Feature Article

The complicated relationship between law enforcement and the public

Cases of police brutality spur collaborations between researchers and public service professionals to improve police training and service provision.
By John Young, PhD

John Young, PhD The relationship between the general public and law enforcement personnel has long been considered tenuous. In a recent study by the Bureau of Justice Statistics, members of the public reported the belief that police officers acted improperly in 19 percent of encounters across the nation (Davis, Whyde, & Langton, 2018). However, we must also acknowledge that the role of law enforcement has changed and become much more complicated over time. The days in which a police officer was solely responsible for maintaining public safety through making arrests and following rigid guidelines and procedures are long gone. Instead, police officers are now engaging in more community-based policing in efforts to better meet the needs of the public. This approach, along with other societal factors, has required officers to respond to increasingly complex situations that require significant problem-solving, critical-thinking, and of course, quick decisions to best respond to these scenarios. Essentially, society is asking more of police officers aside from simply upholding the law.

The lack of appropriately funded and comprehensive community mental health services has required officers to be fluent in case management practices and sensitive to differences in responding to persons with major mental illness. Additionally, the advancement of technology has placed police officers under more intense scrutiny in every decision they make when responding to different scenarios, with both the increased utilization of body cameras and bystanders recording incidents on cell phones. These factors partially help explain the apparent increase of media focus on allegations of police misconduct, and on the strained relationship between law enforcement and society. While police departments have made a number of internal changes in an attempt to address these aforementioned issues (e.g., focus on community policing and changing policies and procedures), social science has also contributed to research and training initiatives to help law enforcement deal with their increasingly complicated roles and responsibilities. Two areas of the social sciences primary focus related to these aforementioned factors and law enforcement have included Crisis Intervention Team (CIT) training, and training to address implicit biases. The social science contributions in each of these areas will be further summarized herein.

Crisis Intervention Team Training

Partly in response to public outcry of the shooting of a person with mental illness in Memphis, Tennessee, in 1987, Randolph Dupont of the University of Tennessee, Memphis, and Maj. Sam Cochran of the Memphis Police Department developed the Crisis Intervention Team (CIT) Memphis model of police training. The CIT model provides law enforcement personnel with crisis intervention training to respond to persons with mental illness and to increase the safety of the officers, person served, family members and citizens of the community (Dupont, Cochran, & Pillsbury, 2007). In addition, the model also acts as a pre-arrest, jail diversion for persons with mental illness. The 40-hour CIT curriculum aims to provide a common knowledge base regarding mental illness and to provide officers with the skills to recognize symptoms of mental illness and co-occurring disorders. The training also helps officers recognize a crisis, de-escalate someone in crisis, know where to take a person in crisis, employ jail diversion options, and learn how to problem solve with the local treatment systems (Florida CIT Coalition, 2005, p. 7-8). Importantly, various portions of CIT training are led by police officers, mental health professionals, mental health advocates, persons with mental illness, and family members of persons with mental illness. The inclusion of a variety of experts, rather than only being led by mental health professionals, helps increase officer engagement in the training and provides the training through a variety of perspectives.

As the CIT model has expanded over time, so has the collection of data and the research on effectiveness of the training. The Substance Abuse and Mental Health Services Administration (SAMHSA) has published a CIT guide on using data to inform practice (2018). The guide proposes three levels of data collection. The first tier includes tracking community partnerships and engagement, the CIT training program, the number of mental health calls an agency receives, and the number of mental health calls responded to by CIT officers. The guide notes that these data points are the “most basic” to collect. The second tier of data includes CIT training outcomes, call dispositions, injury rate and use of force. This tier includes the majority of social science research, most notably focused on CIT training outcomes. The third tier requires the most advanced data and includes assessing impact on jail diversion, impact on recidivism, strengths of community partnerships, treatment continuity and other advanced factors. Research on CIT outcomes has primarily utilized second tier data, including (but not limited to) a drop in arrest rates in a Florida County that implemented CIT (Franz & Borum, 2011); increase in officers’ knowledge about mental illness (Hanafi, Bahora, Demir, & Compton, 2008); increase in officer self-efficacy in responding to person with mental illness (Bahora, Hanafi, Chien, & Compton, 2008); and reduction in stigmatizing attitudes and misperceptions about mental illness (Demire, Broussard, Goulding, & Compton, 2009). Compton et al. (2014a) replicated the aforementioned results in arguably the most comprehensive CIT study to date, finding that CIT significantly improved officer’s attitudes about mental illness, self-efficacy, social distance/stigma from persons with mental illness, de-escalation abilities and referral decisions. However, the aforementioned research does not investigate one of the most central aspects of the CIT model, namely whether CIT training and improved de-escalation abilities translates into a reduction of use of force/injuries of all involved.

