of interest operationalized in the present study. The day variable is binary and refers to calls taking place between 7am–7pm and the veteran status variable indicates if the person in crisis had a reported veteran designation. The balance between crisis events responded to by CIT-certified and non-certified officers is relatively even. Though, the use of force data is highly skewed with few force events in the subsequent analysis. The dispositional variable will operationalize arrests as the baseline for all relevant comparisons, even though arrests only reflect about 8% of the overall sample.

Table 1.

Crisis Data Variables of Interest Descriptive Statistics

Independent variable
	Yes	No
CIT-certified	45.41%	54.59%
Dependent variables
	Yes	No
Deployment	45.44%	54.56%
Use of force	1.63%	98.37%
	Arrest	Emergent det.	Referral	No action
Disposition	7.79%	34.64%	33.05%	24.525%
Covariates
	Yes	No
911 origin	74.56%	25.44%
	Yes	No
Day time	51.43%	48.57%
	Admin	Violent/Priority	Assignments	Disorder	Kids	Ped./Traffic	Warrant	Property
Initial call	2.89%	43.88%	2.90%	44.72%	0.48%	0.42%	0.13%	4.58%
	Unknown	East	North	West	South	Southwest
Precinct	4.46%	19.64%	26.82%	28.29%	10.88%	9.91%
Veteran	Yes	No
	1.65%	98.35%

Note. the covariate statistics were reported for the deployment and use of force models (n = 93,080). The disposition model had (n = 89,271) some excluded events without final dispositions recorded but it did not have a substantial impact on variable inclusion.

Analytical Approach

The analytical approach of the current study was research question specific. The first research question used logistic regression to examine factors associated with the response of a Crisis Intervention Team (CIT)-certified officer during crisis events in Seattle. The binary dependent variable indicated whether a CIT-certified officer was present at the scene, and the model incorporated relevant predictor variables that may influence the marked arrival of a CIT-certified officer compared to a non-certified officer including whether the call originated from 911 or was proactively initiated, whether the incident occurred during day or night time hours, the precinct of the event, a categorical indicator of the initial call type sequenced into categories like violent, property, disorder, admin, warrant, etc., and whether the subject was identified as a veteran. We computed clustered standard errors, calculated marginal effects, and predicted probabilities across each variable included to identify any factors that may influence the likelihood of a CIT-certified officer arriving on scene to determine if there is a case-profile or strategy towards the deployment of CIT-certified officers for crisis events. The presumption is that there may be certain call types, dispatcher strategies or intuitions, or departmental protocols that prioritize CIT-certified officer responses. The analyses are positioned to explore this possibility correlationally, though, the data do not readily permit the identification of officer self-selection into crisis related calls through proactive activities or dispatch responses.

The second research question tests if a CIT-certified officer response to a crisis event influences the likelihood of law enforcement use of force. The binary outcome of interest was whether force was used, and the key predictor was whether the first responding officer was CIT-certified or not. The dataset only captures information for the first officer, meaning the analyses cannot account for the presence of multiple officers or potential instances in which officers have different CIT-certification levels. Due to this data limitation, and because officers are not randomly assigned to calls and CIT-certified officers may be systematically deployed to certain call types or precincts, we used propensity score methods to adjust for potential selection biases. Covariates used to estimate the probability of CIT-certification included precinct number, officer gender, officer race, officer years of experience, call type, veteran status, and time of the event notated as day or night.⁴ Using these variables, we estimated propensity scores and applied inverse probability of treatment weighting (IPTW) to create a balanced, pseudo-randomized comparison between CIT-certified and non-certified officers. Covariate balance diagnostics confirmed successful reweighting across all key variables.⁵ Overall, the propensity score distribution suggests the data meet the necessary assumptions for covariate-adjusted comparisons between CIT-certified and non-certified officers pertaining to use of force in crisis encounter responses. Logistic regression was used to indicate differences in CIT-certified and non-certified use of force likelihood between the two officer groupings. The exploratory findings are positioned to highlight if CIT-certification alters the likelihood of a use of force event after matching on several covariate crisis event predictors.

