Microsoft Word - CJRL Murphy F18.docx IS IT RECORDING?—RACIAL BIAS, POLICE ACCOUNTABILITY, AND THE BODY-WORN CAMERA ACTIVATION POLICIES OF THE TEN LARGEST METROPOLITAN POLICE DEPARTMENTS IN THE USA Julian R. Murphy0F * In recent years, there has been a growing belief that the pressing problem of racial bias in policing might be ameliorated by a technical fix—namely, police body-worn cameras. Accordingly, body-worn cameras have been intro- duced in police departments across the country, giving rise to a variety of different internal guidelines and policies. This Note surveys the body-worn camera policies of the ten largest metropolitan police departments in the United States in order to assess their relative effectiveness at combatting racial bias. Particular attention is paid to “activation” requirements, which specify the sorts of events police officers are required to record on their cameras. The survey shows that, at present, many body-worn camera policies are not appropriately cali- brated to successfully reduce racial bias in policing. In par- ticular, this Note suggests that many current body-worn camera policies do not adequately target two different strains of racial bias: implicit racial bias and deliberate racial profil- ing. This Note concludes by offering a draft activation policy for potential adoption by police departments to better use body-worn cameras to reduce racially biased police practices. I. INTRODUCTION ............................................................. 143 II. EXISTING LITERATURE ON BODY-WORN CAMERA ACTIVATION POLICIES .................................................. 145 III. THE PROBLEM: RACIALLY BIASED POLICING................ 147 * Postgraduate Public Interest Fellow, Columbia Law School. Thanks to Mary Fan, L. Song Richardson, and all the participants in the Columbia Human Rights Institute’s Student Paper Symposium. Special thanks to the editors of the Columbia Journal of Race and Law and to former Notes Editor Corben Green. All errors are my own. 142 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 A. Deliberate Racial Profiling .................................... 148 B. Implicit Racial Bias ............................................... 150 IV. A SOLUTION? BODY-WORN CAMERAS AND THREE THEORIES OF CHANGE .................................................. 153 A. Body-Worn Cameras and the Deterrence Theory of Change ................................................................... 154 B. Body-Worn Cameras and the Accountability Theory of Change ............................................................... 155 1. Wide Acceptance of the Accountability Theory .............................................................. 156 C. Body-Worn Cameras and the Training Theory of Change ................................................................... 157 D. A Prerequisite for Change: Footage Is Reviewed..... 158 V. ACTIVATION POLICIES IN THE TEN LARGEST METROPOLITAN POLICE DEPARTMENTS ....................... 159 A. Limited Discretion Activation Policies .................. 160 B. Minimal Discretion Activation Policies ................. 162 C. A Trend Toward Reducing Police Officer Discretion ............................................................... 163 D. Gradations of Discretion ........................................ 163 VI. ASSESSING ACTIVATION POLICIES AGAINST THE STATED GOAL OF REDUCING RACIALLY BIASED POLICING ........ 164 A. Activation Requirements and Deliberate Racial Profiling ................................................................. 165 1. The Advantages of Continuous Recording ...... 165 2. Obstacles to Continuous Recording ................. 168 3. Recording All Police-Civilian Interactions ...... 168 B. Activation Requirements and Implicit Racial Bias ............................................................. 170 C. Activation Requirements and “Consensual” Interactions ............................................................ 173 1. Racial Bias in “Consensual” Interactions ........ 174 2. Activation Policies and “Consensual” Interactions ...................................................... 175 D. Proposal for a Minimal Discretion Activation Policy .................................................... 176 VII. CONCLUSION ................................................................ 180 VIII. APPENDIX A: STATED OBJECTIVES OF POLICE DEPARTMENT BODY-WORN CAMERA POLICIES ............ 183 No. 4:141] IS IT RECORDING? 143 IX. APPENDIX B: DEGREE OF DISCRETION AFFORDED TO OFFICERS IN ACTIVATION POLICIES ............................. 184 X. APPENDIX C: ACTIVATION POLICIES REQUIRING MANDATORY RECORDING OF “CONSENSUAL” INTERACTIONS .............................................................. 185 XI. APPENDIX D: METHODOLOGY ....................................... 186 A. Choice of Sample .................................................... 186 B. Obtaining the Policies............................................ 187 C. Specific Methodology: Appendix A ........................ 188 D. Specific Methodology: Appendix B ........................ 188 E. Specific Methodology: Appendix C ........................ 189 I. INTRODUCTION Racial bias in policing has been at the forefront of the national public debate since at least the 2014 shooting of Michael Brown in Ferguson and the ensuing Department of Justice investigation.1 The numerous subsequent incidents2 of unarmed Black3 people dying at the hands of police have served, if nothing else, to keep the issue in the news.4 As the 1 See Linda Sheryl Greene, Before and After Michael Brown— Toward an End to Structural and Actual Violence, 49 WASH. U. J.L. & POL’Y 1, 3–4, 9–16 (2015) (providing an account of how Michael Brown’s death may be seen to fit within broader trajectories of racialized police violence). 2 See Roger A. Fairfax, Jr., The Grand Jury and Police Violence Against Black Men, in POLICING THE BLACK MAN: ARREST, PROSECUTION, AND IMPRISONMENT 209 (Angela J. Davis ed., 2017) (overviewing police kill- ings of Black men that have reached the grand jury stage); see also KIMBERLÉ WILLIAMS CRENSHAW & ANDREA J. RITCHIE, SAY HER NAME: RESISTING POLICE BRUTALITY AGAINST BLACK WOMEN (2015), http://static 1.squarespace.com/static/53f20d90e4b0b80451158d8c/t/560c068ee4b0af26f727 41df/1443628686535/AAPF_SMN_Brief_Full_singles-min.pdf [https://perma. cc/BF4C-8V69] (providing an important corrective to the male-focused public narrative around these types of events by reporting on police violence against Black women). 3 I use the adjectives “Black” and “African American” inter- changeably for the same reason other writers in this field do, namely, be- cause police processes do not differentiate between “types” of Blackness. See, e.g., Devon W. Carbado, Blue-on-Black Violence: A Provisional Model of Some of the Causes, 104 GEO. L.J. 1479, 1480 n.1 (2016) (explaining the author’s semantic choices along similar lines). 4 This Note focuses primarily on the problem of racial bias against Black people because it is this particular form of racism that catalyzed the body-worn camera movement. This is not to ignore or diminish the effects 144 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 country looks to ways to combat this invidious problem, one “solution” has risen to particular prominence: police body- worn cameras.5 Such is the momentum behind these devices that the movement has been described as “a police body cam- era revolution.”6 Police departments across the country have piloted and mandated body-worn cameras, giving rise to a variety of different guidelines and policies. This Note sur- veys the body-worn camera policies of the ten largest metro- politan police departments in the United States: New York, Chicago, Los Angeles, Philadelphia, Houston, Washington, D.C., Dallas, Phoenix, Baltimore, and Miami-Dade. Whilst each department’s policy covers everything from camera cleaning and maintenance to metadata storage, this Note focuses solely on “activation” requirements—that is, the policy provisions specifying events that officers are required to record on their cameras.7 Though activation requirements are relatively understudied, the only study conducted to date on the scope of officer discretion in activation highlights their importance. The study found that giving officers discretion as to when to activate their cameras correlated with signifi- cantly higher incidents of officer use of force.8 It is clear, then, that activation requirements are capable of affecting police behavior. This Note’s ultimate aim is to assess the relative merits of different activation policies as against one of racially-biased policing on other minority communities. On the dangers of “Black male exceptionalism,” see Paul Butler, Black Male Exceptionalism? The Problems and Potential of Black Male-Focused Interventions, 10 DU BOIS REV. 485 (2013). 5 See Howard M. Wasserman, Moral Panics and Body Cameras, 92 WASH. U. L. REV. 831, 832 (2015) (“[O]ne significant policy suggestion has emerged from the [Ferguson] controversy: equipping police officers with body cameras.”). 6 Mary D. Fan, Justice Visualized: Courts and the Body Camera Revolution, 50 U.C. DAVIS L. REV. 897, 901 (2017). 7 This Note focuses on activation policies rather than activation functionality on body-worn cameras. When discussing activation policies, however, it is important to know that there can be a slight margin for er- ror because some body-worn cameras have the capacity to retrieve thirty seconds of video, but not audio, from the time immediately prior to an officer commencing recording. See, e.g., Operating Modes, AXON http://help.axon .com/hc/en-us/articles/221132167-Operating-modes [https://perma.cc/WV2V-86 3N]. 8 See Barak Ariel et al., Wearing Body Cameras Increases Assaults Against Officers and Does Not Reduce Police Use of Force: Results from a Global Multi-Site Experiment, 13 EUR. J. CRIMINOLOGY 744 (2016). No. 4:141] IS IT RECORDING? 145 of the primary stated goals of the body-worn camera movement: reducing racially biased policing. The discussion will proceed in five parts. Part II comprises a brief review of the existing literature on body- worn camera activation policies. Part III describes the prob- lem of racial bias in policing and distinguishes two different forms of racial bias—deliberate racial profiling and implicit racial bias. As this Note will describe, deliberate racial pro- filing occurs when a police officer makes a conscious decision to treat someone less favorably because of their race, while implicit racial bias describes subconscious discriminatory behavior. Part IV outlines three theories of body-worn cam- eras as “racial change agents”9—a deterrence theory, an accountability theory, and a training theory. In the application and analysis of the above theoretical framework, in Part V, this Note begins by com- paring the activation policies of the ten largest metropolitan police departments.10 The comparison shows that activation requirements vary significantly between departments, with some departments requiring officers to record most police- civilian interactions and other departments affording their officers more discretion. Part VI assesses the merits of the different policies when measured against the goal of reducing racially biased policing. This Note argues that the most ef- fective way to use body-worn cameras to reduce racial bias would be to mandate that the cameras be activated during all police-civilian interactions, with few exceptions. This re- form proposal is reified in a draft activation policy. Finally, this Note proposes that any attempt at reforming body-worn camera activation policies must involve input from affected communities. II. EXISTING LITERATURE ON BODY-WORN CAMERA ACTIVATION POLICIES Much has been written about body-worn cameras in the short time they have been in operation, both by way of 9 This terminology comes from Rashawn Ray et al., Can Cameras Stop the Killings? Racial Differences in Perceptions of the Effectiveness of Body-Worn Cameras in Police Encounters, 32 SOC. F. 1032, 1035 (2017). 