Measuring whether training programs positively affect law enforcement use of force in the field has proven to be notoriously difficult given a multitude of factors, such as inconsistency of coding calls as “mental health” related and outcomes, discrepancies in use of force documentation between departments/jurisdictions, and a lack of access to encounters outside of those documented by officers who have consented to do so as part of a research study. Nevertheless, many researchers have addressed officer use of force preferences through scenario-based assessment. Compton et al. (2011) provided three scenarios to a group of police officers who had completed CIT training and another group who had not. When confronted with an increasingly agitated and psychotic individual, CIT-trained officers chose to respond by using a significantly lower level of force when compared with non-CIT trained officers. Furthermore, CIT-trained officers perceived non-physical police responses as more effective, and conversely, physical responses less effective when faced with the psychotic individual. Relatively few studies have researched real world implications for reduction of officer and subject injuries and use of force with some mixed results. Research by Skeem & Bibeau (2008) in Memphis showed a decrease in officer injuries for mental health related calls after implementation of CIT (.04 to .01 per 1,000). Additionally, although no control group of non-CIT-trained officers was utilized, CIT officers used force in 6 percent of situations in which subjects were in psychiatric crises and threatened to harm themselves or others. Authors compared this frequency to a study of over 7,500 adult arrests that found police used physical force in 17 percent of cases (Garner & Maxwell, 1999), suggesting that CIT-trained officers used less force. Skeem & Bibeau also found that the use of force by CIT officers resulted in injury to the subject less often than traditional officers. Although the authors acknowledged the imperfect and indirect comparison of groups, this study highlighted potential use of force benefits of CIT training. 

More recent research by Compton et al. (2014) assessed use of force and encounter outcomes using 180 police officers. The study showed promising results in that CIT officers were more likely to use verbal engagement or negotiation as their highest level of force used when compared to non-CIT officers. Additionally, CIT-trained officers were more likely to refer a subject to mental health services or transport to a hospital rather than arrest compared to officers without CIT training. However, the overall statistical analysis of the study indicated that CIT training was not predictive of the level of force utilized. Given the inconsistencies in the research results of some of the effects of CIT training, Watson, Compton, & Draine (2017) recently addressed the question of whether CIT training represented an evidence-based policing practice. They suggested that CIT training can be considered evidence-based for officer level, and cognitive and attitude variables (e.g., reducing social stigma, improving officer knowledge of mental illness, and improving officer confidence when responding to a person in crisis). However, they note that more research is needed to determine whether CIT is evidenced-based for other types of outcomes, including potential changes in use of force in the field.

Implicit Bias Training

Although the CIT model was created after the unfortunate shooting of a person with mental illness in 1987 in order to provide officers skills to respond to persons in crisis, the shooting also had racial undertones given that the police officer was white and the subject was black. Over time, social science has reinforced that implicit values and beliefs that can act as a large contributor to behavior. As such, a significant portion of the recent research and training in this area has focused on implicit bias. Social science has produced research on implicit bias related to race in both members of the public and in law enforcement personnel, and in response, the field has begun to develop and implement training programs to address implicit bias in law enforcement. Arguably, the two most prominent training programs developed and widely implemented thus far include the Fair and Impartial Policing Program and the Counter Bias Training Simulation program. The Fair and Impartial Policing Program was primarily designed by Lori Fridell, PhD, to utilize the science behind implicit bias to offer training programs designed to educate law enforcement personnel about implicit bias and learn skills to reduce/manage biases (Fair and Impartial Policing, n.d.). The program utilizes skills that are based on scientific research in part on contact theory and exposure to counter stereotypes. The program has received positive subjective feedback from jurisdictions and participants, including personal communications from participants and satisfaction surveys completed at the conclusion of training. Additionally, an independent research company is in the process of collecting and analyzing data from the New York City Police Department related to potential effects of the Fair and Impartial Policing Program (L. Fridell, personal communication, September 16, 2019). 