The third research question analyzes whether CIT-certified officers differ from non-certified officers in the final disposition following a crisis encounter. Beyond the potential for use of force during a crisis event, officers must make a final decision surrounding what to do when responding to a crisis event. The dependent variable was the outcome disposition derived from the SPD crisis dataset. The disposition variable was recoded into a categorical measure representing four meaningful outcomes: (1) arrest, (2) emergent detention, (3) referral, and (4) no action. Dispositions coded as unknown or other were excluded from the analysis. The key independent variable was whether the responding officer was CIT-certified or not. Given concerns that CIT-certified officers may be systematically assigned to different call types, precincts, or shifts, we again employed inverse probability of treatment weighting (IPTW) to adjust for confounding due to observable differences prior to estimating the multinomial logistic regression model. Propensity scores were estimated using logistic regression, modeling the likelihood of CIT certification as a function of key covariates: precinct, officer gender, officer race, officer years of experience, initial call type, veteran status, and day/night indicator. These scores were then used to generate IPTW weights, which were applied to all subsequent models. Post-weighting covariate balance was assessed using propensity scores, and results confirmed that treated and control groups were adequately balanced on the measured covariates.⁶ Multinomial logistic regression was used with robust standard errors to see if there were differences in crisis event outcomes across each disposition type relative to arrest between CIT-certified and non-certified officers.

Results

Research Question 1: CIT Deployment Profiles and Strategies

The logistic regression model demonstrated statistically significant effects for multiple predictors of CIT-certified officer response as detailed in Table 2 below. Nighttime crisis events were about 37% less likely to involve a responding CIT-certified officer (p < .001) and crisis events originating from 911 calls were associated with about 12% reduced odds of CIT deployment (p < .001) compared to non-CIT responses. Substantial geographic variation also emerged relative to the baseline East Precinct, with every precinct demonstrating significantly higher odds of CIT-certified officer deployment to crisis events compared to non-certified officer deployment in the East Precinct (even in instances where the precinct was marked as “unknown”). Regarding call types, two categories, disorder and pedestrian/traffic calls (relative to the baseline violent call types), were associated with increased odds of CIT-certified officer response (OR = 1.08/1.46, p < .001) compared to non-CIT response, while most others did not differ significantly from the reference.

Table 2.

CIT Deployment Logistic Regression

CIT certified	Odds ratio	SE	p	95% LL CI	9% UP CI
911 call (vs. Proactive)	0.876	0.014	0.000	0.849	0.903
Day (vs. Night)	0.629	0.008	0.000	0.612	0.646
Initial call type (baseline: Violent)
Admin	1.090	0.044	0.034	1.006	1.180
Assignments	1.002	0.041	0.969	0.924	1.085
Disorder	1.081	0.015	0.000	1.051	1.112
Kids	0.893	0.088	0.251	0.737	1.083
Pedestrian/Traffic	1.462	0.151	0.000	1.194	1.789
Warrant	1.246	0.234	0.241	0.863	1.799
Property	1.047	0.034	0.163	0.982	1.117
Precinct (baseline: East)
Unknown	2.136	0.075	0.000	1.994	2.288
North	1.665	0.033	0.000	1.600	1.732
West	1.175	0.024	0.000	1.130	1.222
South	1.981	0.050	0.000	1.885	2.082
Southwest	2.507	0.066	0.000	2.381	2.640
Veteran status	1.077	0.057	0.156	0.972	1.194
Constant	0.749	0.017	0.000	0.717	0.782
	N	Chi²	p	Pseudo R²
	93,380	3342.19	0.000	0.026

Marginal effects analyses were also undertaken and confirmed that the precinct and the time of day were among the strongest predictors of a crisis event receiving (or not receiving) CIT, with CIT response probability reduced by approximately 11 percentage points during nighttime and increased by as much as 22 percentage points in the Southwest precinct (with precincts varying from about four percentage points to the maximum of 22 points). The predicted probabilities further revealed that veteran status was associated with a modestly higher probability of CIT response, particularly during daytime hours, though this association did not reach conventional significance thresholds in the regression model (p = .15). Table 3 includes the results of the marginal effect analyses.

Table 3.

CIT Deployment Marginal Effect Analysis

	dy/dx	SE	p	95% LL CI	9% UP CI
911 call (vs. Proactive)	−0.032	0.004	0.000	−0.039	−0.025
Day (vs. Night)	−0.112	0.003	0.000	−0.118	−0.106
Initial call type (baseline: Violent)
Admin	0.021	0.010	0.035	0.001	0.040
Assignments	0.000	0.010	0.969	−0.019	0.019
Disorder	0.019	0.003	0.000	0.012	0.025
Kids	−0.027	0.023	0.248	−0.072	0.019
Pedestrian/Traffic	0.091	0.025	0.000	0.043	0.140
Warrant	0.053	0.045	0.243	−0.036	0.141
Property	0.011	0.008	0.164	−0.004	0.026
Precinct (baseline: East)
Unknown	0.183	0.008	0.000	0.167	0.200
North	0.122	0.005	0.000	0.113	0.131
West	0.038	0.005	0.000	0.028	0.047
South	0.165	0.006	0.000	0.153	0.177
Southwest	0.222	0.006	0.000	0.210	0.234
Veteran status	0.018	0.013	0.156	−0.007	0.043

Note. dy/dx for factor levels is the discrete change from the base level.

