Article Crisis Intervention Teams and People With Mental Illness: Exploring the Factors That Influence the Use of Force Crime & Delinquency 58(1) 57–77 © The Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0011128710372456 http://cad.sagepub.com Melissa S. Morabito1, Amy N. Kerr2, Amy Watson2, Jeffrey Draine3, Victor Ottati4, and Beth Angell5 Abstract The Crisis Intervention Team (CIT) program was first developed to reduce violence in encounters between the police and people with mental illness as well as provide improved access to mental health services. Although there is overwhelming popular support for this intervention, scant empirical evidence of its effectiveness is available—particularly whether the program can reduce the use of force. This investigation seeks to fill this gap in the literature by exploring the factors that influence use of force in encounters involving people with mental illness and evaluating whether CIT can reduce the likelihood of its use. Keywords mental illness, force, police 1 University of Massachusetts Boston, Boston, MA, USA University of Illinois at Chicago, Chicago, IL, USA 3 University of Pennsylvania, Philadelphia, PA, USA 4 Loyola University, Chicago, IL, USA 5 Rutgers, The State University of New Jersey, New Brunswick, NJ, USA 2 Corresponding Author: Melissa S. Morabito, Department of Sociology, University of Massachusetts Boston, 100 William T. Morrissey Boulevard, Boston, MA 02125, USA Email: melissa.morabito@umb.edu Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 58 Crime & Delinquency 58(1) The police have regular contact with people with mental illness and report that up to 6% of individuals considered suspects have a serious mental illness (Engel & Silver, 2001; Teplin, 1984). The potential for violence during these interactions has been of concern to practitioners, researchers, and policy makers alike (Cordner, 2006; Council of State Governments, 2002). The powers of coercion and the authority to use force are defining elements of the police role, yet little research examines its use in the persistent interactions with people with serious mental illness. In response to these concerns, a growing body of literature has developed that examines the type and extent of these encounters as well as police attitudes toward people with mental illness. Simultaneously, interventions such as Crisis Intervention Teams (CIT) have been implemented across the country in an effort to improve officer responses to people with mental illness. Despite these developments, gaps in our knowledge remain. In particular, it is unclear under what conditions force is used in these encounters and whether police interventions such as CIT can defuse violence. There is overwhelming support for police-based intervention among many advocates for people with mental illness, mental health policy makers, and some police leaders, yet scant empirical evidence of its effectiveness is available. Specifically, we are aware of no study that measures the effect of a specific intervention such as CIT on the use of force during such encounters. This investigation seeks to fill this gap in the literature by relying on the vast use of force literature (see Alpert, Dunham, & MacDonald, 2004) and past research on CIT (see Compton, Esterberg, McGee, Kotwicki, & Oliva, 2006). We explore the factors that are predictive of the use of force in incidents involving people with mental illness and test whether the implementation of a CIT program affects the likelihood of its use. Police Response to People With Mental Illness Although not at the core of the police role in the same way as use of force, researchers have long noted that interactions with people with mental illness are a regular part of policing (Bittner, 1967; Engel & Silver, 2001; Morabito, 2007; Teplin & Pruett, 1992). Deinstitutionalization returned people with mental illness to communities and interactions between the police and people with mental illness increased, bringing attention to a previously neglected population (Abramson, 1972; Teplin & Pruett, 1992). Police report that these encounters are outside of their expertise and responsibility and that they feel ill-prepared to provide services (Bittner, 1967; Reuland, Schwarzfeld, & Draper, 2009). These feelings of ill preparation can result from officers’ misunderstanding the behavior of some people with mental illness. Although people with mental illness are usually not dangerous, they can behave bizarrely and may not respond to police officer cues in a predictable manner based on the behavior of others (Cordner, 2006; Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 Morabito et al. 59 Reuland et al., 2009). Cordner (2006) notes that police officers may approach encounters with people with mental illness using the same tactics and strategies that they would use for any other type of encounter. Situations can become tense when officers do not get the compliance that they expect and may subsequently escalate to violence or injury. Although there is no evidence that mental illness alone will lead to resistance or resulting use of force (see Terrill & Mastrofski, 2002), officers are concerned about these encounters (Kaminski, DiGiovanni, & Downs, 2004) because of their perceived unpredictability. Some evidence suggests that police encounters might be most dangerous for people with mental illness in terms of injury (Cordner, 2006) but scant documentation of these interactions exists. Not only are encounters with people with mental illness unpredictable their resolution can also be frustrating for police officers. Interactions with people with mental illness can be described as time-consuming, with officers often waiting hours with a person in crisis to be admitted to a hospital only to have that person discharged a mere few hours later (see Bittner, 1967; Lurigio & Swartz, 2000; Reuland et al., 2009). Connecting people with mental illness to emergency services can involve greater police resources than more traditional responses of arresting the individual or offering no formal response. Evidence suggests, however, that these encounters infrequently end in arrest (Engel & Silver, 2001), but because these situations are resolved informally, the end results are relatively unknown though of concern to both practitioners and researchers. Crisis Intervention Teams One approach to improving police response to people with mental illness is the Crisis Intervention Team (CIT). CIT is a police-based prebooking approach with specially trained officers who provide first-line response to calls involving a person with mental illness and who act as liaisons to the mental health system (Borum, Deane, Steadman, & Morrissey, 1998). The intervention is based on a model developed by the Memphis