10 In addition to the discussion in Part V, the results of this study are tabulated in Appendixes A, B, and C. The methodology is described in Appendix D. 146 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 commentary and empirical studies.11 The empirical studies to date have largely examined the effects of body-worn cam- eras on rates of police use of force and instances of civilian complaints.12 Commentary has tended to focus on questions relating to Fourth Amendment compatibility,13 privacy,14 da- ta retention,15 and public access to body-worn camera foot- age.16 Little, however, has been written specifically on activation policies.17 If the basic hope for body-worn cameras 11 See, e.g., Timothy I.C. Cubitt et al., Body-Worn Video: A Systematic Review of Literature, 50 AUSTRALIAN & N.Z. J. CRIMINOLOGY 379 (2017) (providing a helpful summary of the literature). 12 The major United States empirical studies to date are from Rialto, Mesa, Phoenix, Orlando, and Washington, D.C. See Barak Ariel et al., The Effect of Police Body-Worn Cameras on Use of Force and Citizens’ Complaints Against the Police: A Randomized Controlled Trial, 31 J. QUANTITATIVE CRIMINOLOGY 509 (2015); Justin T. Ready & Jacob T. N. Young, The Impact of On-Officer Video Cameras on Police–Citizen Contacts: Findings from a Controlled Experiment in Mesa, AZ, 11 J. EXPERIMENTAL CRIMINOLOGY 445 (2015); CHARLES M. KATZ ET AL., CTR. FOR VIOLENCE PREVENTION & CMTY. SAFETY, EVALUATING THE IMPACT OF OFFICER WORN BODY CAMERAS IN THE PHOENIX POLICE DEPARTMENT (2014); Wesley G. Jennings et al., Evaluating the Impact of Police Officer Body- Worn Cameras (BWCs) on Response-to-Resistance and Serious External Complaints: Evidence from the Orlando Police Department (OPD) Experience Utilizing a Randomized Controlled Experiment, 43 J. CRIM. JUST. 480 (2015); David Yokum et al., Evaluating the Effects of Police Body-Worn Cameras: A Randomized Controlled Trial (Oct. 20, 2017) (unpublished working paper), https://bwc.thelab.dc.gov/TheLabDC_MPD_ BWC_Working_Paper_10.20.17.pdf [https://perma.cc/99N6-RXRU]. 13 See, e.g., Julian R. Murphy, Comment, Chilling: The Constitutional Implications of Body-Worn Cameras and Facial Recognition Technology at Public Protests, 75 WASH. & LEE L. REV. ONLINE 1 (2018). 14 See, e.g., Bryce Clayton Newell, Collateral Visibility: A Socio- Legal Study of Police Body-Camera Adoption, Privacy, and Public Disclosure in Washington State, 92 IND. L.J. 1329 (2017). 15 See, e.g., Tod Newcombe, Body Worn Camera Data Storage: The Gorilla in the Room, GOVTECH (Sept. 9, 2015), http://www.govtech.com/ dc/articles/body-worn-camera-data-storage-the-gorilla-in-the-room.html [https://perma.cc/35AV-SWT6]. 16 See, e.g., Mary D. Fan, Privacy, Public Disclosure, Police Body Cameras: Policy Splits, 68 ALA. L. REV. 395 (2016). 17 For existing literature touching on body-worn camera activation policies, see THE POLICING PROJECT, N.Y. UNIV. SCH. OF LAW, REPORT TO THE NYPD SUMMARIZING PUBLIC FEEDBACK ON ITS PROPOSED BODY-WORN CAMERA POLICY 12–15 (2017); Fanny Coudert et al., Body-Worn Cameras for Police Accountability: Opportunities and Risks, 31 COMPUTER L. & SECURITY REV. 749, 758–59 (2015); LINDSAY MILLER ET AL., POLICE EXEC. RESEARCH FORUM, IMPLEMENTING A BODY-WORN CAMERA PROGRAM: RECOMMENDATIONS AND LESSONS LEARNED 12–14 (2014); Wasserman, supra No. 4:141] IS IT RECORDING? 147 is that, when they are recording, they will serve as an oversight mechanism to improve police behavior, then surely our first question should be: When will they be recording? Indeed, only one scholar has discussed the topic in any depth. Over the course of two articles in 2016 and 2017, Mary Fan published the results of the first thorough compar- ison of police body-worn camera policies in the United States.18 Comparing forty-two municipal police departments, Fan concluded that the policies differ widely with respect to privacy and release of footage but that there is a near consensus as to what must be recorded in the first place.19 Building on Fan’s research, I conducted an updated comparison of body-worn camera activation policies with particular attention to the way in which these policies might curb racially biased policing. This Note thus constitutes only the second scholarly comparison of body-worn camera activa- tion policies and the first comprehensive exploration of the potential interaction of activation policies and racial bias. Before this comparison can occur, however, it is helpful to unpack exactly what “racial bias” means. III. THE PROBLEM: RACIALLY BIASED POLICING A perception of widespread racial bias in policing— particularly in practices such as stop-and-frisk programs targeting Black pedestrians,20 the Ferguson Police Depart- ment’s “predatory policing” of Black residents,21 and the note 5, at 842; Kelly Freund, Note, When Cameras Are Rolling: Privacy Implications of Body-Mounted Cameras on Police, 49 COLUM. J.L. & SOC. PROBS. 91, 110–12, 128–30 (2015); Karson Kampfe, Note, Police-Worn Body Cameras: Balancing Privacy and Accountability Through State and Police Department Action, 76 OHIO ST. L.J. 1153, 1175–78, 1196 (2015); and Diana Taylor, Note, Body Cameras for Police: When Should They Be Turned Off?, 6 L.A. PUB. INT. L.J. 42, 54–61 (2015). 18 Fan, supra note 6; Fan, supra note 16. 19 Fan, supra note 6, at 902, 929–34; Fan, supra note 16, at 426–30. 20 See Floyd v. City of New York, 959 F. Supp. 2d 540 (S.D.N.Y. 2013) (finding racially discriminatory stop-and-frisk practices in New York City to be unconstitutional). 21 See CIVIL RIGHTS DIV., U.S. DEP’T OF JUSTICE, INVESTIGATION OF THE FERGUSON POLICE DEPARTMENT 2 (2015) (“Ferguson’s police and mu- nicipal court practices both reflect and exacerbate existing racial biases, including racial stereotypes. Ferguson’s own data establish clear racial disparities that adversely impact African Americans. The evidence shows that discriminatory intent is part of the reason for these disparities.”); see 148 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 nationwide incidence of police shootings of unarmed Black civilians22—has significantly propelled the body-worn camera movement. In response to these vastly different manifesta- tions of racialized policing, the proposed solution has been largely the same: body-worn cameras.23 Although this policy response appears to be well-intentioned, it is problematic in its failure to account for two very different causes of discrim- inatory police practices: deliberate racial profiling and implicit racial bias. The distinction is important because the different kinds of racial bias will not necessarily be amenable to the same reform policies—they may require separate and distinct ameliorative strategies.24 Thus, in order for a body- worn camera policy to effectively respond to racialized policing, it must address both deliberate racial profiling and implicit racial bias. To envisage how this might be possible, we must first better understand the difference between deliberate and implicit racial bias. A. Deliberate Racial Profiling Deliberate racial profiling—also known as explicit or conscious bias—is the process by which people are intention- ally treated differently because of their race. In the policing context, a paradigmatic example is the practice of some United States police departments in the 1980’s and 1990’s of purposefully targeting Black and Latinx people for drug- also Devon W. Carbado, From Stopping Black People to Killing Black People: The Fourth Amendment Pathways to Police Violence, 105 CALIF. L. REV. 125, 135 (2017) (discussing the racial dimensions to “predatory polic- ing” like that previously practiced by the Ferguson Police Department). 22 See sources cited supra note 2. 23 See, e.g., Kami Chavis Simmons, Body-Mounted Police Cameras: A Primer on Police Accountability vs. Privacy, 58 HOW. L.J. 881, 887 (2015) (“Although body-mounted cameras could deter multiple types of impermissi- ble conduct, these cameras could be extremely helpful in deterring racial profiling.”); see also Coudert et al., supra note 17, at 750, 756–57 (“[I]t is hoped that [body-worn cameras] will act as deterrent against . . . discrimi- nation by police officers . . . .”); Katherine B. Spencer et al., Implicit Bias and Policing, 10 SOC. & PERSONALITY PSYCHOL. COMPASS 50, 58 (2016) (“Cameras also hold potential to have a deterrent effect on [racial] bias in policing.”). 24 See Christine Jolls & Cass R. Sunstein, The Law of Implicit Bias, 94 CALIF. L. REV. 969, 973 (2006) (advocating for a practice “of ‘debi- asing’ actors through legal strategies that are designed to counteract bias- es of various sorts across a variety of domains”). No. 4:141] IS IT RECORDING? 149 related investigation, a practice illuminated by the mention of race in police department training materials.25 Another notorious example is the New York Police Department (NYPD) and Central Intelligence Agency’s post-9/11 strategy of surveilling and questioning men of Middle Eastern appearance at a higher rate than members of other racial groups.26 Today, many scholars view deliberate racial profil- ing as less of a concern than implicit racial bias.27 Neverthe- less, deliberate racial profiling remains a problem in policing and criminal justice and is of significant concern to the general public.28 25 See David A. Harris, Racial Profiling Revisited: “Just Common Sense” in the Fight Against Terror?, 17 CRIM. JUST. 36, 38–39 (2002) (de- scribing the federal Drug Enforcement Administration’s traffic stop poli- cies and intelligence in the 1980’s which explicitly “blamed trafficking in particular drugs on identified ethnic groups”). 26 See U.N. Human Rights Comm., Concluding Observations on the Fourth Report of the United States of America, ¶ 7, U.N. Doc. CCPR/C/USA/CO/4 (Apr. 23, 2014) (“[T]he Committee remains concerned about the practice of racial profiling and surveillance by law enforcement officials targeting certain ethnic minorities and the surveillance of Muslims, undertaken by the Federal Bureau of Investigation (FBI) and the New York Police Department (NYPD), in the absence of any suspicion of wrongdoing.”); see also JACK GLASER, SUSPECT RACE: CAUSES AND CONSEQUENCES OF RACIAL PROFILING 127–60 (2014) (describing racial pro- filing of Middle Eastern men in post-9/11 counterterrorism policies). 27 See, e.g., Devon W. Carbado & Patrick Rock, What Exposes African Americans to Police Violence?, 51 HARV. C.R.-C.L. L. REV. 159, 162 (2016) (“[C]onscious racial animosity likely only accounts for a small per- centage of racially-motivated conduct.”); Jerry Kang et al., Implicit Bias in the Courtroom, 59 UCLA L. REV. 1124, 1126 (2012) (“The problems of overt discrimination have received enormous attention from lawyers, judges, academics, and policymakers. While explicit sexism, racism, and other forms of bias persist, they have become less prominent and public over the past century. . . . [L]ikely more pervasive, are questions surrounding im- plicit bias . . . .”); Brian A. Nosek et al., Pervasiveness and Correlates of Implicit Attitudes and Stereotypes, 18 EUR. REV. SOC. PSYCHOL. 36, 58–60 (2007) (analyzing the results of over 2.5 million online implicit bias tests and showing that implicit bias is more prevalent than explicit bias, at least when the latter is measured on a self-reporting basis). 