Research Question 2: CIT and Use of Force

Logistic regression was used to examine if use of force was more likely in CIT-certified or non-certified crisis event responses. Results (Table 4) show CIT-certified officers were associated with a 17% higher odds of force (coefficient = .154, p < .05), but the practical difference is small: predicted probabilities are 1.52% (non-certified) vs. 1.78% (CIT-certified), a .26 percentage point increase. Given the low overall force rates (∼1.6–1.8%), this may reflect dispatch selection (CIT officers prioritized for higher-risk profiles) rather than CIT causing more force. While this suggests CIT-certified officers are slightly more likely to use force, the difference is small in practical terms. The predicted probabilities calculated via marginal effects are also included in Table 4.⁷

Table 4.

CIT Use of Force Rare Events Logistic Regression

	Coefficient	SE	p	95% CI lower limit	95% CI upper limit
CIT-certified	0.154	0.059	0.008	0.040	0.270
Constant	−4.16	0.044	0.000	−4.25	−4.08
	dy/dx	SE	p	95% CI lower limit	95% CI upper limit
CIT certified	0.003	0.001	0.008	0.001	0.004
		N	Chi²	p	Pseudo R²
		93,080	6.93	0.006	0.001

Research Question 3: CIT and Dispositional Outcomes

Table 5 shows CIT-certified officers had lower relative risks of emergent detention (RRR = 0.829), referral (RRR = 0.862), and no action (RRR = 0.732) compared to arrest (all p < .001).

Table 5.

CIT Dispositional Outcome Multinomial Logistic Regression

	RRR	SE	95% LL CI	95% UL CI
Emergent detention	0.829	0.026	0.780	0.880
Referral	0.862	0.027	0.811	0.917
No action	0.732	0.023	0.688	0.780

Note. baseline is “arrest”.

Marginal effects (Table 6) reveal higher predicted arrest probability (8.7% vs. 7.2%) and referral (34.4% vs. 33.1%), with lower no action (22.4% vs. 25.4%) and similar emergent detention (∼34%) in the CIT vs. non-CIT groups. This suggests CIT-certification shifts toward formal dispositions (arrest/referral) over inaction, though arrests remain altogether rare (∼8% of events overall).

Table 6.

CIT Versus Non-CIT Dispositional Outcomes Marginal Effects

	Margin	SE	p	95% LL CI	95% UL CI
Arrests
Non-CIT	0.072	0.001	0.000	0.069	0.075
CIT	0.087	0.001	0.000	0.084	0.090
Emergent detention
Non-CIT	0.343	0.003	0.000	0.338	0.349
CIT	0.344	0.002	0.000	0.339	0.349
Referral
Non-CIT	0.330	0.003	0.000	0.325	0.337
CIT	0.344	0.002	0.000	0.340	0.349
No action
Non-CIT	0.254	0.002	0.000	0.249	0.259
CIT	0.225	0.002	0.000	0.220	0.229

Discussion/Conclusion

The results of the current study highlight several key findings pertaining to various aspects of crisis encounters documented by SPD. Crisis events show distinct profiles: CIT-certified officer responses are less likely at night (∼37% lower odds) and for 911-originated calls (∼12% lower) but are more likely in certain precincts (relative to the East precinct) and for disorder/pedestrian calls. These patterns suggest operational/dispatch factors (e.g., staffing limits, proactive vs. reactive calls, precinct variation) may influence or underly a strategy towards prioritizing CIT-certified officer response to crisis events rather than random deployment. This finding again critically highlights the important role dispatchers play in resource allocation and crisis response (Watson et al., 2021).

In addition to the observed differences in CIT-response to crisis events pertaining to time and operations, spatial disparities in CIT-certified crisis responses were also demonstrated through precinct variation. Geographic disparities by precinct suggest potential structural or resource-based inconsistencies in the implementation of crisis response strategies like CIT. Due to the voluntary nature of the CIT-certification process in Seattle, certain precincts may have more CIT buy in at key leadership levels, resulting in more formally trained officers and better CIT coverage in crisis response. SPD’s most recent crisis intervention program report (Seattle Police, 2019) indicates the lowest percentage of CIT-certified staff is in the East Precinct (approximately 61% of its officers); a statistic that likely contributes to current findings pertaining to significantly decreased odds of CIT-certified officer response to crisis events in the East Precinct relative to all other tested precinct defined geographies. Prior research has long established a link between leadership style and patrol attitudes (Engel & Worden, 2003) and enforcement activities including the use of force and arrest (Engel, 2000).