Police Department (Council of State Governments, 2002) and involves three elements that are important to our goals in this article. First, officers self-select to participate. Once that self-selection is approved, officers receive 40 hours of specialized training regarding mental illness. The range of specific topics offered differs depending on the department and community priorities but all programs include information on deescalation techniques and education on mental illness. After training is completed, officers are certified as members of the CIT unit. When a call is identified as involving a person with a serious mental illness, a CIT officer is dispatched and is given the authority of officer in charge. The third element of a CIT intervention is a system-level approach to addressing the needs of people with mental illness including partnership with local advocacy groups and providers. CIT is designed so that police are just one part of the response Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 60 Crime & Delinquency 58(1) and are not solely responsible for addressing the needs of people with mental illness in the community. One of the significant difficulties experienced by police officers in addressing people with mental illness is the long wait times associated with accessing mental health care for people in crisis (Bittner, 1967). This partnership is designed to reduce the bureaucracy associated with admission to care and allows officers to more quickly and easily direct people in need to services. Among other outcomes, CIT programs have been hypothesized to improve officers’ abilities to more safely interact with persons with mental illness, including reduced use of force and subsequent injury to both police and citizens (Watson, Morabito, Draine, & Ottati, 2008). CIT is being widely implemented in medium and large size cities across the United States as a policy initiative to demonstrate responsiveness to mental illness. Despite these claims, there has been little empirical research to support the effectiveness of the intervention on shaping police encounters particularly around the use of force. Instead, the bulk of the CIT literature has focused more on process than outcomes (Compton et al., 2006; Compton & Chien, 2008; Watson et al., 2008). Police Use of Force and CIT Effectiveness The goal in all police encounters is for the officer to gain and maintain control of an unknown situation (Alpert et al., 2004). However, the use of force in any encounter between the police and the public is a relatively rare event, occurring in less than 1% of incidents and usually involving a small number of officers (Adams, 2004). In addition to being relatively rare, in practice, the use of force is concentrated at the lower end of the spectrum infrequently involving the use of weapons (Adams, 2004). Similarly, the injuries that result from these encounters are also minor— described mostly as abrasions and bruises (Alpert & Dunham, 1999). Although the use of force is relatively rare among police, it is an issue that receives a disproportionate amount of attention (Adams, 2004) as there is always present a concern that force would lead to more serious injury or death increasing interest in an intervention like CIT, which might be associated with reduced use of force. The criminal justice literature overwhelmingly suggests that situational factors such as demeanor, hostility, and impairment explain most of the variation in the use of force by police officers (Alpert et al., 2004; Alpert & Dunham, 1999). Police typically use force when they are trying to make an arrest and the suspect is resisting (Adams, 2004). Other factors such as the type of and seriousness of the crime (MacDonald, Manz, Alpert, & Dunham, 2003) may also influence the likelihood of the use of force. These situational characteristics provide important context for encounters that involve people with mental illness because their behavior is often misunderstood. Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 Morabito et al. 61 Evidence suggests that it is a combination of substance abuse related activity and hostile demeanor that result in the use of force during police encounters (Cordner, 2006; Kaminski et al., 2004; Terrill & Mastrofski, 2002). Yet, even the existing literature on drug impairment and force cites mixed results. Some evidence suggests that the use of drugs and alcohol does matter (see Garner, Maxwell, & Heraux, 2002; Kaminski et al., 2004) whereas other studies find no relationship between alcohol use and force (Alpert & Dunham, 1999). Interestingly, most of these studies focus on suspects under the influence of alcohol, a legal drug, combined with street drugs (see Engel, Sobol, & Worden, 2000; Garner et al., 2002)—without noting the presence of a mental illness. These distinctions are particularly relevant because mental illness and substance abuse are commonly co-occurring disorders (Abram, 1990). As such, police officers may encounter individuals who have a mental illness and are also under the influence of drugs or alcohol—increasing their difficulty in managing the incident and perhaps making it difficult for the officer to recognize the mental illness. People with mental illness may behave in ways that many police view as bizarre, nonresponsive or even hostile (Cordner, 2006), putting themselves at greater risk of injury than other citizens. Manifestations of mental illness may be construed as hostile or resistant to arrest. Encounters handled by CIT-trained officers may play out differently if officers are able to interpret seemingly hostile behavior as signs of mental illness. CIT officers are trained to recognize mental illness and provide linkage to available resources. When a CIT-trained officer responds to an incident involving a person with a mental illness, we might expect that fewer formal sanctions are employed—including the use of force. Rather, when there is no or minor illegal activity, police will employ other tools—such as deescalation techniques—to resolve the situation. These techniques may be learned in their CIT training. As such, it can be expected that in encounters between CIT trained officers and people with mental illness, violence and injury will be less likely than in similar encounters with nonCIT trained officers. This would suggest that CIT training could be particularly powerful in the prediction of the use of force. Current Focus and Research Hypotheses This investigation focuses on the use of force by police officers in interactions with people with mental illness, and the impact of CIT on the use of force. First, we examine whether the factors that influence the use of force in these encounters are similar to those that influence the use of force in general police