28 See, e.g., Floyd v. City of New York, 959 F. Supp. 2d 540, 561 (2013) (“[T]he NYPD has an unwritten policy of targeting ‘the right people’ for stops. In practice, the policy encourages the targeting of young [B]lack and Hispanic men . . . . This is a form of racial profiling.” (footnote omitted)). 150 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 B. Implicit Racial Bias Implicit racial bias is an umbrella term attaching to a number of unconscious attitudes, stereotypes, and prejudices about people of different races.29 Our understanding of im- plicit racial bias is based on the “science of implicit cognition [that] suggests that actors do not always have conscious, intentional control over the processes of social perception, impression formation, and judgment that motivate their ac- tions.”30 In the policing context, implicit racial bias most commonly describes the process by which police officers unconsciously associate Black civilians with criminality or dangerousness.31 Implicit racial bias operates subconscious- ly despite a police officer’s best intentions. As L. Song Richardson has explained, “implicit stereotypes can cause an officer who harbors no conscious racial animosity and who rejects using race as a proxy for criminality to unintentional- ly treat individuals differently based solely upon their physi- cal appearance.”32 Implicit racial bias can infect police interactions with Black civilians in deeply troubling ways. Two such potential manifestations discussed here have been chosen for their relevance to the genesis of the body-worn camera movement: the increased scrutiny of Black people in daily life and the increased use of force against Black civilians. First, implicit racial bias can contribute to increased police scrutiny of Black civilians—for example, by police- initiated questioning of Black people on the street or by the more invasive practice of Terry stops.33 It is well established 29 Anthony G. Greenwald & Linda Hamilton Krieger, Implicit Bias: Scientific Foundations, 94 CALIF. L. REV. 945, 951 (2006). 30 Id. at 946. 31 See Carbado & Rock, supra note 27, at 167–73 (“[In the policing context,] African-American men are associated not only with criminality but also with violence and dangerousness.”); Jennifer L. Eberhardt et al., Seeing Black: Race, Crime, and Visual Processing, 87 J. PERSONALITY & SOC. PSYCHOL. 876, 876 (2004) (“The stereotype of Black Americans as violent and criminal has been documented by social psychologists for almost 60 years.”). 32 L. Song Richardson, Arrest Efficiency and the Fourth Amendment, 95 MINN. L. REV. 2035, 2039 (2011). 33 “Terry stops” describe the police practice of stopping civilians for questioning and/or frisking, despite the absence of probable cause. The practice is named after the Supreme Court case in which it was deemed constitutional: Terry v. Ohio, 392 U.S. 1 (1968). No. 4:141] IS IT RECORDING? 151 that police subject Black people to these practices at a higher rate than other civilians.34 This racial disproportionality can plausibly be attributed, at least in part, to police officers’ implicit racial biases, whereby officers are more likely to pay attention to, and thus approach, a Black civilian than a White civilian.35 This vein of implicit bias has been called “attentional bias”36 and is borne out in studies showing that Black individuals both attract attention more often37 and retain attention for longer38 than White individuals. As Richardson has written: Implicit biases affect whether behavior catches attention in the first place and whether the observer will interpret that behavior as suffi- 34 See, e.g., Floyd v. City of New York, 959 F. Supp. 2d 540, 560 (2013) (“Blacks and Hispanics are more likely than whites to be stopped [by the NYPD] within precincts and census tracts, even after controlling for other relevant variables. . . . Together, these results show that [B]lacks are likely targeted for stops based on a lesser degree of objectively founded suspicion than [W]hites.”); Jeffrey Fagan & Amanda Geller, Following the Script: Narratives of Suspicion in Terry Stops in Street Policing, 82 U. CHI. L. REV. 51, 55 (2015) (observing race-based suspicion leading to stops and searches of Black and Latinx people in New York); David A. Harris, Factors for Reasonable Suspicion: When Black and Poor Means Stopped and Frisked, 69 IND. L.J. 659, 679–80 (1994) (“Police are much more likely to stop African-American men than [W]hite men.”); Anthony C. Thompson, Stopping the Usual Suspects: Race and the Fourth Amendment, 74 N.Y.U. L. REV. 956, 983–90 (1999) (reviewing scientific literature on stereotyping and analyzing how racial stereotypes likely played a role in the stop under review in Terry v. Ohio, 392 U.S. 1 (1968)). 35 See Jeffrey Fagan & Garth Davies, Street Stops and Broken Windows: Terry, Race, and Disorder in New York City, 28 FORDHAM URB. L.J. 457, 481 (2000) (“The prominence of race in the decision to stop citizens may not rise to the threshold of racial profiling, but it does seem to create a racial classification of ‘suspicion.’”); Richardson, supra note 32, at 2052 (“Based on the science, it is reasonable to conclude that the police target, stop, and search [B]lacks more often than [W]hites based on the operation of implicit biases.”). 36 Eberhardt et al., supra note 31, at 881. 37 See Sophie Trawalter et al., Attending to Threat: Race-Based Patterns of Selective Attention, 44 J. EXPERIMENTAL SOC. PSYCHOL. 1322, 1322 (2008) (“[T]he stereotype that young Black men are threatening and dangerous has become so robust and ingrained in the collective American unconscious that Black men now capture attention . . . .”). 38 See Eberhardt et al., supra note 31, at 887 (describing a study which bore out the researchers’ hypothesis that racial biases led partici- pants to look at Black faces quicker than White faces and to look at Black faces for longer than White faces). 152 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 ciently suspicious to warrant further investi- gation. . . . Hence, police attention may be drawn to [B]lack individuals . . . regardless of whether these individuals are engaged in suspicious behavior.39 Secondly, implicit racial bias can result in higher rates of police use of force against Black civilians. Studies explain this manifestation of implicit racial bias by demon- strating that people, including police officers, are more likely to view a Black person’s behavior as threatening.40 One such study suggests that, as a result of social learning, people are more likely to respond fearfully to Black than to White indi- viduals.41 Another study suggests that people will more readily detect hostility in Black faces than in White faces with identical expressions.42 A slew of “shooter bias” simulation studies have shown that people are generally quicker to shoot armed and unarmed Black people than they are to shoot similarly situated White people.43 This review of 39 Richardson, supra note 32, at 2052–53 (footnote omitted). 40 See Nosek et al., supra note 27, at 56 (“[M]ost participants [of over 2.5 million online implicit bias tests] showed stronger associations of Blacks with weapons and Whites with harmless objects compared to the reverse pairing . . . .”). 41 See Matthew D. Lieberman et al., An fMRI Investigation of Race-Related Amygdala Activity in African-American and Caucasian- American Individuals, 8 NATURE NEUROSCIENCE 720 (2005) (reporting re- sults of a study using functional magnetic resonance imaging to show that both Black and White participants recorded higher levels of amygdala ac- tivity in response to Black targets than to White targets; the amygdala is, broadly speaking, the section of the brain associated with fear and survival instincts). 42 Kurt Hugenberg & Galen V. Bodenhausen, Facing Prejudice: Implicit Prejudice and the Perception of Facial Threat, 14 PSYCHOL. SCI. 640, 643 (2003). 43 See, e.g., Joshua Correll et al., The Police Officer’s Dilemma: Using Ethnicity to Disambiguate Potentially Threatening Individuals, 83 J. PERSONALITY & SOC. PSYCHOL. 1314, 1327 (2002) (“In four studies, participants showed a bias to shoot African American targets more rapidly and/or more frequently than White targets.”); Anthony G. Greenwald et al., Targets of Discrimination: Effects of Race on Responses to Weapons Holders, 39 J. EXPERIMENTAL SOC. PSYCHOL. 399, 405 (2003) (“Race of target can affect both (a) perceptual ability to discriminate a weapon from a harmless object and (b) bias to respond as if a weapon is present.”); Melody S. Sadler et al., The World Is Not Black and White: Racial Bias in the Decision to Shoot in a Multiethnic Context, 68 J. SOC. ISSUES 286, 297– 98 (2012) (replicating early “shooter bias” results in samples of police No. 4:141] IS IT RECORDING? 153 the psychological literature reveals that implicit racial bias operates in a very different way than deliberate racial profil- ing. Whereas deliberate racial profiling involves conscious decisions and behaviors intended to negatively discriminate against Black people, implicit racial biases are instinctive attitudes and responses operating at a subconscious level. To address the root causes of racialized policing, we must adopt a multifaceted understanding of racial bias that integrates both subconscious and conscious discrimination against Black civilians. We may not always know whether a racially skewed policing phenomenon—such as stop-and- frisk—is the product of deliberate or implicit racial bias; so, we need to be prepared to respond to both possibilities. Thus, any body-worn camera policy that hopes to reduce the incidence of racially motivated policing must be calibrated to change both deliberate and implicit racial bias. Proponents of body-worn cameras generally advance three theories as to how this change might be achieved: a deterrence theory, an accountability theory, and a training theory. IV. A SOLUTION? BODY-WORN CAMERAS AND THREE THEORIES OF CHANGE Claims that body-worn cameras can mitigate “the implicit and explicit bias against people of color and particu- larly African American youth”44 are usually made by refer- ence to three theories of change. This Note refers to these as the deterrence theory, the accountability theory, and the training theory. officers). But see Lois James et al., Racial and Ethnic Bias in Decisions to Shoot Seen Through a Stronger Lens: Experimental Results from High- Fidelity Laboratory Simulations, 10 J. EXPERIMENTAL CRIMINOLOGY 323, 336 (2014) (finding that, although subjects held subconscious racial biases associating Black people with threat, these biases did not express themselves in decisions to shoot.). 44 Press Release, Lawyers’ Comm. for Civil Rights Under Law et al., A Unified Statement of Action to Promote Reform and Stop Police Abuse (Aug. 18, 2014), https://www.aclu.org/sites/default/files/assets/black_leaders_joint _statement_-_final_-_8-18.pdf [https://perma.cc/W78J-JJLX]. 