The Counter Bias Training Simulation (CBT SIM) program was developed by Lois James, PhD, and Stephen James, PhD, as an evidence-based training designed to help law enforcement personnel improve awareness of what goes into their decision-making. This program is primarily based around scenario-based training conditions (e.g., simulators) that require officers to make a decision on use of force, which is followed by “…meaningful debriefing and self-evaluation session to identify and understand motivations behind their decisions to use or hold deadly force” (CBT SIM, n.d.). This training program appears to build on the research by Plante & Peruche that suggested that more exposure to simulators had the potential to reduce implicit bias (likely through counterconditioning). According to a personal communication with Lois James (September 9, 2019), researchers are currently collecting data comparing the effects of CBT SIM to a more traditional classroom-based implicit bias training, but the data collection includes twelve months of pre- and post- data so the research will not be complete for some time. In sum, while the two widely utilized programs designed to reduce implicit bias in law enforcement have been seemingly well received, the results from the ongoing research will be a notable addition to the body of literature on the effectiveness of implicit bias training for law enforcement personnel.

Although these two programs are still collecting and analyzing data on the effectiveness of the trainings, other social science research has addressed the presence and potential for change of implicit bias with a variety of subjects. A meta-analysis of research on implicit bias with over 72,000 participants suggested that implicit bias can be malleable and that certain procedures can produce limited changes in biases. The studies suggested that associated set of concepts (e.g., observing stimuli that weaken the associated bias), invoking goals/motivations (e.g., instructing participants to respond in a manner that weakened the associated bias), and taxing people’s resources (e.g., thought suppression or performing mentally effortful task) produced the largest changes in implicit biases. However, the authors also noted that there was little evidence that change in implicit biases mediated changes in explicit biases or participant’s behavior (Forscher et al., 2016). The data on the effects of implicit bias on behavior in research simulators is somewhat more mixed. Correll, Park, Judd, & Wittenbrink (2002) asked community members to utilize a simulator with shoot/don’t shoot decision points that were moderated by race and whether the target was armed. The results suggested that white participants shot correctly more quickly if the target was black but correctly shot slower if the target was white. Additionally, researchers found that both black and white participants had almost equivalent level of racial biases. This research suggested that the study participants set a lower decision criterion (e.g., whether to shoot) for black subjects than for white subjects.

Plante & Peruche (2005) took this research a step further by examining police officer decision-making related to their decision to shoot in a simulator. In the early trials, officers were more likely to shoot unarmed black subjects compared with unarmed white subjects. However, after extensive training with the program in conditions where the target’s race was not related to the presence of a weapon, the officers were able to eliminate the biased response/increased likelihood of shooting an unarmed black subject. Researchers suggested that exposure to counterconditioning was a possible explanation for the reduction of biased responses. James, James, & Vila (2016) later replicated the shoot/don’t shoot scenario procedures comparing police officers’ responses with community members. Results showed that both community members and police officers exhibited biases, but police officers outperformed community members in accuracy and overall speed in reaching the decision point. Police officers also did not set the decision criterion (e.g., whether to shoot or not) lower for black targets compared with white targets. This research suggests that although training may not affect the speed at which stereotype-incongruent targets are cognitively processed, it may affect the ultimate decision of how to respond, thereby affecting behavior.

Research has also suggested that other variables can moderate the salience and strength of racial bias in police officers. James (2017) studied whether sleep (and fatigue) affected implicit bias expressed by police officers using the Weapons Implicit Association Test (IAT) across four occasions. The results suggested that the officers’ implicit biases were not stable over time and were affected by the amount of sleep. When officers slept less prior to completing the IAT, they had stronger connections between black subjects and weapons. James proposed that based on the notion that implicit biases could be more trait-based than state-based, policymakers and researchers should investigate whether trainings designed to address those dynamic factors such as sleep and fatigue could be more effective than other trainings designed to address the underlying construct of implicit bias.

Looking Into the Future

Researchers have identified a number of areas for future research in both responding to persons with mental illness and reducing implicit biases. There has been sufficient research to suggest a number of positive officer-level knowledge outcomes related to CIT training, but the research on reduction of use of force is still somewhat mixed. Additionally, data collection related to use of force statistics, coding of incident type, and incident outcome is often inconsistent across jurisdictions, making it challenging for researchers to compare and analyze data. Nonetheless, while there has been some data to suggest that CIT could reduce use of force in simulators and paper-and-pencil scenarios, there is a lack of research to suggest whether these effects would generalize to officers’ actual response to real-world incidents. Concerning implicit bias, Forscher et al. argues that we must first improve our understanding of the causal mechanisms of implicit biases, including using more reliable and valid assessment measures and more heterogeneous samples. Research method difficulties aside, the research suggesting that the use of simulators can reduce some aspects of implicit bias is a promising start. However, it is important to note that it is unclear whether the results from simulators will actually generalize to real-life encounters in the field, where officers will be faced with the same decisions as the simulators but with more significant consequences. Similarly, the initiation of the Fair and Impartial Policing Program and the CBT SIM program are excellent starts to attempting to address implicit bias for law enforcement personnel. However, we anxiously await the ongoing data collection and analyzation regarding the effectiveness of those programs, as additional research will only help social scientists further improve the trainings and interventions we can offer to law enforcement in these areas. 