Two unique call types were also shown to have increased odds of a CIT-certified officer responding to the crisis event. Calls initially coded as disorder or pedestrian/traffic calls had significantly increased likelihoods of a CIT-certified officer responding. This may reflect a unique “crisis” case type, separate from criminogenic calls for service to things like property or violent crime, that is particularly adept for a CIT-certified response. This may be an artifact of coincidence, or a reflection of a potential underlying dispatch strategy to direct CIT resources towards these types of calls (both disorder and pedestrian/traffic calls are likely to involve an individual on the street, for example). There is empirical precedent to speculate on the potential for an underlying, even implicit or subconscious CIT response strategy, as prior research indicates that 911 call-takers and dispatchers are key components of many CIT-related training efforts (Hendy et al., 2022; Watson et al., 2021). These results underscore the importance of examining both organizational and situational dynamics when assessing both equity and effectiveness in crisis intervention policing. It is reasonable to assume dispatchers leverage available call information and knowledge of CIT programming towards crisis events, though this is a stated presumption of the present study. Despite our efforts to adjust for observable call characteristics, dispatch assignment constitutes a second layer of selection that cannot be fully accounted for in the data. Dispatchers rely on finer-grained information than is captured by administrative call categories, and CIT-certified officers may therefore be systematically assigned or self-select into to more complex or higher-risk incidents, which limits the interpretation of the observed differences being a function of CIT-certification alone. Though, it appears that there are underlying factors that may dictate the likelihood of a CIT-certified officer response to a crisis event.

The second research question focused on potential differences in the likelihood of use of force in CIT and non-certified crisis responses. CIT-certified officers showed modestly higher force likelihood, but the low base rates and potential dispatch strategy in deploying officers to crisis events, may suggest that CIT simply maintains (rather than reduces) the status quo on use of force. To this effect, future research should be strategically designed to explicitly measure if CIT-certified officers are indeed systematically assigned or self-select into more complex or higher-risk incidents. This would allow for more conclusive determinations to be made regarding whether the observed increased likelihood for the use of force on behalf of CIT-certified officers is mitigated by this proposed phenomenon of systematic assignment/self-selection or by CIT-certification alone, or if it is an actual artifact of CIT programming that lends itself towards the use of force. To date, prior research on CIT effectiveness towards reducing force remains mixed, with studies finding bi-directional effects of CIT programming (Compton et al., 2011; Morabito et al., 2017; Peterson & Densley, 2018; Willis et al., 2023).

The third research question examined potential differences in crisis event resolution with or without a CIT-certified officer as the first respondent. For dispositions, a CIT-certified officer response seemed to shift events toward formal outcomes (e.g., higher arrest/referral, lower no action), potentially indicating greater confidence in assessment or priority for action-oriented calls. The potential to increase referrals, even relative to arrest, is conceivable within the present study’s findings, though potential barriers may still inhibit some referral actions as prior research has suggested much of the referral process hinges on law enforcement having active community partners (Rogers et al., 2019). Certainly, this suggestion relates to another crucial implication for current and future practice pertaining to CIT implementation, as it is unclear whether SPD has an active partnership with streamlined mental health intake facilities maintaining a no-refusal policy for patients dropped off by law enforcement officers. It is also important to note, though, that the overall arrest rate for crisis events was altogether relatively low with only about 8% of all crisis events culminating in an arrest.

The results offer several exploratory insights on CIT. First, there may be an underlying dispatch strategy or subconscious officer patrol approach in the response to crisis events. The findings from the first research question suggest that there were several preliminary differences in the profile of crisis events handled by CIT-certified officers relative to non-CIT officers. Though, without more specific dispatch data and narratives, these results are correlational in suggesting a presumed relationship between dispatch decision-making and CIT response.

Second, the analyses found a modest increase in force in crisis events with a CIT-certified compared to non-certified officer. The effect was marginal, but there may be differences in dispatch strategy, officer approach to crises, or overall training that lead patrol officers to make discretionary decisions surrounding the use of force, which in many such instances is founded to be both appropriate and necessary. The data did not have the nuance to explore force levels or appropriateness, but preliminary indications suggest that CIT does not necessarily reduce the overall use of force. Therefore, it is critical for future research to develop (and subsequently utilize) more nuanced datasets with the capability of documenting and analyzing force levels/appropriateness to conclusively determine the extent to which CIT may effectively reduce force, or more simply align with more consistent and appropriate force applications.