encounters. Next, we test whether the implementation of CIT can affect the likelihood of the use of force. The following are our hypotheses: Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 62 Crime & Delinquency 58(1) Table 1. District Description CIT Saturation High Low Disadvantage High Low District A District B District C District D Hypothesis 1: The same factors that are related to police use of force in encounters with the general population are also related to use of force in encounters with people with mental illness. Hypothesis 2: The adoption of CIT decreases the likelihood of the use of force in encounters with people with mental illness. Data and Method The current study is of the Chicago Police Department (CPD) CIT program. Between 2002 and 2005, CPD expanded their preservice academy mental health training from 5 to 9 hours with additional mental health lecturers. In 2005, they piloted the CIT program in two districts. Some officers in those districts volunteered to participate in a 40-hour training session that provides sworn officers with additional skills to handle interactions with persons with mental illness and information on mental health resources including designated drop-off points for psychiatric emergencies. In 2007, the department began city-wide expansion of the program. At the time of the study, all districts had officers that were CIT trained. However, the CIT pilot districts had many more CIT officers, and they had been trained for a much longer period of time. The purpose of the evaluation was to determine the effectiveness of the additional training and services provided to police and people with mental illness in Chicago. We surveyed sworn CIT-trained and non-CIT-trained officers from the two pilot districts and two comparison districts that did not have as many CIT-certified officers (see Table 1). Because little official data exist about encounters involving police and people with mental illness (Engel & Silver, 2001), we interviewed officers about their involvement. The scant official data available could not provide the level of detail necessary to learn about these calls for service, particularly in parsing out the levels of force and resistance used. Officers who agreed to participate in the study were asked to complete an inperson interview. In the following section, we will describe the sampling and data collection procedures. Next, we provide descriptions of the encounters involving police officers and people with mental illness. Then, using logistic regression, we Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 63 Morabito et al. Table 2. District Demographics District A B C D Population size Racial distribution African American (%) Caucasian (%) Latino (%) Property crime Violent crime Average dispatched calls per month Crisis Intervention   Team–trained officers (%) Disadvantage variables Average proportion of single   mother headed households Average proportion of households   on public assistance Average proportion of households   living below the poverty line 91,600 98,391 105,360 107,516 98.43 13.26 98.72 3.95 0.39 65.93 0.39 74.28 0.92 11.64 0.60 15.89 0.069 0.030 0.064 0.043 0.026 0.004 0.021 0.004 3,183 908 3,649 944 9.56 15.27 4.02 4.69 0.306 0.088 0.226 0.067 0.193 0.040 0.106 0.021 0.365 0.159 0.194 0.096 explore the individual and district level predictors of use of force in calls involving individuals with mental illness. Finally, we discuss the effect of CIT on the use of force in these incidents. Sampling and Data Collection Procedures We sampled officers from the two CIT pilot districts, District A and District B. At the onset of the study, these districts were highly saturated with exposure to CIT1 but differed greatly in terms of available mental health resources, crime rates and level of structural disadvantage. These pilot CIT districts were matched with comparable districts that had fewer CIT-trained officers. Comparison Districts C and D were selected to mirror the location, resource levels, population demographics, crime rates and disadvantage variables of the pilot districts (see Table 2). Once comparison districts were selected, district personnel lists of CPD officers with a minimum of 18 months service served as the sampling frame.2 A total of 80 officers from each district were invited to participate, including all CIT officers and a random sampling of non-CIT officers based on the proportion of officers assigned to each watch. Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 64 Crime & Delinquency 58(1) The researchers visited each district to announce the study during all watch roll calls in February 2008. The 80 officers selected in each district received a letter inviting them to learn more about the study in an in-person meeting. We then worked with watch commanders to set up times to meet with officers during their duty hours to discuss the study and ask them to participate. If the officer agreed to participate, the researcher obtained consent from the officer and then completed a 45-minute interview. Participation in the study was voluntary and 17.4% of the 333 invited officers declined. Additionally 17.7% of the invited officers were not contacted because of situations such as extended leave or departmental transfer. However, the demographics of the obtained sample are similar to that of the Chicago police force, and thus it seems that our sample is representative of the CPD population (Chicago Police Department, 2008). Participants completed a four-part interview with the researcher that first asked them to recall and answer questions about the most recent incident involving a person with mental illness while working. The next section included questions concerning the officer’s perceptions about the effectiveness of the district’s psychiatric services, the district’s ability to handle mental health calls and the usefulness of their policing skills in handling mental health calls. Officers were also asked to give their perspectives on the CIT program. Finally, demographic questions such as the officer’s race, length of time in the district, and familiarity with mental illness were included (see Table 3). Defining Calls Involving Persons With Mental Illness For this investigation, mental illness is defined based on the observations of officers. Research suggests that officers transporting individuals to crisis centers typically make accurate judgments on the need for mental health care (Strauss et al., 2005; Teplin, 1984). In this study, officers were asked to identify whether they were involved in incidents with adults with mental illness and asked them to describe their most recent incident through a series of structured and semistructured questions. No other information was used to determine the presence