154 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 A. Body-Worn Cameras and the Deterrence Theory of Change The deterrence theory postulates that when police officers understand their actions to be reviewable, they are deterred from violating commonly held norms of conduct.45 The first randomized controlled trial of body-worn cameras appears to have borne out this hypothesis. In that trial, researchers found that officers wearing cameras were in- volved in fewer use-of-force incidents than an analogous con- trol group without cameras.46 The researchers concluded plausibly that the officers wearing cameras behaved better, and resorted to force less often, because of their awareness that they were being “watched.”47 There is survey evidence to suggest that the public strongly believes that body-worn cameras are capable of having this effect.48 Though it re- mains true that not everyone accepts the plausibility of the 45 See, e.g., THE CONSTITUTION PROJECT, GUIDELINES FOR THE USE OF BODY-WORN CAMERAS BY LAW ENFORCEMENT: A GUIDE TO PROTECTING COMMUNITIES AND PRESERVING CIVIL LIBERTIES 12 (2016) (“With respect to officer behavior, body cameras may deter officer misconduct, including improper use of force, unjustified stops or arrests, and other violations of the law or department policy.”); Barak Ariel et al., Report: Increases in Police Use of Force in the Presence of Body-Worn Cameras Are Driven by Officer Discretion: A Protocol-Based Subgroup Analysis of Ten Randomized Experiments, 12 J. EXPERIMENTAL CRIMINOLOGY 453, 455 (2016) (“The theoretical basis for the use of cameras—that being monitored changes behavior—is deterrence theory. . . . BWCs increase the perceived certainty of apprehension for rule violations. . . . When BWCs are actually turned on and appropriately activated, they can efficiently detect rule violations and law breaking by officers . . . and this process can send a credible deterrence threat.” (citations omitted)); Ariel et al., supra note 8, at 747–50 (reviewing literature on deterrence theory and applying it to the use of body-worn cameras); Coudert et al., supra note 17, at 750, 756–57 (“[B]y exposing bad . . . behaviour, it is hoped that [body-worn cameras] will act as deterrent against the [mis]use of force and discrimination by police officers . . . because of this deterrent effect, body-worn cameras are expected to improve policing . . . .”). 46 Ariel et al., supra note 12, at 523. 47 Id. at 526 (“We interpret this to reflect a fundamental tendency of humans to exhibit more desirable behaviors when they know they are under surveillance . . . .”). 48 THE POLICING PROJECT, supra note 17, at 10 (reporting that eighty- nine percent of respondents to a survey of 2500 people believed that body- worn cameras would improve the conduct of police officers interacting with members of the public). No. 4:141] IS IT RECORDING? 155 deterrence function of body-worn cameras,49 there is such broad support for the theory among the public, policymakers, and scholars that it is likely to inform any public discussion about body-worn camera policies. B. Body-Worn Cameras and the Accountability Theory of Change Meanwhile, the accountability theory is based on the simple idea that when a person is held accountable for unsat- isfactory behavior, they are more likely to refrain from that behavior in the future. Body-worn cameras increase the like- lihood that an officer acting on racial biases will be discov- ered, investigated, and disciplined.50 Importantly, under the accountability theory, body-worn cameras not only improve the behavior of the individual officer disciplined, but also im- prove that of other officers who, having seen the disciplinary consequence, will modify their behavior to avoid disciplinary action. 49 See, e.g., Ben Bradford & Jonathan Jackson, Enabling and Constraining Police Power: On the Moral Regulation of Policing, in ROUTLEDGE HANDBOOK OF CRIMINAL JUSTICE ETHICS 219, 226 (Jonathan Jacobs & Jonathan Jackson eds., 2017) (“Almost all discussion of the influ- ence of BWV and other recording technology on police officers has revolved around rational choice and deterrence theory—the presence of cameras deters them from behaviours they might otherwise have engaged in simply because the cameras increase the risk of censure and sanction. The empha- sis is, then, on extrinsic motivations for behaviour that, we argue below, are not necessarily particularly strong or efficacious. On this basis alone increased surveillance of police is unlikely to solve on its own the problems thrown up by stop and search (and many other practices besides).”). 50 THE CONSTITUTION PROJECT, supra note 45, at 13 (“[Video evi- dence] allows departments to more rapidly take appropriate disciplinary action in response to misconduct.”); Coudert et al., supra note 17, at 755 (“As accountability mechanism, [body-worn cameras] could act in two ways: to apportion responsibility and to encourage learning.”); see also David A. Harris, Picture This: Body-Worn Video Devices (“Head Cams”) as Tools for Ensuring Fourth Amendment Compliance by Police, 43 TEX. TECH L. REV. 357, 363–64 (2010) (discussing the use of body-worn cameras to substantiate civilian complaints against police officers). 156 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 1. Wide Acceptance of the Accountability Theory Police departments, the public, and the judiciary all appear to subscribe to the accountability theory. As shown in Appendix A, eight of the ten police departments included in the comparative analysis included “accountability” or “transparency” as an objective of their body-worn camera pol- icies. Furthermore, the Police Executive Research Forum (PERF) interviewed forty police executives whose depart- ments had implemented, or considered implementing, body- worn cameras. PERF registered a widespread belief among the executives that body-worn cameras could enhance ac- countability and transparency.51 Public support for the ac- countability theory is evident in two separate studies. One survey of public responses to a draft NYPD policy recorded respondents’ hopes that body-worn cameras would enable the department to better identify police officer misconduct.52 Another survey, this time of Maryland community members, recorded similar sentiments that body-worn cameras would improve police accountability.53 Finally, at least some mem- bers of the judiciary believe in the accountability theory, as indicated by the inclusion of body-worn cameras in judicial orders following complaints of racially biased policing in Ferguson, Baltimore, Newark, and New York.54 In the litiga- tion surrounding the NYPD’s racialized stop-and-frisk policy, the trial court justified its inclusion of body-worn cameras in the final order by reasoning: [Body-worn cameras] will provide a contempo- raneous, objective record of stops and frisks, allowing for the review of officer conduct . . . . 51 See MILLER ET AL., supra note 17, at 5 (“The police executives whom PERF consulted cited many ways in which body-worn cameras have helped their agencies strengthen accountability and transparency.”). 52 THE POLICING PROJECT, supra note 17, at 10. 53 See Ray et al., supra note 9, at 10–11 (quoting respondents to a survey about body-worn cameras who believe that they will improve police accountability). 54 Consent Decree at 52–57, United States v. City of Ferguson, No. 4:16-cv-00180-CDP (E.D. Mo. Apr. 19, 2016), ECF No. 41; Consent Decree at 95–96, United States v. Police Dep’t of Balt., No. 1:17-cv-00099-JKB (D. Md. Jan. 12, 2017), ECF No. 2-2; Consent Decree at 38–39, United States v. City of Newark, No. 2:16-cv-01731-MCA-MAH (D.N.J. 2016), ECF No. 4-1. No. 4:141] IS IT RECORDING? 157 [that] may either confirm or refute the belief of some minorities that they have been stopped simply as a result of their race . . . .55 C. Body-Worn Cameras and the Training Theory of Change Some proponents of body-worn cameras also advance a final theory of change, the training theory. The training theory is rooted in a belief that body-worn cameras allow po- lice to develop training programs specifically targeted at par- ticular practices of racially biased policing. At the level of an individual officer, body-worn camera footage may allow a supervisor to identify racially biased behavior—such as stop- ping Black civilians more regularly than White civilians— and require that officer to engage in training designed to change this behavior. At a department-wide level, aggregate data derived from body-worn camera footage may reveal that there is a widespread or systemic problem of racial bias that would require the entire force to engage in remedial training.56 There is evidence that police believe in the training theory of change: Appendix A shows that nine of the ten po- lice departments in my comparative analysis included “training” as an objective of their body-worn camera policies. The theory appears to have the support not just of the policy drafters, but also of frontline police officers. One survey con- ducted after a body-worn camera study recorded that almost eighty percent of police officers agreed that reviewing their own conduct on body-worn camera footage would help them improve their behavior.57 55 Floyd v. City of New York, 959 F. Supp. 2d 668, 685 (S.D.N.Y. 2013). 56 See MILLER ET AL., supra note 17, at 8–9 (describing body-worn cameras as a useful tool to train individual officers and “to address wide- reaching structural problems within the department”); Harris, supra note 50, at 364–65 (“[S]upervisors could use the recordings for more general (i.e., not complaint responsive) assessment, training, and disciplinary deci- sions.”). 57 Jennings et al., supra note 12, at 484 (recording that 79.5% of officers involved in a body-worn camera study agreed, in a post-study survey, that “reviewing BWC video after an incident would help them . . . identify issues that they may need improvement on”). 158 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 D. A Prerequisite for Change: Footage Is Reviewed Before departing the topic of theories of change, some concluding comments are in order. Each of the above theo- ries of change can only operate effectively if internal supervi- sors or external auditors not only record, but actually review body-worn camera footage.58 If someone is not reviewing the footage, then police officers will not be deterred or disci- plined, and they will not receive targeted training. Though some critics have doubted that police departments have the funds and resources to regularly review countless hours of body-worn camera footage,59 there is reason to believe that police departments are systematically reviewing such foot- age. A nationwide survey of 254 police departments found that ninety-four percent used body-worn camera footage to train officers and assist in administrative reviews.60 This suggests that there are already procedures in place laying a foundation for each of the three theories of change just postu- lated. However, before further examining how these theories of change might inform the ideal body-worn camera activa- tion policy, it is helpful to survey the policies that are currently in place. 58 See Coudert et al., supra note 17, at 757 (“[The deployment of body-worn cameras] should be accompanied by organizational [sic] measures ensuring that police officers receive sufficient and meaningful feedback from their supervisors after incidents . . . . In that sense, body-worn cameras can also act as [a] forward-looking mechanism, to encourage learning.”); Developments in the Law—Policing, 128 HARV. L. REV. 1706, 1802, 1802 n.55 (2015) (“But this benefit is contingent on departmental review policies, as some police departments only allow supervisors to ac- cess footage if a civilian complaint has been made—prohibiting random screenings of officer conduct in the field.”). 59 See Spencer et al., supra note 23, at 58 (“[A]nalyzing thousands of hours of audio and video recordings can be logistically complicated and prohibitively costly.”). 60 MILLER ET AL., supra note 17, at 7 (“Many police agencies are discovering that body-worn cameras can serve as a useful training tool to help improve officer performance. For example, agencies are using footage from body-worn cameras to provide scenario-based training, to evaluate the performance of new officers in the field, and to identify new areas in which training is needed.”). No. 4:141] IS IT RECORDING? 