The policing profession has been under intense scrutiny over the past several years.1 The 2014 shooting of Michael Brown in Ferguson, Missouri, brought allegations of racially motivated policing to the mainstream public conversation about the trustworthiness of police.2 At about the same time, the police profession began to pay attention to the science of implicit bias, in particular, the idea that officers might have biases that influence their behavior, judgments, and decisions, even if they are not explicitly or overtly biased against a particular group of people.3 Police agencies need a resolution to this dilemma, and training is a common approach used by police departments to maneuver around public criticism.4 Historically, when a social problem arises, society clamors for police training to deal with the crisis. In this sense, the “crisis of legitimacy” is no different. In response to broad concerns about racially motivated policing, implicit bias training is becoming a staple among police departments across the United States.
Implicit bias training aims to increase fairness in officer decision-making and to enhance the outcomes of police-citizen encounters. The problem, however, is that no empirical evidence exists on the impact of implicit bias training on officer decision-making in the field, whether officers who are trained in implicit bias are perceived to be fairer by citizens, which training modality (e.g., classroom vs. simulation based) is most effective in producing persistent changes in police behavior, or how long training effects last. Randomized controlled trials (RCTs) are desperately required to inform policy and practice related to implicit bias training. The critical question is: Can implicit bias training reduce police officer bias, improve officer fairness in behavior, and ultimately promote public trust in police legitimacy?
Understanding Implicit Bias
Implicit bias is more subtle than explicit bias—which is conscious attitudes, stereotypes, and beliefs that individuals are aware of and “own” as part of their worldview. Implicit bias is made up of unconscious or semiconscious attitudes that influence behavior.5 Explicit bias is a conscious choice, whereas implicit bias can be passively acquired with no conscious awareness. For example, a police officer might not consider themselves to be prejudiced against many groups of people—such as people of color, members of the homeless community, individuals who are substance dependent, or people with mental illness—but may nevertheless implicitly associate members of these groups with either danger or weakness because of stigma that exists within police culture.6 As such, implicit bias is not denying or ignoring a prejudice, but being unaware that the prejudice or bias exists at all.
Up until relatively recently, researchers believed that implicit bias was stable or “trait like,” established in childhood and changed only under extreme circumstances.7 More recently, however, researchers have suggested that implicit attitudes are susceptible to change by simply altering one’s environment.8 For example, if a police officer works in a high-crime neighborhood that is predominantly African American, he or she might implicitly associate African Americans with crime. If, however, the officer transfers to a neighborhood with lower crime rates that is also predominantly African American, his or her implicit biases are likely to be reduced through a change in associations between African Americans and crime. This could then fundamentally shift the officer’s beliefs about African Americans and even result in behavioral changes.9 In other words, exposure to counter-stereotypical association reduces the strength of the negative association or even results in a positive association.
Police officers are likely to be just as susceptible to implicit bias as any other professional group—perhaps more so, given the nature of their work, which often focuses on negative aspects of human behavior. It is, however, critical that officers make decisions based on legally defined criminal behavior and policy and not be influenced by individual traits that (by themselves) do not dictate criminal behavior. Even if a group as a whole is more prone to crime, a particular individual from that group cannot be treated as a likely criminal; doing so would be discriminatory.
Perhaps the most salient example of implicit bias among police comes from widespread allegations of racially motivated policing. Cases in which police officers shoot unarmed black citizens typically result in public outcry, with the belief that officers were driven by racial bias. Officers, on the other hand, tend to assert that they were, in fact, responding to threat cues that they perceived. Interestingly, both groups might be correct. An illustrative example comes from George Fachner and Steven Carter’s threat perception failures (TPF).10 These researchers showed that implicit bias can influence an officer’s decision to shoot or not to shoot, by influencing the officer’s perception of reality. In this case, an officer incorrectly perceives that the suspect poses a deadly threat, due to the misperception of an object (such as a cellphone) or an action (such as reaching for a cellphone). When these researchers analyzed officer-involved shootings by the Philadelphia, Pennsylvania, Police Department (PPD), they found that TPF more frequently explained the shootings of unarmed African Americans compared to unarmed whites, speculating that officers may be subconsciously “on guard” with African American suspects.