of a citizen with mental illness in the encounters recalled by the police. Method Dependent Variable Police use of force in an encounter with a person with a mental illness is the dependent variable in this investigation. Objectively reasonable use of force is difficult to ascertain particularly without observing the incident in question (Terrill, Alpert, Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 65 Morabito et al. Table 3. Officer/Participant Demographics Count A (n = 61) B (n = 51) C (n = 58) D (n = 46) CIT training No 27 19 47 32 Yes 34 32 11 14 Race/ethnicity White 18 29 17 32 Black 28 5 34 2 Hispanic 12 10 6 9 Other 3 7 1 0 Gender Male 46 39 45 40 Female 15 12 13 6 Highest level   of education High school 2 0 1 2 Associate’s degree 11 4 6 4 Some college 12 12 13 18 College degree 26 19 26 15 Graduate degree 10 16 12 7 Current rank Patrol 53 12 49 39 Sergeant 3 7 2 6 Lieutenant 1 2 2 1 FTO 4 0 3 0 Other 0 0 2 0 Watch assignment First 14 13 10 9 Second 20 17 23 18 Third 22 15 17 10 Fourth 5 6 7 9 0 0 1 0 Other Years on police 9.65 (6.00) 12.62 (8.03) 10.54 (8.75) 13.83 (6.73)   force; mean (SD) Age in years; 39.82 (9.25) 42.92 (8.80) 39.50 (9.63) 43.09 (8.14)   mean (SD) Note. CIT = Crisis Intervention Team; FTO = Field training officer. Dunham, & Smith, 2003). Because this study does not involve direct observation, the purpose of this article is not to measure whether officers were using an appropriate level of force in encounters with people with mental illness. However, because Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 66 Crime & Delinquency 58(1) Table 4. Distribution of Use of Force Counta Mere presence Verbal warnings, commands, and/or persuasion Physical control of the suspect Use of a weapon other than my firearm Use of my firearm Total Percentage 54 29.3 91 49.5 34 18.5 3 1.6 2 1.1 184 100 a. Counts of less than 216 are a result of missing data. this is a comparison between conditions with conceptual and design controls for other explanatory variables, we are able to draw conclusions about the level of force used relative to the implementation of CIT.3 In Chicago, force is measured across a continuum. As such, the language in our measure corresponds with the force continuum officers learn through the CPD Academy and located in each officer’s manual to facilitate officer comprehension (Chicago Police Department, 2003). As is accepted practice in the literature and based on the CPD standard operating procedure, use of force was measured on a 5-point scale (Alpert et al., 2004; Lawton, 2007; Terrill & Reisig, 2003). Officers were asked to self-report the highest level of force that the most recent situation involving a person with mental illness necessitated. Officers chose from the following categories: (1) My mere presence was enough; (2) Verbal warnings, commands, and/or persuasion were necessary; (3) Physical control of the suspect, such as holding, open hand strike, and/or knee strike, was necessary; (4) The use of a weapon other than my firearm, such as a taser, baton, and/or chemical weapon was necessary; (5) The use of my firearm was necessary. A breakdown of these incidents is included in Table 4. We chose to create a dichotomous variable where Category 1 became “No Force” and Categories 2 through 5 were collapsed to create “Force.” As highlighted by Klinger (1996), research that divides force into physical and nonphysical loses the verbal command portion of the use of force continuum most used by officers. Although verbal warnings, commands, and/or persuasion may be used to deescalate, they are ultimately coercive as they imply that the officer has to assert their authority verbally to gain compliance.4 Independent Variables The predictor variables included in our model are drawn from the use of force literature. These include the following. Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 Morabito et al. 67 Resistance. Citizens who are antagonistic are more likely to be on the receiving end of police force than others who are more compliant (see Crank, 1997; Reisig, McCluskey, Mastrofski, & Terrill, 2004). This is particularly an issue with people with mental illnesses who may encounter the police when experiencing psychosis and as a result may behave in ways perceived as resistant. CIT training should increase officer understanding about seemingly resistant behavior. As such, we might expect an interaction between resistant behavior and officer CIT training. To measure the level of violence or conflict presented by the person with mental illness, we operationalize resistance with two different indicators: physical resistance and a resistive demeanor. Physical resistance is measured as the highest level of resistance exhibited by the suspect during the interaction using a 4-point scale with the following categories: (1) Cooperative: The subject exhibited no resistance such that they were cooperative with or without direction. (2) Passive Resister: The subject exhibited passive resistance in that he or she made nonmovements in response to verbal and other directions such as stiffening to dead weight. (3) Active Resister: The subject exhibited active resistance such that he or she made movements to avoid physical control such as fleeing or pulling away. (4) Assailant: The subject attacked you or another officer such that the subject’s actions were likely to cause death or serious physical injury with or without weapons. These categories were adapted from the Chicago Police Department continuum of force as with the force scale (Chicago Police Department, 2003). Because of the low incidences of high levels of resistance, Categories 2, 3, and 4 were collapsed to create a dichotomous variable that captured the presence of resistance or lack of resistance (resistance = 0, no resistance = 1).5 We measured resistive demeanor using the following statements: (1) The subject displayed combative/assaultive behavior; (2) The subject was verbally abusive; (3) The subject was upset/angry/agitated; (4) The subject had a calm demeanor (Kaminski et al., 2004). The officers chose their level of agreement with each statement from a 4-point scale (not at all = 1, very little = 2, somewhat = 3, to a great extent = 4). Statement 4 was reverse scored and the rankings provided by the officers were averaged to obtain the score for the demeanor. A higher score on the demeanor scale indicates a more resistant subject. Differential racial status and subject gender. We captured race by coding subject and officer as either majority (White) or minority (non-White). Hispanic ethnicity was coded as minority regardless of race (Alpert et al., 2004; Skogan, 2006). A majority officer–minority subject encounter was coded as High Officer Differential Racial Status (1). Encounters with