159 V. ACTIVATION POLICIES IN THE TEN LARGEST METROPOLITAN POLICE DEPARTMENTS Body-worn cameras do not have a long history in the United States. First trialed in continental Europe and the United Kingdom,61 body-worn cameras began to enter United States police departments in the late 2000’s.62 The pilot programs and initial trials of these devices proceeded on a city-by-city basis rather than as a uniform national rollout. The result has been a lack of uniformity in policies and pro- cedures.63 Indeed, some departments appear to lack any pol- icies at all.64 Perhaps because of the piecemeal and localized nature of police body-worn camera programs, there has been little scholarly effort directed at comparing and critiquing the policies governing the use of these devices. Mary Fan’s pioneering work has gone some way toward improving this area of academic study. Fan looked at the municipal police departments serving the largest 100 cities in the United States, of which eighty-eight had used or planned to use body-worn camer- as.65 As of December 2015, the date of Fan’s study, only thir- ty-nine departments had publicly available guidelines.66 Three other police departments were located in states that had legislative guidelines regarding body-worn cameras.67 Accordingly, Fan coded and compared the guidelines for a total of forty-two police departments. My December 2017 survey of the body-worn camera policies of the ten largest 61 See Coudert et al., supra note 17, at 751–53 (describing use of body-worn cameras by police departments in the United Kingdom); Kampfe, supra note 17, at 1156–57 (describing the history of body-worn cameras in Denmark and the United Kingdom). 62 Harris, supra note 50, at 361–62 (summarizing the emergence of body-worn cameras in the United States). 63 See Kampfe, supra note 17, at 1155 (tracking the rapid imple- mentation of body-worn cameras and suggesting that there is little, if any, policy uniformity). 64 See MILLER ET AL., supra note 17, at 2 (“PERF received responses from 254 departments . . . Of the 63 agencies that reported using body-worn cameras, nearly one-third did not have a written policy governing body-worn camera usage. Many police executives reported that their hesitance to imple- ment a written policy was due to a lack of guidance on what the policies should include . . . .”). 65 Fan, supra note 16, at 426; Fan, supra note 6, at 931. 66 Fan, supra note 16, at 426; Fan, supra note 6, at 931. 67 Fan, supra note 16, at 426; Fan, supra note 6, at 931. 160 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 metropolitan police departments in the United States builds on Fan’s research. My findings, which are discussed further below, align with Fan’s insofar as they identify two primary models for body-worn camera activation policies: “limited discretion” policies and “minimal discretion” policies. As the labeling suggests, the point of differentiation between these two mod- els is the degree of discretion they afford police officers. Limited discretion policies allow officers more autonomy as to what they need to record on their cameras; minimal dis- cretion policies allow officers less autonomy. The fact that discretion is the point of difference between the two models should not be surprising. Discretion has long been a signifi- cant topic of debate in modern theories of policing. On the one hand, it is argued that individual police officer discretion should be reduced in order to ensure that the law is applied equally to everyone rather than applied variably according to the whim of a particular officer.68 On the other hand, advo- cates of discretion suggest that it is necessary to ensure that police officers are able to respond flexibly to different circum- stances and thus to avoid the harshness of blindly applied universal rules.69 As this Note will discuss, these competing views of discretion underlie the two common models of body- worn camera activation policies. A. Limited Discretion Activation Policies Limited discretion activation policies operate by pre- sumptively providing police officers discretion as to whether or not to record their activities while on duty.70 This pre- sumption of officer discretion is only displaced in specific circumstances (in which recording is mandatory). An illus- trative limited discretion policy is that of the NYPD, which states: 68 K. C. DAVIS, POLICE DISCRETION 145 (1975). 69 Wayne R. La Fave, The Need for Discretion, in LAW AND THE LAWLESS: A READER IN CRIMINOLOGY 299 (1969). 70 Cf. Fan, supra note 16, at 427 (“A limited-discretion model cur- tails officer discretion by requiring recording of several specified law en- forcement activities, while leaving some situations up to officer discretion.”); Fan, supra note 6, at 931 (“Under a limited discretion model, police are directed to record specified enforcement activities and given discretion over whether to record at other times.”). No. 4:141] IS IT RECORDING? 161 Activate BWC prior to engaging in, or assisting another uniformed member of the service with, the following police actions: a. Arrests b. Summonses . . . c. Vehicle stops d. Interactions with persons suspected of criminal activity e. A search of an individual and/or his/her belongings . . . f. Interactions with an emotionally dis- turbed person g. Use of force . . . h. Public interactions that escalate and become adversarial . . . . Uniformed members of the service may record other official activities when, in the uniformed member’s judgment, it would be beneficial to record . . . .71 This NYPD policy is characteristic of the limited discretion model because there is no generalized prescription requiring mandatory recording as there is, for instance, in the Chicago Police Department (“the Chicago PD”) policy reproduced be- low.72 Instead, the NYPD policy grants a general discretion to officers as to whether to record their activities and only displaces this discretion in the specific circumstances listed in subsections a through h. The limited discretion model places significant trust in the individual police officer to make the right “judgment” about whether it is “beneficial” to record. In this sense, the NYPD policy can be understood to come down on the pro-discretion side of the long-raging de- bate about the desirability of discretion in modern policing. As the following section shows, the Chicago PD can be seen to fall into the opposing anti-discretion camp. 71 NYPD, DRAFT 16, PILOT PROGRAM—USE OF BODY-WORN CAMERAS, §§ 5, 8 (2017) (emphasis added). 72 CHI. POLICE DEP’T, SPECIAL ORDER S03-14, BODY WORN CAMERAS (2018). The Chicago policy states: “The decision to electronically record a law-enforcement-related encounter is mandatory, not discretionary, except where specifically indicated.” Id. § III(A)(1) (emphasis added). 162 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 B. Minimal Discretion Activation Policies The other dominant type of body-worn camera activa- tion policy among United States police departments is a minimal discretion policy.73 Such policies start from a pre- sumption that all enforcement activities will be recorded, subject only to narrow exceptions.74 An example of a minimal discretion model can be seen in the Chicago PD policy, which states: 1. The decision to electronically record a law- enforcement-related encounter is mandato- ry, not discretionary, except where specifi- cally indicated. 2. The Department member will activate the system . . . at the beginning of an incident and will record the entire incident for all law-enforcement-related activities. . . . . 3. A Department member may utilize discre- tion to activate the BWC for non-law- enforcement-related activities [including] . . . community caretaking functions . . . .75 The difference between the limited and minimal discretion policies is illustrated by the fundamentally different starting point of the Chicago PD policy as compared to the earlier- quoted NYPD policy. In a limited discretion model like the NYPD’s, the default position is voluntary recording; whereas in a minimal discretion model like the Chicago PD’s, the de- fault position is mandatory recording. Thus, in the terms of the discretion debate adverted to above, the Chicago PD’s 73 Fan prefers to label this sort of policy as a “highly-limited- discretion model.” See Fan, supra note 16, at 427; Fan, supra note 6, at 932. Allyson Roy prefers the language of “a strict-control policy.” See Allyson Roy, On-Officer Video Cameras: Examining the Effects of Police Department Policy and Assignment on Camera Use and Activation 12 (May 2014) (unpublished M.S. thesis, Arizona State University), https://repository.asu.edu/attachments/ 134979/content/Roy_asu_0010N_13803.pdf [https://perma.cc/BEA5-QFNG]. 74 Cf. Fan, supra note 16, at 427 (“This highly-limited-discretion approach requires that body cameras record during all enforcement en- counters with the public, with only limited exceptions.”); Fan, supra note 6, at 932. 75 CHI. POLICE DEP’T, SPECIAL ORDER S03-14, BODY WORN CAMERAS § III(A)(1)–(3) (2018) (emphasis added). No. 4:141] IS IT RECORDING? 163 minimal discretion policy can be seen to be more concerned with reducing police officer discretion and ensuring uniform policy application than with allowing officer flexibility in re- sponding to different situations. Having described the two common models of body-worn camera activation policy, it is now possible to assess their relative popularity. C. A Trend Toward Reducing Police Officer Discretion In Fan’s 2015 study, the vast majority (eighty percent) of police departments opted for the policy that afforded officers more discretion about what to record on their cameras—the limited discretion policy.76 These results are at odds with the results of my smaller, more recent study in which only twenty percent of departments afforded their officers such discretion, as shown in Appendix B. This signif- icant difference suggests that, since 2015, large metropolitan police departments have been moving increasingly to reduce police officer discretion as to what must be recorded on their body-worn cameras. The implications of this development for racial bias are discussed in Part VI, where this Note argues that reducing police officer discretion even further will result in activation policies best suited to combating racially biased policing. Before commencing that discussion, however, it is necessary to qualify the endorsement of the limited/minimal discretion binary. D. Gradations of Discretion The apparent binary between limited and minimal discretion policies outlined above, and in Fan’s work, is somewhat reductive. A closer analysis of the specific phrase- ology of the policies reveals significant gradations of record- ing requirements, even within policies on the same side of the limited/minimal equation. For example, compare the above-quoted Chicago PD policy, which requires mandatory recording of all enforcement activities, with the minimal dis- cretion policy of the Baltimore Police Department (“the Baltimore PD”), which requires mandatory recording of all 76 Fan, supra note 6, at 932. 