Bahora, M., Hanafi, S., Chien, V., & Compton, M. (2008). Preliminary evidence of effects of crisis intervention team training on self-efficacy and social distance. Administration and Policy in Mental Health, 35, 159-167.

Compton, M., Bakeman, R., Broussard, B., Hankerson-Dyson, D., Husbands, L., Krishan, S., Stewart-Hutto, T., D’Orio, B., Oliva, J., Thompson, N., & Watson, A. (2014a). The police-based crisis intervention team (CIT) model: 1. Effects on officers’ knowledge, attitudes, and skills. Psychiatric Services, 65 (4), 517-522.

Compton, M., Bakeman, R., Broussard, B., Hankerson-Dyson, D., Husbands, L., Krishan, S., Stewart-Hutto, T., D’Orio, B., Oliva, J., Thompson, N., & Watson, A. (2014b). 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.

Compton, M., Demir Neubert, B., Broussard, B., McGriff, J., Morgan, R., & Oliva, J. (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.

Correll, J., Park, B., Judd, C., & Wittenbrink, B. (2002). The police officer’s dilemma: Using ethnicity to disambiguate potentially threatening individuals. Journal of Personality and Social Psychology, 83 (6), 1314-1329.

Correll, J., Park, B., Judd, C., Wittenbrink, B., Sadler, M., & Keesee, T. (2007). Across the thin blue line: Police officers and racial bias in the decision to shoot. Journal of Personality and Social Psychology, 92 (6), 1006-1023.

Davis, E., Whyde, A., & Langton, L. (2018, October). Contacts between police and the public, 2015. Retrieved from https://www.bjs.gov/content/pub/pdf/cpp15.pdf.

Demir, B., Broussard, B., Goulding, S., & Compton, M. (2009). Beliefs about causes of schizophrenia among police officers before and after crisis intervention team training. Community Mental Health Journal, 45, 385-392.

Dupont, R., Cochran, S., & Pillsbury, S. (2007). Crisis intervention team core elements. The University of Memphis, Retrieved from http://cit.memphis.edu/pdf/CoreElements.pdf Fair and Impartial Policing. (n.d.). Retrieved from https://fipolicing.com/.

Florida CIT Coalition (2005). The Florida crisis intervention team (cit) program. Retrieved from http://www.floridacit.org/Florida%20CIT%20Program%20document%20for%20the%20website.pdf.

Forscher, P., Lai, C., Axt, J., Ebersole, C., Herman, M., Devine, P., & Nosek, B. (2016). A meta-analysis of change in implicit bias.

Franz, S., & Borum, R. (2011). Crisis intervention teams may prevent arrests of people with mental illnesses. Police Practice and Research, 12(3), 265-272.

Garner JH, Maxwell CD: Measuring the amount of force used by and against the police in six jurisdictions, in Use of Force by Police: Overview of National and Local Data. Washington, DC, US Department of Justice, Bureau of Justice Statistics, 1999. Available at www.ojp.usdoj.gov/bjs/abstract/ufbponld.htm.

Graham v. Connor. 490 U.S. 386. (1989).

Hanafi, S., Bahora, M., Demir, B., & Compton, M. (2008). Incorporating crisis intervention team (CIT) knowledge and skills into the daily work of police officers: A focus group study. Community Mental Health Journal, 44, 427-432.

International Association of the Chiefs of Police, Police Use of Force in America, 2001, Alexandria, Virginia, 2001.

James, L. (2017). The stability of implicit racial bias in police officers. Police Quarterly, 21 (1), 30-52.

James, L., James, S., & Vila, B. (2016). The reverse racism effect: are cops more hesitant to shoot black than white suspects? Criminology & Public Policy, 15 (2).

Plant, A., & Peruche, M. (2005). The consequences of race for police officers’ responses to criminal subjects. Psychological Science, 16 (3), 180-203.

Skeem, J., & Bibeau, L. (2008). How does violence potential relate to crisis intervention team responses to emergencies? Psychiatric Services, 59, 201-204.

Substance Abuse and Mental Health Services Administration, Crisis Intervention Team (CIT) Methods for Using Data to Inform Practice: A Step-by-Step Guide. HHS Pub. No. SMA-18-5065. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2018.

Watson, A., Compton, M., Draine, J. (2017). The crisis intervention team (CIT) model: An evidence-based policing practice? Behavioral Sciences and the Law, 35 (5), 431-441.