Police Implicit Bias Training
Implicit bias is a universal human condition, and police officers are very unlikely to be exempt from it without directly addressing it. As mentioned previously, police training is commonly implemented as a response to changes in societal problems or pressure from the public. As societal pressure mounts, police agencies will train an entire organization before an incident arises in the hopes of alleviating public scrutiny during and after a critical incident. More progressive departments may choose to implement training without societal pressure to do so to help prevent these issues from arising. Police chiefs often espouse training as evidence of progressiveness, accountability, and innovation, in addition to using training as a deflection for scrutiny.
Due to this rush to train, law enforcement implicit bias training across the United States can be haphazardly designed and implemented before any evidence even suggests that the training will be effective.11 Agencies fall into several different categories for how they design and implement training. The first method is to bring in an outside expert to come in and teach a class. The second method is to send an internal trainer to learn about the course content via a train-the-trainer (TOT) course, who will then return and conduct training at his or her agency. The third method is to ask another agency for a copy of the course materials, assign it to an officer, and have that officer create an internal training course. These three methods are typical of how all training is diffused across the police profession, which aligns with diffusion of innovation research and how information is shared.12
The Society of Prevention Research argues that evidence-based interventions (EBIs) should be tested for effectiveness before an EBI is scaled up across a profession.13 Unlike use-of-force, vehicle, or arrest training, which focuses on teaching actual behaviors to officers, implicit bias training classes focus on giving officers information in the hopes of altering behavior and consequently improving the public’s perceptions of the police. Many implicit bias trainings have not even reached the EBI standard yet, and none have evaluated behavioral modification following training.
Figure 1 demonstrates this leap in logic that police leaders tend to make when implementing public perception–driven training. The first step assumes officers will change behavior after obtaining new knowledge. The second step assumes the behavioral changes will impact the public’s perception of the police. Often the biggest leap in logic occurs when managers believe the police training will lead to a change in public perception, as this assumption surpasses the first step entirely. However, training effects will never be truly understood without evaluating a training program to see if it (1) creates positive behavioral change in police and (2) if those behavioral changes improve the public’s perception of police.
A notable few studies have attempted to move beyond these assumptions in other domains of police training such as procedural justice.14 However, no experimental evaluations of implicit bias training have yet been conducted. In the current era’s heightened public sensitivity around racially motivated policing, these trainings are critically important—providing they can fulfill the assumptions outlined in Figure 1.
The Effectiveness of Implicit Bias Training—The Known and the Unknown
A recent meta-analysis of the effectiveness of different procedures to change implicit bias synthesized evidence from 426 studies using non-police samples. The authors found that most procedures produced some changes in implicit bias. They concluded that procedures that
associate sets of concepts, invoke goals or motivations, or tax people’s mental resources produced the largest changes in implicit bias, whereas procedures that induced threat, affirmation, or specific moods/emotions produced the smallest changes.15
Most of the procedures tested had a greater impact on implicit measures than on behavioral measures. Thus, changes in subconscious attitudes were observed (when tested on measures such as the Implicit Association Test) without any real changes in behavior, which is troubling for the first assumption of Figure 1.
Several studies have examined changes in implicit bias among police officers during laboratory testing. A 2005 study found that police participants showed bias against black suspects in shooting decisions at the start of testing (i.e., they were significantly more likely to mistakenly press a button labeled “shoot” when rapidly presented with an image of an unarmed black suspect than an unarmed white suspect).16 This bias was significantly reduced, however, when officers were exposed to approximately 15 minutes of repeated trials in which the race of the suspect was not predictably related to whether they were armed or unarmed. These findings suggest that counter-stereotyping may reduce implicit biases, or (more importantly for policing) the effects of these biases on behavior.
A concerning possibility that has been raised by social justice scholars is that implicit bias training may have an unintended negative effect by increasing the expression of bias. Researcher Joshua Correll has suggested “there are a number of very compelling studies that show that if you just ask somebody to try really hard not to show racial bias you can actually inadvertently increase racial bias.”17 For example, a negative effect was found in 2015 when researchers tested participants following training to increase awareness of stereotyping.18 The study showed that this type of training may actually normalize stereotyping by promoting the notion that implicit bias is common and expected, thereby decreasing trainees’ guilt and motivation to overcome it. Furthermore, some research discusses a “rebound effect” whereby actively trying to inhibit stereotyping results in bias eventually resurging with greater insistence than it would had it not been suppressed.19 This possibility could have catastrophic implications for law enforcement if implicit bias training programs are implemented without empirical evaluation.