164 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 investigative and enforcement activities.77 On its face, the Baltimore PD’s inclusion of “investigative activities” signifi- cantly extends its mandatory recording requirements beyond those of the Chicago PD. The Baltimore PD policy would appear to require recording of, for example, all preliminary civilian inquiries, door knocks, witness interviews, crime- scene investigations, and evidence collection activities (unless covered by an enumerated exclusion). The purpose of adverting to the shades of discretion even within the same policy model is to emphasize that the optimal body-worn camera policy will be drafted with attention to whether each word, clause, and concept enlarges or reduces police officer discretion. It is not enough to simply endorse one model or the other in the abstract. With this in mind, the next section turns to consider how activation policies might best be draft- ed to address deliberate and implicit racial bias. VI. ASSESSING ACTIVATION POLICIES AGAINST THE STATED GOAL OF REDUCING RACIALLY BIASED POLICING The preceding discussion introduced the dual problem of deliberate racial profiling and implicit racial bias; ex- plained the potential solution of body-worn cameras by refer- ence to three theories of change; and surveyed the current array of body-worn camera activation policies. In Sections VI.A and VI.B, this Part will now analyze the most effective way in which a policy might leverage the theories of change to reduce deliberate and implicit racial bias. Section VI.C will give special consideration to activation requirements for “consensual”78 interactions between police and civilians be- cause of the heightened risk of racial bias manifesting in 77 Compare CHI. POLICE DEP’T, supra note 75, with BALT. POLICE DEP’T, POLICY 824, BODY WORN CAMERAS PILOT PROGRAM 2 (2015). 78 This Note considers “consent” in the stop-and-search context to be a legal fiction with little connection to everyday understandings of con- sent. Accordingly, this Note uses quotation marks around the word “con- sent” and its derivatives when the legal concept is being described. Cf. Oren Bar-Gill & Barry Friedman, Taking Warrants Seriously, 106 NW. U. L. REV. 1609, 1662 (2012) (noting the authors’ deliberate use of quotation marks around the word “consent” in their article); Alafair S. Burke, Consent Searches and Fourth Amendment Reasonableness, 67 FLA. L. REV. 509, 513 n.12 (2015) (collecting some of the scholarly literature that insists on quotation marks around the word “consent” in the policing context). No. 4:141] IS IT RECORDING? 165 these interactions. Finally, Section VI.D introduces a draft activation policy for consideration by interested police de- partments, policymakers, and affected communities. A. Activation Requirements and Deliberate Racial Profiling Each of the three theories of change—deterrence, accountability, and training—relies for its effectiveness on police officers activating their body-worn cameras. If a cam- era is not activated during an act of deliberate racial profil- ing, then the camera is serving little deterrence function; it is not going to catalyze a disciplinary response and it will not result in any responsive training.79 The necessary implica- tion of this observation is that the more often cameras are recording, the more effective they will be in reducing deliber- ate racial profiling. Taken to its fullest extent, this logic suggests that the activation policy that would most likely reduce deliberate racial profiling would be one that requires police officers to activate their cameras for the entirety of their shift.80 This can be labeled a “continuous recording” policy. 1. The Advantages of Continuous Recording Barak Ariel, the leading empirical researcher in the field, and his coauthors have advocated for exactly this ap- proach, arguing that “cameras should remain on throughout 79 But see Barak Ariel et al., “Contagious Accountability”: A Global Multisite Randomized Controlled Trial on the Effect of Police Body-Worn Cameras on Citizens’ Complaints Against the Police, 44 CRIM. JUST. & BEHAV. 293 (2017) (hypothesizing that body-worn cameras may in fact influence officer behavior even when they are not recording, because offic- ers will become so accustomed to the oversight of body-worn cameras that they will assume constant surveillance, either from their own device or from that of another officer; this possibility does not undermine my focus on activation policies—if we want to increase “contagious accountability,” then we need expansive activation policies that encourage officers to as- sume that there will always be a camera recording). 80 Cf. Bryce Clayton Newell, Crossing Lenses: Policing’s New Visibility and the Role of “Smartphone Journalism” as a Form of Freedom- Preserving Reciprocal Surveillance, 2014 J.L. TECH. & POL’Y 59, 85 (2014) (suggesting that if body-worn cameras are “always on” they will better “serve to support citizen oversight and law enforcement accountability”). 166 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 the entire shift—that is, during each and every interaction with citizens.”81 The American Civil Liberties Union (ACLU) has also recognized that, “from an accountability perspective, the ideal policy for body-worn cameras would be for continu- ous recording throughout a police officer’s shift.”82 To require anything other than continuous recording reduces the capac- ity of cameras to protect against deliberate racial profiling, because an officer can avoid detection by simply not activat- ing their camera. A number of scholars and civil rights or- ganizations have expressed concern about this possibility.83 The ACLU has noted: [P]olicies and technology must be designed to ensure that police cannot ‘edit on the fly’—i.e., choose which encounters to record with limit- less discretion. If police are free to turn the cameras on and off as they please, the camer- as’ role in providing a check and balance against police power will shrink and they will no longer become a net benefit.84 The public shares this concern. In the Policing Project’s sur- vey of more than 25,000 people in New York City, “a 81 Ariel et al., supra note 45, at 461. 82 JAY STANLEY, ACLU, POLICE BODY-MOUNTED CAMERAS: WITH RIGHT POLICIES IN PLACE, A WIN FOR ALL 3 (2015), https://www.aclu.org/ sites/default/files/assets/police_body-mounted_cameras-v2.pdf [https://perma .cc/ML4F-RXRB]. 83 See, e.g., Emmeline Taylor, Lights, Camera, Redaction . . . Police Body-Worn Cameras: Autonomy, Discretion and Accountability, 14 SURVEILLANCE & SOC’Y 128, 129 (2016) (“The ability of officers to ‘edit on the fly’ fundamentally undermines any potential benefits the cameras in- troduce.”); Jay Stanley, Police Body Cameras: The Lessons of Albuquerque, ACLU: FREE FUTURE (Mar. 24, 2015, 11:30 AM), https://www.aclu.org/blog/ police-body-cameras-lessons-albuquerque [https://perma.cc/3CQQ-424Y] (“In Albuquerque the police have body worn cameras . . . . But they’ve been used sporadically—police use cameras when it suits them, and they don’t when it doesn’t. The selective recording that goes on in these police misconduct cases has been a real problem for us. There have been police shootings where the cameras simply weren’t turned on.” (quoting Alexandra Smith, legal director of the ACLU of New Mexico)). 84 STANLEY, supra note 82, at 2–3; see also Coudert et al., supra note 17, at 756 (“Body-worn cameras have the potential to act as [a] power- ful deterrent mechanism by exposing behaviour that was previously not scrutinised. Still, the latitude let to police officers to turn the camera on and off will be a decisive factor in that regard.”). No. 4:141] IS IT RECORDING? 167 substantial number of [survey respondents] . . . expressed concern that officers would fail to record contentious encounters.”85 The worry that contentious encounters will not be recorded unless the cameras are continuously recording is supported by empirical evidence. One study of a body-worn camera program in Phoenix required officers to activate their cameras in all enforcement encounters. In reality, however, they recorded “only 13.2 to 42.2 percent of incidents.”86 The results were almost identical in a Denver pilot project. In that study, the policy required officers to record all uses of force but, in fact, less than half of the uses of force were caught on camera.87 In a British study where the body-worn camera policy required recording of all police attendances for domestic violence, only one in six officers recorded every incident.88 Finally, in New Orleans, a court-appointed mon- itor reported that only about one third of uses of force were recorded, despite a consent decree requiring the recording of all uses of force.89 Of course, not all failures to record suggest that there has been deliberate wrongdoing, but the failure to record does have at least the potential to mask such wrongdoing. What the aforementioned statistics suggest is that the only way to ensure that all potential incidents of deliberate wrongdoing—such as deliberate racial profiling—are record- ed is to require continuous recording. There are, however, some significant obstacles to implementing a continuous recording policy, most obviously privacy-oriented objections from both police and civilians. 85 THE POLICING PROJECT, supra note 17, at 11. 86 KATZ ET AL., supra note 12, at 3. 87 NICHOLAS E. MITCHELL, DENVER OFFICE OF THE INDEP. MONITOR, THE DENVER POLICE DEPARTMENT’S BODY WORN CAMERA PILOT PROJECT: A FOCUS ON POLICY AND LESSONS LEARNED 1 (2015). 88 CATHERINE OWENS ET AL., THE ESSEX BODY WORN VIDEO TRIAL: THE IMPACT OF BODY WORN VIDEO ON CRIMINAL JUSTICE OUTCOMES OF DOMESTIC ABUSE INCIDENTS 2 (2014). 89 2014 NEW ORLEANS CONSENT DECREE MONITOR THIRD Q. REP. 23–25; see also Ken Daley, Cameras Not on Most of the Time When NOPD Uses Force, Monitor Finds, NOLA.COM (Sept. 4, 2014), http://www.nola.com/ crime/index.ssf/2014/09/cameras_not_on_most_of_the_tim.html [https://perma .cc/T7HG-S96G]. 168 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 2. Obstacles to Continuous Recording Despite the fact that continuous recording would best protect against deliberate racial profiling, there is little pro- spect of such a policy receiving popular support.90 The near consensus is that continuous recording policies unjustifiably infringe on the privacy of police officers and civilians. Di- verse interest groups from across the political spectrum converge on this position, including police unions,91 civil rights groups,92 and scholars,93 all of which have expressed views that the benefits of continuous recording are incapable of outweighing the collateral privacy incursions.94 The ACLU’s position provides a summary of the range of con- cerns: “Continuous recording would . . . mean a lot of mass surveillance of citizens’ ordinary activities. . . . [and] would also impinge on police officers when they are sitting in a station house or patrol car shooting the breeze.”95 3. Recording All Police-Civilian Interactions Given the apparently insurmountable opposition to continuous recording, it is necessary to ask: What is the next 90 Cf. Josh Feit, Seattle State Senator, ACLU Call for Tougher Body Cam Guidelines than in SPD Pilot, SEATTLE MET (Feb. 9, 2015), https://www.seattlemet.com/articles/2015/2/9/aclu-body-cam-bill-calls-for- tougher-oversight-than-spd-version-february-2015 [https://perma.cc/8HVA- KHWK] (describing an unsuccessful bill introduced in the Washington state legislature that sought to impose continuous recording requirements on police officers). 91 See, e.g., NYPD, NYPD RESPONSE TO PUBLIC AND OFFICER INPUT ON THE DEPARTMENT’S PROPOSED BODY-WORN CAMERA POLICY 11 (2017) (“Non-stop recording is impractical both because it would require a vast increase in long-term data storage capacity and because it would represent an invasion of the privacy of many people whom the police encounter.”). 92 See, e.g., STANLEY, supra note 82, at 3 (“Continuous recording would also mean a lot of mass surveillance of citizens’ ordinary activities.”). 93 See, e.g., Coudert et al., supra note 17, at 758 (“Constant recording does not seem an acceptable interference into the right to privacy of policemen and of the persons filmed.”). 94 See Fan, supra note 16, at 426 (“Continuous recording is controversial to privacy proponents and law enforcement officers because of the heavy burden on the privacy of officers and members of the public they encounter.”). 