either matched officer–subject status or minority officer–majority subject status were coded as Low Officer Differential Racial Status (0). The subjects’ ages were unavailable so age differentials were excluded from the authority calculation. The gender of the subject was reported by the officer Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 68 Crime & Delinquency 58(1) (male = 1, female = 2). Based on previous research, we expect that officer authority, based on race, will be correlated with decreased the use of force (Alpert et al., 2004). Subject impairment. The subject’s level of drug or alcohol impairment can factor into explanations of the level of force an officer deems necessary. The officer’s perception of impairment may increase perception of threat because the presence of drugs or alcohol has potential to exacerbate resistant behavior and demeanor. Subject intoxication is coded as 1 in cases where the officer perceived the subject to be intoxicated at the time of the incident. Officer experience. We measure the experience as the length of an officer’s service through the self-reported number of years the officer had been with the CPD. Officer CIT training. To determine each individual officer’s CIT training status, we obtained the list of CIT training completion dates from the CPD and coded the officer as CIT trained if their completion date occurred prior to the interview. Anyone who had not completed training was coded as a non-CIT-trained officer (CIT = 1, non-CIT = 0). District. The district variables capture the level of saturation, disadvantage, as well as other situational differences. Each district is dummy coded as a separate variable (see Table 1). Analysis Strategy To examine the relationships between force, police and subject characteristics, and CIT, we used a logistic regression model that tested our two hypotheses. District C serves as the reference category for the regression because it is low in CIT saturation and low in mental health resources. We report descriptive statistics, odds ratios, the standardized logistic regression coefficients and the standard errors, as well as possible interaction effects. Findings Descriptive Results Table 5 shows the independent and dependent variable frequencies and mean distributions for the variables included in the model. The use of verbal or physical force by the officers seems to be more common than withholding force. Surprisingly though, the percentage of incidents with reported subject physical resistance is slightly less than half. However, our second measure of resistance, demeanor, averaged high on the scale with 61.5% scoring an average of greater than 2. Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 69 Morabito et al. Table 5. Independent and Dependent Variable Descriptives Code Counta (%) Dependent variable Force Independent variables Subject physical resistance Subject demeanor Subject gender Subject impairment Officer experience (years) Individual officer saturation   (CIT training) Differential racial status District A District B District C District D 0 (no) 61 (29.3) 1 (yes) 147 (70.7) 0 (no) 110 (53.9) 1 (yes) 98 (47.1) — 208 2.76 (0.93) 0 (male) 123 (58.9) 1 (female) 86 (41.1) 0 (no) 153 (73.2) 1 (yes) 56 (26.8) — 216 11.46 (7.56) 0 (no) 125 (57.9) 1(yes) 0 (low) 1 (high) — — — — Mean (SD) 91 (42.1) 143 (69.1) 64 (30.9) 61 (28.2) 51 (23.6) 58 (26.9) 46 (21.3) a. Counts of less than 216 are a result of missing data. Demeanor and physical resistance had a large correlation (r = .60, p < .01) according to Cohen’s (1988) standards. However this did not affect their ability to separately significantly predict the force variable. The remaining variables all had small to medium bivariate correlations (rs < .40). Regression Analyses The Hosmer and Lemeshow test, χ2(8) = 8.71, p > .05, as well as the −2 log likelihood chi-square test, χ2(9) = 83.77, p < .001, tells us that our model as a whole fits significantly better than the constant only model. The model correctly predicts 83.4% of the cases. Based on our analyses—presented in Table 6—there are mixed findings about the effects of CIT. Most notably, our findings only suggest a marginal effect of CIT training on officer use of force in the opposite direction of our hypothesis. However when looking at the interaction between CIT and resistant demeanor, Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 70 Crime & Delinquency 58(1) Table 6. Analyses of the Relationship Between Force and Individual and Situational Characteristics Odds Lower Upper Predictor Variables B SE Ratio 95% CI 95% CI Physical resistance Resistive demeanor District A District B District D District Ca Officer CIT training Officer experience Differential racial status Subject gender Subject impairment CIT × resistive   demeanor interaction 3.00*** 0.71 20.15 4.99 81.40 1.57** 0.52 4.79 1.73 13.30 −2.13** 0.76 0.12 0.027 0.53 −1.86* 0.75 0.75 0.036 0.68 −1.61* 0.76 0.20 0.045 0.88 — — — — — 1.50 15.98 0.84 302.70 2.77† 0.039 0.035 1.04 0.97 1.11 −0.54 0.51 0.58 0.22 1.58 0.34 0.47 1.40 0.56 3.48 0.027 0.53 1.03 0.36 2.93 −1.31* 0.65 0.27 0.075 0.97 Note. CIT = Crisis Intervention Team; CI = confidence interval. a. District C served as the reference category. † p < .10. *p < .05. **p < .01. ***p < .001 the officers with CIT training are less likely to use force as the subject’s demeanor becomes more resistant than the non-CIT officers. Additionally, CIT saturation and disadvantage of the districts seem to impact the use of force. As shown in Table 6, the officers working in high CIT saturation districts (Districts A and B) report using significantly less force than officers in one of the low CIT saturation districts (C). The officers in the low saturation and low disadvantage district (D) also report using significantly less force than the low saturation, high disadvantage district (C). By working in District C rather than District A, B, or D, the odds of an officer reporting using force in a call with a person with a mental illness increase. There is some support for our hypothesis that the correlates that predict the use of force in incidents with people with mental illness are the same as those that influence the use of force in interactions with the general population. Whereas individual officer and citizen characteristics are not significant predictors of use of force (differential racial status, officer experience, subject drug/alcohol impairment or subject gender), other characteristics do influence its use. Similar to the findings in the existing criminal justice literature, the use of force is significantly related to physical resistance and demeanor in encounters with people with serious mental illness (see Alpert et al., 2004; Klinger, 1996). Based on the odds ratios, it appears