Lastly, the dispositional analyses suggested that CIT-certified officers were simultaneously more likely to make arrests and more likely to use diversionary outcomes like referral. This may reflect that CIT-certified officers are simply more willing to take formal action relative to no action, and that they may be more confident in their assessment and decision-making surrounding a crisis event. Indeed, this assumption is consistent with prior findings of improved officer-level outcomes stemming from CIT training, such as increased feelings of preparedness and knowledge pertaining to PMI/individuals in crisis (Compton et al., 2008; Pelfrey & Young, 2020; Watson & Compton, 2019). Taken together, the study finds preliminary support that CIT-certified officers may be differentially dispatched or proactively selecting into crisis events, and they may consider both use of force and dispositional decisions differently, possibly because of CIT programming, relative to non-CIT-certified officers.

Limitations

While the above results are informative, some limitations of the current study are detailed below. First, the dataset was limited to only crisis encounters documented by SPD; this may notably miss unreported crisis events or events misclassified by the SPD. Therefore, the findings may not be wholly representative of crisis events occurring in Seattle nor be generalizable to other police departments across the United States. By extension, SPD’s implementation of CIT may also not be consistent with CIT implementation in other police departments. SPD’s requirement that all sworn police officers receive a minimum of eight hours of standard CIT training, separate from the voluntary CIT-certification process, is a unique adaptation of CIT programs, as not all police departments follow this model (Pelfrey & Young, 2020).

Another limitation stems from the way de-escalation effectiveness (and by way of that, reductions in use of force) are conceptualized and measured as outcomes of successful CIT training. We recognize that while officers who attend CIT training that emphasizes de-escalation techniques should be more effective and confident in appropriate use of force and diverting PMI to appropriate services, de-escalation success for individual officers is a combination of training as well as personal traits, experience, and prior skills/knowledge. For this fact, measuring de-escalation effectiveness can be somewhat elusive, although we believe the robustness of our sample and comparison of two training modalities (40 hr and 8 hr) provide empirical support to CIT programs being related to potential changes in de-escalation and use of force. The present study is exploratory in nature, and future studies should address this connection with more robust and experimental methods.

The results are also more exploratory and preliminary as a first attempt at examining the crisis data more rigorously. Early research on crisis events in Seattle focused on available data and dispositional decision-making (Todd & Chauhan, 2021) but did not do so through a comparative framework. The current study attempts to evaluate differences in CIT-certification and non-certification in response to crisis events in a case-controlled research design. The analytical approach involved balancing on available officer-traits and geographic confounders prior to comparing crisis event responses to increase the validity of the results. Though, the covariate matching approach was not without limitation and there was not nuanced enough data to cluster on unique officer identifiers, consider the level of force used, or the appropriateness of force used. The available data were limited in confounders, only identified the first responding officer which may introduce treatment misclassification or bias estimates towards the null if multiple officers (or officers with mixed CIT status) arrived, and the outcomes of interest often had low baseline occurrence rates. Results should be interpreted as preliminary and correlational and applied with caution.

Conclusion

This exploratory study offers an examination of CIT-certified officers’ response and behavior in crisis events. The findings reveal that CIT-certified officers were significantly more likely to respond to certain crisis incidents based on space, time, and operational capacities. The results further demonstrated that CIT-certified officers may use force and arrest more than non-CIT officers, but these effects were marginal when examining predicted probabilities and may be a byproduct of underlying dispatch and operational strategies to prioritize CIT-officers to confirmed or potential crisis events, officer self-selection into crisis events, or CIT programming more generally.

While the observational nature of the data and the lack of randomized assignment preclude causal claims, the results add a growing literature base on CIT programming (Boddy et al., 2024; Peterson & Densley, 2018). As jurisdictions across the country consider alternatives to traditional enforcement-centric approaches, the evidence presented here preliminarily supports the continued and targeted expansion of CIT-certification as a potentially promising harms reduction strategy in shifting police responses further away from force and custody and toward referral and care-based resolutions. As a next step, future research should advocate for data systems and research designs that allow researchers to better disentangle the effects of dispatch decision-making, officer self-selection, and on-scene dynamics. Leveraging resources like dispatch narratives, officer identifiers, and measures of force severity and appropriateness could help better isolate the effects of CIT programming from other competing factors on outcomes of interest. From a practice perspective, agencies may benefit from examining how CIT-certified officers are deployed across shifts, precincts, and call types to ensure alignment between training objectives and operational use, and agency uniformity across differing leadership levels and structures. At the policy level, these findings underscore the importance of integrating CIT training with broader crisis response infrastructure, including dispatcher training and community-based service capacity, rather than treating CIT-certification as a standalone solution for patrol officers exclusively. Though, continued, and more rigorous, research is necessary to better establish the link between CIT programming and its purported goals and effects.