95 STANLEY, supra note 82, at 3. No. 4:141] IS IT RECORDING? 169 best activation policy to reduce deliberate racial profiling? Scholarship has called for body-worn camera policies that require recording of all police-civilian interactions.96 Yet, none has specifically considered how such a policy might help curb deliberate racial profiling. My claim that such an acti- vation policy would reduce deliberate racial profiling relies on three separate sub-claims. First, police officers who ad- here to the policy and record all interactions with civilians but nevertheless engage in deliberate racial profiling will likely be discovered and disciplined. Secondly, a policy requiring the recording of all police- civilian interactions makes it more difficult for officers to remain undetected when they keep their cameras off so as to engage in deliberate racial profiling. Such behavior will raise a “red flag”97 for supervisors reviewing or auditing the officer’s behavior.98 The more a policy requires cameras to be turned on, the more suspicious periods of non-recording will be to a supervisor reviewing an officer’s body-worn camera log. As has been explained by others adhering to this theory, “[r]isks resulting from on-the-fly editing [to avoid recording deliberate misconduct] . . . could be mitigated by tracking how often officers switch recording on and off and flagging those who do it exceptionally often.”99 Finally, where activation is required for all police- civilian interactions, a police officer considering turning their camera off in order to engage in deliberate racial profiling 96 See, e.g., Harris, supra note 50, at 365 (“In order for recordings [to most effectively improve police behavior] . . . the law, departmental rules, or both would have to require officers to record every interaction with citizens.”); Newell, supra note 80, at 85 (“The deployment of officer- mounted cameras may only serve to support citizen oversight and law enforcement accountability when . . . officers adhere to strict guidelines requiring activation during every citizen encounter (unlikely) . . . .”); Freund, supra note 17, at 128 (“This Note recommends that officers turn on their cameras whenever they exit the vehicle to interact with a member of the public, but may subsequently turn off their camera if requested by a member of the public.”). 97 Roy, supra note 73, at 40 (noting that more expansive activation policies will mean that when officers fail to activate their cameras, this will raise “red flags” for supervisors reviewing the officers). 98 See Coudert et al., supra note 17, at 756 (“[I]n order for body- worn cameras to be an efficient tool leading to socially-desirable behaviour of the officers who wear them . . . [there must be] a given level of certainty of being apprehended in case of misbehaviour . . . .”). 99 Id. at 759. 170 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 may be deterred by the likelihood that another police of- ficer—such as a partner or team member—will be recording. Ariel and his colleagues have labeled this secondary deter- rence mechanism “contagious accountability,”100 and two of the ACLU’s body-worn camera experts have endorsed the idea.101 In light of these three arguments, this Note suggests that an activation policy requiring recording of all police- civilian interactions will reduce instances of racial profiling. Remembering that deliberate racial profiling functions on a conscious rather than an unconscious level (contrary to im- plicit bias), the following section turns to best practices for curtailing the unconscious bias problem. B. Activation Requirements and Implicit Racial Bias How exactly do we think body-worn cameras can reduce implicit racial bias? Most researchers seem to accept that the deterrence benefits of body-worn cameras do not ex- tend to non-deliberative behavior, including behavior based on implicit racial bias.102 This rather pessimistic conclusion stands to reason that because implicit racial bias operates at a subconscious level, it is hard to believe that it would be af- fected by deliberative assessments of the possibility of one’s actions being observed.103 One empirical study confirms this intuition. Researchers at Stanford University used body- 100 See Ariel et al., supra note 79 (hypothesizing that body-worn cameras may in fact influence officer behavior even when they are not recording, because officers will become so accustomed to the oversight of body-worn cameras that they will assume constant surveillance, either from their own device or from that of another officer). 101 See Chad Marlow & Jay Stanley, Should We Reassess Police Body Cameras Based on Latest Study?, ACLU: FREE FUTURE (Nov. 20, 2017, 4:15 PM), https://www.aclu.org/blog/privacy-technology/surveillance- technologies/should-we-reassess-police-body-cameras-based [https://perma.cc /8TPU-EXNS]. 102 See, e.g., Ariel et al., supra note 12, at 518 (limiting their causal hypothesis to deliberative or thoughtful behavior); Lawrence Rosenthal, Good and Bad Ways to Address Police Violence, 48 URB. LAW. 675, 684 (2016) (doubting that implicit bias can be “solved” by body-worn cameras). 103 See Spencer et al., supra note 23, at 58 (“[E]ven the specter of one’s behavior being reviewed may not help officers avoid the impact of implicit biases on their behaviors, any more than they can prevent bias from affecting them on the shooter task.”). No. 4:141] IS IT RECORDING? 171 worn camera footage from routine police traffic stops in Oakland to assess the respectfulness of officers interacting with civilians of different races. The study found that “of- ficers speak with consistently less respect toward [B]lack versus [W]hite community members.”104 This was the case notwithstanding the fact that the police officers knew that they were being recorded. Similarly, a number of studies of doctor-patient interactions have shown that doctors consist- ently offer a lower level of care to patients who do not share the doctor’s race or ethnicity, despite the fact that the doctors know that they are being recorded on audio and/or audio- visual equipment.105 Although the authors of these studies do not purport to identify a cause for this observed effect, it is at least plausible that the police officers and doctors treated ci- vilians and patients differently based on implicit racial biases rather than any deliberate racial profiling. The plau- sibility of this conclusion rests on the assumption, supported above,106 that implicit racial bias is more widespread than deliberate racial profiling and thus that the police officers and doctors are more likely to be discriminating subconsciously. If we accept that body-worn cameras are inapt in deterring implicit racial biases, we need to rely on the accountability and training theories for any hope of change. Under these theories, we can hypothesize that body-worn cameras will allow police departments to more readily identi- fy officers acting on implicit racial biases. Such officers could then be disciplined and/or subjected to remedial training.107 For instance, a supervisor reviewing body-worn camera footage of pedestrian stops might observe that a particular 104 Rob Voigt et al., Language from Police Body Camera Footage Shows Racial Disparities in Officer Respect, 114 PROC. NAT’L ACAD. SCI. U.S. 6521, 6521 (2017). 105 See, e.g., Lisa A. Cooper et al., Patient-Centered Communication, Ratings of Care, and Concordance of Patient and Physician Race, 139 ANNALS INTERNAL MED. 907 (2003); Joke C.M. van Wieringen et al., Intercultural Communication in General Practice, 12 EUR. J. PUB. HEALTH 63 (2002). 106 See sources cited supra note 27. 107 On implicit bias training, see Richardson, supra note 32, at 2054–55 (“Even asking people to be nonprejudiced can reduce implicit biases. Consequently, courts and police departments may be able to implement strategies for reducing their effects on behavior.” (footnote omitted)). But see Spencer et al., supra note 23, at 58 (“[T]here is little empirical evidence to support the effectiveness of such trainings, and they are rarely systematically evaluated.”). 172 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 officer appears to be disproportionately stopping Black civil- ians absent any objective grounds. The supervisor could then bring this to the attention of the officer and, absent an explanation,108 discipline the officer and/or require them to participate in de-biasing training. Similarly, if supervisors reviewing body-worn camera footage detect a department- wide trend suggesting widely shared implicit racial biases, they could implement mandatory de-biasing training for the entire department.109 Encouragingly, body-worn camera footage is already being reviewed by police departments for exactly these pur- poses. In San Diego, one of the key motivations for the body- worn camera program was to inform a better understanding of the way racial biases operate within the police force and how it might best be addressed. The San Diego chief of police explained: When it comes to collecting data, the raw numbers don’t always fully capture the true scope of a problem . . . . But by capturing an audio and video account of an encounter, cam- eras provide an objective record of whether ra- cial profiling took place, what patterns of 108 Of course, body-worn camera footage will rarely provide all the necessary information about a given police-civilian interaction. On the limits of body-worn camera footage, see Natalie Todak, De-Escalation in Police- Citizen Encounters: A Mixed Methods Study of a Misunderstood Policing Strategy 172–73 (Aug. 2017) (unpublished Ph.D. dissertation, Arizona State University), https://repository.asu.edu/attachments/189627/content/Todak_ asu_0010E_17217.pdf [https://perma.cc/9QQW-4F32] (“[B]ody worn camera footage comes with its own set of limitations. . . . Numerous details pertaining to the social interaction, background including the officer’s previous knowledge of the citizen or the type of call, the officer’s fatigue levels and previous work shifts and stressors, as well as the aftermath of the interaction are substantially clouded in a body camera video.”); and see also Gráinne Perkins, Commentary, Lights, Camera, Action! Body-Worn Cameras: Challenges and Opportunities in Police Research, 12 POLICING 120, 122 (2018) (“BWCs are only one method of understanding police ac- tions. How the police operate can be better understood in the examination of why they act in a certain manner.”). 109 The Department of Justice recently implemented mandatory implicit bias training for more than 33,000 federal agents and prosecutors. See Julia Edwards, Justice Dept. Mandates “Implicit Bias” Training for Agents, Lawyers, REUTERS (June 27, 2016) https://www.reuters.com/article/ us-usa-justice-bias-exclusive/exclusive-justice-department-mandates-implicit- bias-training-for-agents-lawyers-idUSKCN0ZD251 [https://perma.cc/Z83 Q-F2W7]. No. 4:141] IS IT RECORDING? 