that physical resistance is by far the strongest predictor of force. With the presence Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 Morabito et al. 71 of resistance, the odds of the officer using some form of force increase. The odds of the officers using force also increase when there is an increase in resistant demeanor. Discussion and Conclusions Our findings indicate that the interaction of CIT and demeanor, district characteristics, and subject resistance produce significant effects on officer use of force. Additionally, officer CIT training marginally affects the use of force though opposite of the predicted direction. Although the marginal effect suggests CIT officers are more likely to use higher levels of force, this must be interpreted in conjunction with the differential effect demeanor has on CIT- and non-CIT-trained officers. The interaction suggests that a CIT officer is likely to respond with less force for an increasingly resistant demeanor in comparison with non-CIT officers. Interestingly, physical resistance does not differentially affect CIT-trained officers. The interaction effects seem to fit directly with the goals of the CIT training. The knowledge and skills that they gain are intended to assist in preventing escalation as well as deescalating an already resistant subject while also ensuring the safety of the officers and subject at the encounter. Previous research has found that officers often respond to calls with a mental health component as they would to any other call (Cordner, 2006; Reuland et al., 2009). Unfortunately, the command and control techniques that are used in other types of calls may inadvertently escalate a situation involving subject experiencing a mental health crisis. Our research suggests that CIT training may lead to different responses from trained officers in encounters with people with mental illness who exhibit seemingly resistant behavior. CIT officers may be able to recognize resistant demeanor as symptoms of a mental illness and thus implement deescalation techniques. However, in a situation involving a physically resistant subject, all officers may find force necessary to control the situation and maintain safety of all involved. Our findings suggest that implementing CIT in urban police districts can affect the force officers use in encounters with people with serious mental illness. In Districts A and B, both with a high level of CIT implementation, officers reported using less force than those officers in District C where CIT implementation is limited. Although there are a number of district level characteristics that may account for some of this difference, including the racial distribution, crime rates, mental health resources, and disadvantage, our analyses suggest that CIT saturation, the common link between Districts A and B, plays some role in the differences in police behavior. Future research including more officers and more districts would allow for hierarchical analyses and the power to test other differences between the districts that could account for the finding. Given the unique Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 72 Crime & Delinquency 58(1) qualities of Chicago, one of the largest police forces in the nation, and the diversity of police departments, future research is also needed to study the effects of CIT in a variety of cities with differing characteristics. Ideally, an experimental design with random assignment would be used to control the district and city variables but for the current investigation, practical, programmatic, and ethical barriers associated with policing make this impossible. Given that we would also expect the districts with higher saturation of the CIT program to see less use of force in encounters with people with serious mental illness, then low saturation and high disadvantage districts would have the highest level of force (Watson et al., 2008). We would expect that Districts C and D officers would use more force than Districts A and B because of the implementation of CIT. Our actual findings, however, were not so clear-cut. District C officers reported significantly more use of force than those of A, B, and D. Yet there were no differences between District D and District A or B. One explanation could be that the officers’ use of force could not be reasonably lowered without increased risk to police and community members. The high resources and relative stability of Districts D and B may reduce the need for the use of force. The level of force used by officers in District A may not be different from District D because the high level of CIT saturation could translate into more officers that have the skills, knowledge of resources, and cooperation from the limited mental health resources necessary to make force unnecessary in their encounters with people with serious mental illness. Thus it seems that the use of force in encounters with people with serious mental illness can be influenced by both community characteristics, such as disadvantage level, as well as CIT. Although our study lends support for the full system implementation of CIT, it is important to note that individual officer CIT training only affected the use of force when taking into account the subject demeanor. Despite the claims of the program, we had to question why CIT training at the individual officer level did not directly reduce the use of force in encounters with people with mental illness. There are a number of reasons that might explain this disconnect. First, the majority of incidents involving the use of force in our sample were located on the lower end of the continuum. In these encounters, low-level force may have been an appropriate and justifiable response. CIT training is not meant to completely eliminate the use of force. Additionally, because CIT is considered a system-level rather than individuallevel approach, police training must be implemented in conjunction with support and participation of the mental health system within the community as well as support of fellow officers and supervisors. It may be that individual officers in Chicago also did not feel this support. At the time of this study, CIT was not implemented fully department and officers may have felt that CIT was not accepted by the department or the city. Finally, CIT breaks with the traditional police chain of command. In theory, the CIT officer becomes the officer-in-charge at a mental health encounter, applies Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 Morabito et al. 73 nontraditional tactics, and requires longer than average time to resolve the call. Consequently, lower ranking officers may have authority over their superiors, may be breaking tactical policing norms, and may be diverting more calls to