A Potential Solution
The authors argue that training evaluations should meet the standards suggested by the Society for Prevention Research. Police training is an intervention, and social interventions can create harm.20 The Society for Prevention Research lays out a framework for the scaling of an intervention. The first step is to run a small pilot program to test for the effects of the training. If the pilot test reveals positive results, then the intervention is implemented under optimal conditions as an RCT, monitored by researchers or supervisors to ensure proper application of standards. This is an efficacy trial. If the efficacy trial demonstrates positive results, then an effectiveness trial is implemented under real-world conditions. An effectiveness trial is “delivered under the same types of conditions as one would expect in community institutions.”21 The Society for Prevention Research advocates for both efficacy and effectiveness trials before an intervention is scaled up or widely adopted.
In line with the Society for Prevention Research’s recommendations, several RCTs are currently being initiated that will evaluate the effectiveness of implicit bias training for reducing bias and promoting fairness in officer behavior. One of these RCTs is being funded by the National Institute of Justice (NIJ) and will test the impact of classroom versus simulation-based implicit bias training on a variety of outcome measures. These measures include fairness in officer decision-making behavior (assessed via body-worn camera footage), citizen perceptions of police bias (assessed via citizen complaints), arrestee perceptions of police bias (assessed via anonymous survey), and police perceptions of training effectiveness (assessed via surveys and focus groups). Using a true experimental design, the study will be used to evaluate (1) whether implicit bias training has a positive impact on study outcomes; (2) whether differences in study outcomes are observed between the different types of implicit bias training (classroom vs. simulation); and (3) how long training effects last. In addition to these measures, police perceptions of training effectiveness will also be gathered, as police buy-in could be critical in the widespread adoption of different types of implicit bias training.
Conclusion
The lack of impartial, objective information on the impact of implicit bias training leaves officers, their supervisors, and the public in the dark. Law enforcement training in general is somewhat of a “black box,” with very little empirical connection between how officers are trained and how they behave on the street. Regarding implicit bias training, no connection has yet been made between training or training modality and increased fairness in officer decision making behavior. Despite this, implicit bias training is growing in popularity; becoming a staple in academies and departments across the United States. In part, this is likely due to federal recommendations such as those by former President Obama’s Task Force on 21st Century Policing. In addition, people tend (perhaps rationally) to assume that teaching officers about implicit bias and how to overcome it is a good idea.
This, however, is just an assumption, because, as stated, no RCTs have yet been conducted to rigorously evaluate the impact of implicit bias training. Furthermore, if implicit bias training is effective, police departments have options available to them regarding training modality. RCTs testing different types of implicit bias training modalities are required to help identify best practices for implicit bias training. Should RCTs discover that implicit bias training is effective in promoting fairness in officer behavior and improving public perception of police legitimacy, then those findings provide justification for the widespread adoption of such programs. Should RCTs discover that implicit bias training is not effective, the insight gained from such results could be highly beneficial for departments’ cost savings and for researchers striving to improve such training programs. To the latter point, conducting focus groups and surveys with officers who have been trained on implicit bias could provide insight into pathways for such improvement.
Notes:
1 Justin Nix et al., “A Bird’s Eye View of Civilians Killed by Police in 2015: Further Evidence of Implicit Bias,” Criminology & Public Policy 16, no. 1 (2017): 309–340; Eugene. A. Paoline, Jacinta M. Gau, and William Terrill, “Race and the Police Use of Force Encounter in the United States,” The British Journal of Criminology 58, no. 1 (2016): 54–74
2 William Terrill, “Deadly Force: To Shoot or Not to Shoot,” Criminology & Public Policy 15, no. 2 (2016): 491–496.
3 Nilanjana Dasgupta, “Implicit Attitudes and Beliefs Adapt to Situations: A Decade of Research on the Malleability of Implicit Prejudice, Stereotypes, and the Self-Concept,” in Advances in Experimental Social Psychology Vol. 47, eds. Patricia Devine, Ashby Plant (San Diego, CA: Academic Press, 2013), 233–279.