Footnotes

ORCID iD

Nathan T. Connealy

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Notes

Author Biographies

Amanda L. Harrison is a master’s student at the University of Tampa. Her research interests lie at the intersection of policing, legal system processing, and policy reform. She has recently published and presented research in several academic forums including the Journal of White Collar and Corporate Crime.

Nathan T. Connealy is an Assistant Professor in the Department of Criminology and Criminal Justice at the University of Tampa. His research focuses on the spatial analysis of crime patterns, the intersection of policing, crime prevention, and technology, and quantitative research designs. His scholarship has recently been published in peer-reviewed outlets including Criminology, Justice Quarterly, Criminology & Public Policy, and Crime & Delinquency among others.

Brandon Dulisse is an Associate Professor and Director of the Criminology Institute for Research and Training at the University of Tampa. His work includes studies on institutional corrections, cybercrime, and criminal justice policy. Recently his work has been featured in the Journal of Criminal Justice, Journal of White Collar and Corporate Crime, and Evidence Base.

References

Adams

I. T.

Alpert

G. P.

(2023). Use of force in policing. Oxford Research Encyclopedia of Criminology and Criminal Justice.

Balfour

M. E.

Zeller

S. L.

(2023). Community-based crisis services, specialized crisis facilities, and partnerships with law enforcement. In Focus, 21(1), 18–27. https://doi.org/10.1176/appi.focus.20220074

Bittner

(1967). Police discretion in emergency apprehension of mentally ill persons. Social Problems, 14(3), 278–292. https://doi.org/10.1525/sp.1967.14.3.03a00040

Bittner

(1970). The functions of the police in modern society: A review of background factors, current practices, and possible role models.

Boddy

Athaide-Victor

Law

F. M.

Sharp

(2024). CIT Training Among Law Enforcement: Does It Work? Journal of Police and Criminal Psychology, 39(3), 486–498.

Brakel

S. J.

South

G. R.

(1968). Diversion from the criminal process in the rural community: Final report of the American bar foundation project on rural criminal justice. American Criminal Law Quarterly, 7(3), 122–173.

City of Memphis Police Services . (n.d.). Crisis intervention team (CIT). https://www.memphistn.gov/cit/

City of Seattle . (2017). Crisis data. https://www.seattle.gov/police

Compton

M. T.

Bahora

Watson

A. C.

Oliva

J. R.

(2008). A comprehensive review of extant research on crisis intervention team (CIT) programs. Journal of the American Academy of Psychiatry and the Law, 36(1), 47–55.

10.

Compton

M. T.

Bakeman

Broussard

Hankerson-Dyson

Husbands

Krishan

Stewart-Hutto

D’Orio

B. M.

Oliva

J. R.

Thompson

N. J.

Watson

A. C.

(2014). The police-based crisis intervention team (CIT) model: II. Effects on level of force and resolution, referral, and arrest. Psychiatric Services, 65(4), 523–529. https://doi.org/10.1176/appi.ps.201300108

11.

Compton

M. T.

Demir-Neubert

B. N.

Broussard

McGriff

J. A.

Morgan

Oliva

J. R.

(2011). Use of force preferences and perceived effectiveness of actions among crisis intervention team (CIT) police officers and non-CIT officers in an escalating psychiatric crisis involving a subject with schizophrenia. Schizophrenia Bulletin, 37(4), 737–745. https://doi.org/10.1093/schbul/sbp146

12.

Cross

A. B.

Mulvey

Schubert

C. A.

Griffin

P. A.

Filone

Winckworth-Prejsnar

DeMatteo

Heilbrun

(2014). An agenda for advancing research on crisis intervention teams for mental health emergencies. Psychiatric Services, 65(4), 530–536. https://doi.org/10.1176/appi.ps.201200566

13.

Di Nota

P. M.

Chan

J. F.

Huhta

J.-M.

Andersen

J. P.

(2021). Considering objective and subjective measures for police use of force evaluation. International Journal of Environmental Research and Public Health, 18(10), 5351–5363. https://doi.org/10.3390/ijerph18105351

14.

Draine

Salzer

M. S.

Culhane

D. P.

Hadley

T. R.

(2002). Role of social disadvantage in crime, joblessness, and homelessness among persons with serious mental illness. Psychiatric Services, 53(5), 565–573. https://doi.org/10.1176/appi.ps.53.5.565

15.

Engel

R. S.

(2000). The effects of supervisory styles on patrol officer behavior. Police Quarterly, 3(3), 262–293. https://doi.org/10.1177/1098611100003003003

16.

Engel

R. S.

Worden

R. E.

(2003). Police officers’ attitudes, behavior, and supervisory influences: An analysis of problem solving. Criminology, 41(1), 131–166. https://doi.org/10.1111/j.1745-9125.2003.tb00984.x

17.