fellow officers thus adding to their workload. Creating this situation would require that the non-CIT police personnel, including administration, supervisors and lower ranking officers, be open to calls handled in a nontraditional manner and be willing to comply with any subsequent changes. In conversations with the police department following the completion of this project, we learned that in the saturated districts (A and B), more sergeants were CIT trained. It is possible that officers in these districts were more willing to take risks by using their CIT training than their peers in the other districts because they felt supported by their supervisors. Further research is needed to investigate the reasons why CIT training may not have an impact on individual officer behavior. One limitation of this study is the use of single, dichotomous force and resistance measures based on recall. Ideally more levels of force and resistance could be used to capture additional variability in officer and subject behavior if adequate power is available. Because the use of force is a relatively rare event, the number of incidents reported by the officers in the study did not allow for these types of analyses. Additionally, researchers have suggested that documenting the sequence of force and resistance throughout the encounter can provide a more accurate picture of how coercion was used to control the situation (Alpert et al., 2004). As with any research involving individual recall of an event, the results may be biased by the officer’s perceptions and memory. However having the officer recount a real scenario avoids the biases and problems with external validity associated with providing a hypothetical situation. Although our sample was demographically representative of the Chicago police force, a number of officers declined or were unable to be contacted. This could introduce a sampling bias into our study. Officers were told that the study concerned encounters with individuals with mental illness before given the option to decline. Potentially officers who declined could have a different perspective or experiences with individuals with mental illness that influenced their decision not to participate. Future research could limit the type of information given to officers prior to volunteering for the study. Overall our findings suggest that the outcomes for people with mental illness may be somewhat beyond the scope of a police intervention alone and instead must also become a community issue. Officers can adjust their method of policing persons with mental illness but access and availability of services seem to be particularly important. It may be that in communities with ample services people with mental illness are less likely to reach a crisis point when out on the street where they are likely to be addressed by the police. In communities where services are lacking, police may be more likely to encounter people with mental illness in crisis Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 74 Crime & Delinquency 58(1) necessitating more force to maintain safety. Differences we may hear through the media and anecdotal evidence may be the result of poverty, access to services and area of residence rather than police response. If indeed the implementation of CIT in the disadvantaged district led to the lowered use of force as our research suggests, then further research should explore which, if any, elements of CIT are integral to its success. Is it the proportion of officers trained, relationships with mental health services, individual training or some combination of the three? This could be helpful to departments looking to create or modify existing programs designed to address the needs of people with mental illness. Authors’ Note These data were provided by and belongs to the Chicago Police Department. Any further use of these data must be approved by the Chicago Police Department. Points of view or opinions contained within this document are those of the authors and do not necessarily represent the official position or policies of the Chicago Police Department. The contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health. Declaration of Conflicting Interests The authors declared no conflicts of interest with respect to the authorship and/or publication of this article. Funding The authors disclosed receipt of the following financial support for the research and/ or authorship of this article: Work supported by NIMH R34 MH081558. Notes 1. We define CIT saturation as the level of exposure to the CIT program both on the individual and district level. 2. Officers with less experience were not included because an officer must have completed the 18-month probationary period to take part in the CIT. 3. We considered using injury as well as use of force to determine the impact of CIT but were unable to do so because of the limited number of injuries to officers or subjects in our study, which are rare anyway. Future studies with a larger sample size should test the effect of CIT on injury to individuals involved in the encounter. 4. Although a three level dependent variable that separates no force, verbal commands, and physical force would provide us additional insight into the use of force, the current research sample size does not allow for such a test of the data. This is important for future studies with larger sample sizes. Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 Morabito et al. 75 5. As with the dependent variable use of force, use of additional categories for the physical resistance level is preferable to understand the situational variables that influence use of force. However, the sample size restricts testing the data with multiple levels. Future studies with larger sample sizes would benefit from additional levels to the physical resistance variable. References Abram, K. M. (1990). The problem of co-occurring disorders among jail detainees: Anti-social disorder, alcoholism, drug abuse and depression. Law and Human Behavior, 14, 333-345. Abramson, M. F. (1972). The criminalization of mentally disordered behavior: Possible side effects of a new mental health law. Hospital and Community Psychiatry, 23, 101-107. Adams, K. (2004). What we know about police use of force. In Q. Thurman & J. Zhao (Eds.), Contemporary police: Controversies, challenges, and solutions (pp. 187-199). Los Angeles, CA: Roxbury. Alpert, G. P., & Dunham, R. G. (1999). The force factor. Washington, DC: Department of Justice. Alpert, G. P., Dunham, R. G., & MacDonald, J. M. (2004). Interactive police-citizen encounters that result in force. Police Quarterly, 7, 475-488. Bittner, E. (1967). Police discretion in the emergency apprehension of mentally ill persons. Social Problems, 14, 278-292. Borum, R., Deane, M. W., Steadman, H. J., & Morrissey, J. (1998). Police perspectives on responding to mentally ill people in crisis: Perceptions of program effectiveness. Behavioral Sciences & the Law, 16, 393-405. Chicago Police Department, Use of Force, GO-02-08-01A. (2003). Chicago Police Department. (2008). Annual report 2007: A year in review. Retrieved October 10, 2008, from http://www.CityofChicago.org/Police. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Compton, M. T., & Chien, V. H. (2008). Factors related to knowledge retention after crisis intervention training for police officers. Psychiatric Services, 59, 1049-1051. Compton, M. T., Esterberg, M. L., & McGee, R., Kotwicki, R. J., & Oliva, J. R. (2006). Crisis intervention team training: Changes in knowledge, attitudes, and stigma related to schizophrenia. Psychiatric Services, 57, 1199-1202. Cordner, G. (2006). People with mental illness. Washington, DC: Office of Community Oriented Policing Services. Council of State Governments. (2002). Criminal Justice/Mental Health Consensus Project. New York, NY: Council of State Governments. Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 76 Crime & Delinquency 58(1) Crank, J. P. (1997). Understanding police culture. Cincinnati, OH: Anderson. Engel, R., & Silver, E. (2001). Policing mentally disordered suspects: A re-examination of the criminalization hypothesis. Criminology, 39, 225-252. Engel, R., Sobol, J., & Worden, R. (2000). Further exploration of the demeanor hypothesis: The interaction effects of suspects’ characteristics and demeanor on police behavior. Justice Quarterly, 17, 235-258. Garner, J., Maxwell, C., & Heraux, C. (2002). Characteristics associated with the prevalence and severity of force used by the police. Justice Quarterly, 19, 705-746. Kaminski, R., DiGiovanni, C., & Downs, R. (2004). The use of force between the police and persons with impaired judgment. Police Quarterly, 7, 311-338. Klinger, D. (1996). More on demeanor and arrest in Dade County. Criminology, 34, 61-82. Lawton, B. (2007). Levels of nonlethal force: An examination of individual, situational and contextual factors. Journal of Research in Crime and Delinquency, 44, 163-184. Lurigio, A., & Swartz, J. (2000). Changing the contours of the criminal justice system to meet the needs of persons with serious mental illness. In J. Horney (Ed.), Policies, processes, and decisions of the criminal justice system (pp. 45-108). Washington, DC: National Institute of Justice. MacDonald, J. M., Manz, P. W., Alpert, G. P., & Dunham, R. G. (2003). Police use of force: Examining the relationship between calls for service and the balance of police force and suspect resistance. Journal of Criminal Justice, 31, 119-127. Morabito, M. S. (2007). Horizons of context: Understanding the police decision to arrest people with mental illness. Psychiatric Services, 58, 1582-1587. Reisig, M. D., McCluskey, J. D., Mastrofski, S. D., & Terrill, W. (2004). Suspect disrespect toward the police. Justice Quarterly, 21, 240-268. Reuland, M., Schwarzfeld, M., & Draper, L. (2009). Law enforcement responses to people with mental illnesses: A guide to research-informed policy and practice. New York, NY: Council of State Governments Justice Center. Skogan, W. (2004). Community policing: Common impediments to success. In L. Wycoff and M. A. Wycoff (Eds.), Community policing: The past, present, and future (pp. 159-167). Washington, DC: Annie E. Casey Foundation, Police Executive Research Forum. Strauss, G., Glenn, M., Reddi, P., et al. (2005). Psychiatric disposition of patients brought in by crisis intervention team police officers. Community Mental Health Journal, 41, 223-228. Teplin, L. (1984). Criminalizing mental disorder: The comparative arrest rates of the mentally ill. The American Psychologist, 29, 794-803. Teplin, L., & Pruett, N. S. (1992). Police as streetcorner psychiatrists: Managing the mentally ill. International Journal of Law and Psychiatry, 142, 593-599. Terrill, W., Alpert, G. P., Dunham, R. G., & Smith, M. R. (2003). A management tool for evaluating police use of force: An application of the force factor. Police Quarterly, 6, 150-171. Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016 Morabito et al. 77 Terrill, W., & Mastrofski, S. D. (2002). Situational and officer-based determinants of police coercion. Justice Quarterly, 19, 215-248. Terrill, W., & Reisig, M. (2003). Neighborhood context and police use of force. Journal of Research in Crime and Delinquency, 40, 291-321. Watson, A. C., Morabito, M. S., Draine, J., & Ottati, V. (2008). Improving police response to persons with mental illness: A multi-level conceptualization of CIT. International Journal of Law and Psychiatry, 31, 359-368. Bios Melissa S. Morabito is an assistant professor in the Department of Sociology at the University of Massachusetts Boston. She received her PhD from American University and was previously a postdoctoral fellow at the Center for Mental Health Services and Criminal Justice Research. Her research focuses on police innovation and responses to public health problems. Amy N. Kerr is currently a doctoral student in Loyola University Chicago’s Social Psychology program and works as a project coordinator under Dr. Amy Watson at the University of Illinois Chicago. Amy Watson, PhD, is an assistant professor at the Jane Addams College of Social Work at the University of Illinois–Chicago. Her research focuses on the interface of the criminal justice and mental health systems. Jeffrey Draine, PhD, is affiliated with the University of Pennsylvania. In addition, he is a codirector of the Center for Behavioral Health Services and Criminal Justice Research at Rutgers University. He conducts research concerning the intersection of behavioral health with the criminal justice system. Victor Ottati, PhD, is a social psychologist currently serving as a professor in the Department of Psychology at Loyola University Chicago. He received his BA from the University of Michigan and received his PhD in psychology from the University of Illinois–Champaign. His research interests include social cognition, attitudes, persuasive communication, stereotyping, prejudice, mental illness stigma, political psychology, and cross-cultural psychology. Beth Angell is an associate professor in the School of Social Work and a faculty member at the Institute for Health, Health Care Policy, and Aging Research at Rutgers, the State University of New Jersey. She is also affiliated with the Center for Behavioral Health Services & Criminal Justice Research. Her research focuses on mental health service delivery issues, particularly regarding mandated treatment and co-occurring involvement in the criminal justice system. Downloaded from cad.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016