4 Renee J. Mitchell, “Procedural Justice Training: The Elixir Vitae or a Cure That Can Harm?” Australia & New Zealand Journal of Evidence Based Policing 1, no. 2 (2016): 22–27.
5 Dasgupta, “Implicit Attitudes and Beliefs Adapt to Situations.”
6 Ni He, Jihong Zhao, and Carol A. Archbold, “Gender and Police Stress: The Convergent and Divergent Impact of Work Environment, Work-Family Conflict, and Stress Coping Mechanisms of Female and Male Police Officers,” Policing: An International Journal of Police Strategies & Management 25, no. 4 (2002): 687–708.
7 Timothy D. Wilson, Samuel Lindsey, and Tonya Y. Schooler, “A Model of Dual Attitudes,” Psychological Review 107, no. 1 (2000): 101.
8 Irene V. Blair, “The Malleability of Automatic Stereotypes and Prejudice,” Personality and Social Psychology Review 6, no. 3 (2002): 242–261; Nilanjana Dasgupta, “Mechanisms Underlying the Malleability of Implicit Prejudice and Stereotypes,” in Handbook of Prejudice, Stereotyping, and Discrimination, ed. T.D. Nelson (Mahwah, NJ: Erlbaum, 2009), 267–284.; Bertram Gawronski and Galen V. Bodenhausen, “Associative and Propositional Processes in Evaluation: An Integrative Review of Implicit and Explicit Attitude Change,” Psychological Bulletin 132, no. 5 (2006): 692.
9 Dasgupta, “Implicit Attitudes and Beliefs Adapt to Situations.”
10 George Fachner and Steven Carter, An Assessment of Deadly Force in the Philadelphia Police Department (Washington, DC: Office of Community Oriented Policing Services, U.S. Department of Justice, 2015).
11 Mark Sedevic. “Procedural Justice & Police Legitimacy Training in Chicago: Reaping the Benefits of the Golden Rule,” 9-1-1, October 10, 2016.
12 Everett M. Rogers, Diffusions of Innovation (New York, NY: Simon and Schuster, 2003).
13 Brian R. Flay et al., “Standards of Evidence: Criteria for Efficacy, Effectiveness and Dissemination,” Prevention Science 6, no. 3 (2005): 151–175.
14 Levin Wheller et al., The Greater Manchester Police Procedural Justice Training Experiment: Technical Report (London, UK: The College of Policing, 2013).
15 Patrick Forscher et al., A Meta-Analysis of Change in Implicit Bias (2016), 3.
16 E. Ashby Plant and B. Michelle Peruche, “The Consequences of Race for Police Officers’ Responses to Criminal Suspects,” Psychological Science 16, no. 3 (2005): 180–183.
17 Martin Kaste, “Police Officers Debate Effectiveness of Anti-Bias Training,” NPR, April 6, 2016.
18 Michelle M. Duguid and Melissa C. Thomas-Hunt, “Condoning Stereotyping? How Awareness of Stereotyping Prevalence Impacts Expression of Stereotypes,” Journal of Applied Psychology 100, no. 2 (2015): 343.
19 C. Neil Macrae et al., “Out of Mind but Back in Sight: Stereotypes on the Rebound,” Journal of Personality and Social Psychology 67, no. 5 (1994): 808.
20 Joan McCord, “Cures That Harm: Unanticipated Outcomes of Crime Prevention Programs,” The Annals of the American Academy of Political and Social Science 587, no. 1 (2003): 16–30.
21 Denise C. Gottfredson et al., “Standards of Evidence for Efficacy, Effectiveness, and Scale-up Research in Prevention Science: Next Generation,” Prevention Science 16, no. 7 (2015): 893–926.
Please cite as:
Renee J. Mitchell and Lois James, “Addressing the Elephant in the Room: The Need to Evaluate Implicit Bias Training Effectiveness for Improving Fairness in Police Officer Decision-Making,” Police Chief Online, November 28, 2018.