Foss

R. W.

(2019). The impact of critical incident team training on arrest and diversion rates [ProQuest dissertations & Theses, Doctoral dissertation, Walden University].

18.

Garner

J. H.

Hickman

M. J.

Malega

R. W.

Maxwell

C. D.

(2018). Progress toward national estimates of police use of force. PLoS One, 13(2), 1–23. https://doi.org/10.1371/journal.pone.0192932

19.

Garner

J. H.

Schade

Hepburn

Buchanan

(1995). Measuring the continuum of force used by and against the police. Criminal Justice Review, 20(2), 146–168. https://doi.org/10.1177/073401689502000204

20.

Hendy

Newcombe

Walton

(2022). From the street corner to call Centre: Changes in the characteristics of mental-health related calls for service received by police. Policing and Society, 32(7), 862–877. https://doi.org/10.1080/10439463.2021.1988085

21.

Hollis

M. E.

(2018). Measurement issues in police use of force: A state-of-the-art review. International Journal, 41(6), 844–858. https://doi.org/10.1108/PIJPSM-11-2017-0137

22.

Iqbal

M. N.

Levin

C. J.

Levin

F. R.

(2019). Treatment for substance use disorder with co-occurring mental illness. In Focus, 17(2), 88–97. https://doi.org/10.1176/appi.focus.20180042

23.

Jachimowski

K. G.

Cooper

J. A.

(2020). Police response to mental health calls for service: Gatekeepers and street corner psychiatrists. Rowman & Littlefield Publishing Group.

24.

Klahm IV

C. F.

Frank

Liederbach

(2014). Understanding police use of force: Rethinking the link between conceptualization and measurement. Policing: An International Journal of Police Strategies & Management, 37(3), 558–578. https://doi.org/10.1108/PIJPSM-08-2013-0079

25.

Koziarski

Ferguson

Huey

(2022). Shedding light on the dark figure of police mental health calls for service. A Journal of Policy and Practice, 16(4), 696–706. https://doi.org/10.1093/police/paac006

26.

Lamb

R. H.

Weinberger

L. E.

(2020). Deinstitutionalization and other factors in the criminalization of persons with serious mental illness and how it is being addressed. CNS Spectrums, 25(2), 173–180. https://doi.org/10.1017/S1092852919001524

27.

Laniyonu

Goff

P. A.

(2021). Measuring disparities in police use of force and injury among persons with serious mental illness. BMC Psychiatry, 21(2), 1–8. https://doi.org/10.1186/s12888-021-03510-w

28.

Livingston

J. D.

(2016). Contact between police and people with mental disorders: A review of rates. Psychiatric Services, 67(8), 850–857. https://doi.org/10.1176/appi.ps.201500312

29.

Matthews

A. R.

(1970). Observation for police policy and procedures for emergency detention of the mentally ill. The Journal of Criminal Law, Criminology, and Police Science, 61(2), 283–295. https://doi.org/10.2307/1142223

30.

McLean

Stoughton

S. W.

Alpert

G. P.

(2022). Police uses of force in the USA: A wealth of theories and a lack of evidence. Cambridge Journal of Evidence-Based Policing, 6(3-4), 87–108. https://doi.org/10.1007/s41887-022-00078-7

31.

Merenda

Trent

Rinke

C. R.

(2021). Untangling the role of interactions in police satisfaction: Examining direct and indirect contacts with the police. The Police Journal, 94(4), 496–514. https://doi.org/10.1177/0032258x20960790

32.

Morabito

M. S.

Socia

Wik

Fisher

W. H.

(2017). The nature and extent of police use of force in encounters with people with behavioral health disorders. International Journal of Law and Psychiatry, 50, 31–37. https://doi.org/10.1016/j.ijlp.2016.10.001

33.

Pelfrey

W. V.

Young

(2020). Police crisis intervention teams: Understanding implementation variations and officer-level impacts. Journal of Police and Criminal Psychology, 35, 1–12. https://doi.org/10.1007/s11896-019-9314-4

34.

Peterson

Densley

(2018). Is crisis intervention team (CIT) training evidence-based practice? A systematic review. Journal of Crime and Justice, 41(5), 521–534. https://doi.org/10.1080/0735648x.2018.1484303

35.

Piza

E. L.

Connealy

N. T.

(2022). The effect of the Seattle police-free CHOP zone on crime: A microsynthetic control evaluation. Criminology & Public Policy, 21(1), 35–58. https://doi.org/10.1111/1745-9133.12570

36.

Rogers

M. S.

McNeil

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. https://doi.org/10.29158/jaapl.003863-19

37.

Rohrer

A. J.

(2021). Law enforcement and persons with mental illness: Responding responsibly. Journal of Police and Criminal Psychology, 36(2), 342–349. https://doi.org/10.1007/s11896-021-09441-2

38.

Rosenthal

R. A.

(2021). The Seattle consent decree: Excessive or effective force in police reform? Doctoral Dissertation, Simon Fraser University.

39.

Rossler

M. T.

Terrill

(2017). Mental illness, police use of force, and citizen injury. Police Quarterly, 20(2), 189–212. https://doi.org/10.1177/1098611116681480

40.

Seattle Police (2018). Crisis intervention program report. https://www.seattle.gov/documents/Departments/Police/Publications/Crisis-Intervention-Outcome-Report-101818.pdf

41.

Seattle Police (2019). Crisis intervention program report. https://spdblotter.seattle.gov/wp-content/uploads/sites/11/2019/10/Dkt-588-3-SPDs-Crisis-Intervention-Program-Report.pdf

42.

Seattle Police Department (n.d.). Crisis response team. https://www.seattle.gov/police/about-us/issues-and-topics/crisis-response-team

43.

Segal

A. F.

Winfree

Jr., T.

Friedman

(2018). Mental health and criminal justice. Aspen Publishing. https://reader.yuzu.com/reader/books/9781543802948

44.

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. https://doi.org/10.1177/0887403414556289

45.

Terrill

Paoline III

E. A.

(2013). Examining less lethal force policy and the force continuum: Results from a national use-of-force study. Police Quarterly, 16(1), 38–65. https://doi.org/10.1177/1098611112451262

46.

Todd

T. L.

Chauhan

(2021). Seattle police department and mental health crises: Arrest, emergency detention, and referral to services. Journal of Criminal Justice, 72(1), 101718. https://doi.org/10.1016/j.jcrimjus.2020.101718

47.

Tucker

A. S.

Van Hasselt

V. B.

Russel

S. A.

(2008). Law enforcement response to the mentally ill: An evaluative review. Brief Treatment and Crisis Intervention, 8(3), 236–250. https://doi.org/10.1093/brief-treatment/mhn014

48.

Usher

Watson

A. C.

Bruno

Andriukaitis

Kamin

Speed

Taylor

(2019). Crisis intervention team (CIT) programs: A best practice guide for transforming community responses to mental health crises. CIT International. https://www.opioidlibrary.org/files/wp-content/uploads/2019/10/cit-guide-desktop-printing-2019_08_16-1.pdf

49.

van Lith

Hutter

R. I.

Sexton

M. M.

Vandeviver

Schoonmade

L. J.

Lindegaard

M. R.

(2025). What do we mean by De-Escalation in police-citizen encounters? A scoping review on conceptualization, techniques, and effectiveness. Journal of Police and Criminal Psychology, 40(3), 1–20. https://doi.org/10.1007/s11896-025-09761-7

50.

Watson

A. C.

Compton

M. T.

(2019). What research on crisis intervention teams tells us and what we need to ask. Journal of the American Academy of Psychiatry and the Law, 47(4), 422–426. https://doi.org/10.29158/JAAPL.003894-19

51.

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. https://vera-institute.files.svdcdn.com/production/downloads/publications/crisis-response-services-for-people-with-mental-illnesses-or-intellectual-and-developmental-disabilities.pdf

52.

Watson

A. C.

Owens

L. K.

Wood

Compton

M. T.

(2021). The impact of crisis intervention team response, dispatch coding, and location on the outcomes of police encounters with individuals with mental illness in Chicago. Policing: A Journal of Police Policy and Practice, 15(3), 1948–1962. https://doi.org/10.1093/police/paab010

53.

Wells

Schafer

J. A.

(2006). Officer perceptions of police responses to persons with a mental illness. Policing: An International Journal of Police Strategies & Management, 29(4), 578–601. https://doi.org/10.1108/13639510610711556

54.

Willis

Comartin

Victor

Kern

Kubiak

(2021). Individuals with mental illness who have multiple encounters with law enforcement. Journal of Offender Rehabilitation, 60(2), 117–137. https://doi.org/10.1080/10509674.2020.1863298

55.

Willis

Kern

L. J.

Hedden

B. J.

Nelson

Comartin

Kubiak

(2023). The impact of crisis intervention team (CIT) training on police use of force. Journal of Offender Rehabilitation, 62(3), 157–173. https://doi.org/10.1080/10509674.2023.2182863

56.

Wood

L. L.

Barrenger

Bonfine

(2024). Contextual factors of mental health crisis calls to law enforcement: A brief report. Community Mental Health Journal, 60(2), 403–409. https://doi.org/10.1007/s10597-023-01185-z