173 officer behavior are present, and how often the problem occurs.110 The Oakland study mentioned earlier appears to have been driven by similar ideas. The Stanford researchers who con- ducted the study hoped that the identification of the racial biases displayed in the body-worn camera footage would offer an opportunity for advancing police training.111 In light of the above, the activation policy that rec- ommends itself to best reduce implicit racial bias is again a policy requiring presumptive recording of all police-civilian interactions. Such a policy would capture as much police conduct as possible so as to maximize the opportunity of re- viewing officers to detect, discipline, and retrain officers who appear to be acting on implicit racial biases. A draft proposal for such a policy is provided in Section VI.D, but one important matter—“consensual” interactions—should be dis- cussed so that it may then be assimilated into the draft proposal. C. Activation Requirements and “Consensual” Interactions The final relation between body-worn camera policies and racial bias that warrants discussion is the recording re- quirements for “consensual” interactions between police and civilians.112 Fan describes “consensual” interactions as of- ficer-initiated contact with a civilian “typically in situations where there is either no articulable basis yet for reasonable suspicion or it is unclear if there is a sufficient basis.”113 110 MILLER ET AL., supra note 17, at 8. Although the speaker uses the term “racial profiling,” it is apparent that he is not referring to deliberate racial profiling but instead to “patterns” of implicit racial bias. 111 Voigt et al., supra note 104, at 6521. 112 See Fan, supra note 6, at 933–34. 113 Id. at 933; cf. Devallis Rutledge, Consensual Encounters, POLICE MAG. (Feb. 5, 2016) http://www.policemag.com/channel/patrol/articles/2016/ 02/consensual-encounters.aspx [https://perma.cc/T8EX-6MSX] (“Consensual encounters, in which police use no commands, force, red or blue lights or sirens, but simply approach a person and engage him in conversation and make plain-view observations without any official restraints, and for which no level of justification is needed, because unlike arrests and detentions, the consensual encounter is not a Fourth Amendment ‘seizure’ of the person.”). 174 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 1. Racial Bias in “Consensual” Interactions Because “consensual” interactions do not attract Fourth Amendment scrutiny,114 police officers do not need probable cause or even a reasonable suspicion; they can act on a hunch or no hunch at all.115 It should be unsurprising, then, that “consensual” interactions are common sites for de- liberate racial profiling and implicit racial bias.116 People of color are the subjects of “consensual” police interactions at a disproportionate rate.117 This is especially true with respect 114 See Florida v. Rodriguez, 469 U.S. 1, 5–6 (1984) (“The initial contact between officers and [civilian], where they simply asked if he would step aside and talk with them, was clearly the sort of consensual encounter that implicates no Fourth Amendment interest.”); Florida v. Bostick, 501 U.S. 429, 434–36 (1991) (“Our cases make it clear that a sei- zure does not occur simply because a police officer approaches an individual and asks a few questions. . . . The encounter will not trigger Fourth Amendment scrutiny unless it loses its consensual nature.”). 115 See Margaret Raymond, The Right to Refuse and the Obligation to Comply: Challenging the Gamesmanship Model of Criminal Procedure, 54 BUFF. L. REV. 1483, 1486 (2007) (“Police are free to initiate a consensual encounter with an individual for any reason or no reason, perhaps based on a whim or a ‘hunch’ that cannot be supported by specific and articulable facts.”); Daniel J. Steinbock, The Wrong Line Between Freedom and Restraint: The Unreality, Obscurity, and Incivility of the Fourth Amendment Consensual Encounter Doctrine, 38 SAN DIEGO L. REV. 507, 509 (2001) (“[A] consensual encounter can be initiated for no reason or for any reason at all . . . .”). 116 See I. Bennett Capers, Rethinking the Fourth Amendment: Race, Citizenship, and the Equality Principle, 46 HARV. C.R.-C.L. L. REV. 1, 40 (2011) (“[A]n officer’s decision to single out an individual for a limited detention or consensual encounter is more likely to be based on implicit racial biases unknown to the officer rather than deliberate racism.”); Burke, supra note 78, at 547 (“When there is no other justification for the search, the odds are higher that police are requesting consent, either consciously or unconsciously, based on racial or other stereotypes.”); Wesley MacNeil Oliver, With an Evil Eye and an Unequal Hand: Pretextual Stops and Doctrinal Remedies to Racial Profiling, 74 TUL. L. REV. 1409, 1410–11 (2000) (describing “consensual” vehicular searches as “[t]he archetypal example of racial profiling”). 117 See Racial Disparity in Consent Searches and Dog Sniff Searches, ACLU ILL. (Aug. 13, 2014), https://www.aclu-il.org/en/publications/racial- disparity-consent-searches-and-dog-sniff-searches [https://perma.cc/PA5X -P8DD] (describing racial disparities in people subjected to “consensual” vehicular searches); see also Steinbock, supra note 115, at 537 (“[C]onsensual encounters are more likely to be used against minorities and the poor . . . .”). No. 4:141] IS IT RECORDING? 175 to “consensual” searches, whereby police will search a person after obtaining their verbal “consent.”118 2. Activation Policies and “Consensual” Interactions Of the police departments that I studied, all required the recording of “consensual” searches. Most also required recording of the request for consent to search. The New York and Philadelphia police departments, however, did not ex- plicitly require the request and consent to be recorded—just the search itself. Unfortunately, Fan does not appear to have considered “consensual” searches as a discrete category of “consensual” interaction. Thus, we are unable to assess whether my finding of a widespread requirement for the re- cording of “consensual” searches is a novel development. The position with respect to “consensual” stops is less satisfactory. Of the police departments that Fan studied, only a quarter required officers to activate their cameras for “consensual” stops.119 These numbers were largely consistent with my study, where, as shown in Appendix C, four of the ten department policies required recording “consensual” stops. This is concerning. In light of the acute danger of racial bias in “consensual” interactions, it is imperative that such interactions be covered by a mandatory activation requirement. As this Note has explained in the preceding sections, the greater the mandatory recording coverage of a body-worn camera policy, the more effective such a policy will be at reducing deliberate racial profiling and implicit racial bias. Take the example of a police officer who is patrolling a par- ticular neighborhood and disproportionately, and without reasonable suspicion, approaches and questions young Black males. The officer would appear to be deliberately racially profiling or acting on implicit racial bias. If the officer is de- liberately racially profiling, then mandatory recording may result in disciplinary action. On the other hand, if the officer is motivated by implicit racial bias, then mandatory record- ing of his behavior will likely result in it being detected by a superior who could then engage him in remedial training. 118 GLASER, supra note 26, at 36. 119 Fan, supra note 6, at 932. 176 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 Either way, the camera needs to be recording in order to en- gage the relevant theory of change. It is for this reason that the policy proposed in the next section requires recording of all “consensual” interactions, both stops and searches. D. Proposal for a Minimal Discretion Activation Policy The differences, highlighted above, between Fan’s December 2015 study of activation policies and my December 2017 study of policies suggest that, at least among the larger police departments, there is a trend from limited to minimal activation policies. The preceding discussion argued that this trend furthers the capacity of body-worn cameras to combat racially biased policing.120 Importantly, however, this Note has suggested that we ought to consider an even more ambitious activation policy—one that requires recording of all police-civilian interactions—if we want body-worn cameras to more effectively limit both deliberate racial profiling and implicit racial bias. What follows is a draft proposal for such a policy.121 Definitions 1. For the purposes of this rule, the following definitions apply: a. Activation: the turning on of the re- cording function on a body-worn camera.122 120 Cf. Roy, supra note 73, at 14 (“[Potential downsides of a mini- mal discretion policy include] a more rigid or inflexible form of manage- ment control could convey to line officers a message of distrust or stifle their ability to make sound decisions under pressure.”). 121 For other aspirational draft policies, see Kampfe, supra note 17, at 1196 (“Participants should manually activate both the audio and visual features of their BWC prior to ‘responding to a call for service or at the initiation of any other law enforcement or investigative encounter between a police officer and a member of the public,’ unless doing so would be ‘unsafe, impossible, or impractical.’ ” (footnote omitted)); and ACLU, A MODEL ACT FOR REGULATING THE USE OF WEARABLE BODY CAMERAS BY LAW ENFORCEMENT § 1(b) (2017), https://www.aclu.org/sites/default/files/field_ document/aclu_police_body_cameras_model_legislation_jan_2017.pdf [https:// perma.cc/32CC-SVDS]. 122 Cf. PHILA. POLICE DEP’T, DIRECTIVE 4.21, BODY-WORN CAMERAS § 5(A) (2016) (“Activate—Any process or action that causes a Body-Worn No. 4:141] IS IT RECORDING? 177 b. Deactivation: the turning off of the recording function on a body-worn camera. c. Confidential police informant: a person who an officer reasonably be- lieves to be seeking to anonymously report a crime or anonymously as- sist in an ongoing law enforcement investigation.123 Mandatory Activation of Body-Worn Cameras 2. While on duty, a member must activate their body-worn camera immediately prior to all interactions with civilians.124 3. If, for whatever reason, a member is unable to activate their body-worn camera prior to the commencement of a civilian interaction, the officer must activate the device as soon as possible thereafter.125 4. The civilian interactions described in Rules 2 and 3 include, but are not limited to: a. Welfare checks; b. Community caretaking functions; c. Calls for service; d. Consensual interactions; e. Consent searches (including search- es of people, items, vehicles, build- ings, and places); f. Requests for consent to search; g. Nonconsensual searches (including searches of people, items, vehicles, buildings, and places); h. Investigatory or enforcement stops (including stops of pedestrians and vehicles); Camera to begin recording and storing both audio transmissions and visual images.”). 123 This provision is based on ACLU, supra note 121, § 1(d)(3). 124 Cf. PHILA. POLICE DEP’T, supra note 122, at § 4(A) (“Authorized Body-Worn Cameras will be activated prior to responding to all calls for service and during all law enforcement related encounters and activities involving the general public.”). 125 Cf. NYPD, supra note 71, § 7 (“Activate the BWC as soon as it is feasible and safe to do so after taking necessary police action to preserve human health and safety.”). 178 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 i. Pursuits (including pursuits of pedestrians and vehicles); j. Interviews with, or statements by, victims, witnesses, or suspects; k. Summonses; l. Arrests; m. Arrestee transport; and n. Uses of force. Notification of Recording 5. As soon as reasonably practicable after activating their body-worn cameras, a mem- ber must notify civilians that the interaction is being recorded.126 Exceptions to Mandatory Activation Requirements127 6. The mandatory activation requirements described above in Rules 2, 3, and 4 do not apply in the circumstances described be- low. In these circumstances, and only these circumstances, a member must refrain from activating their body-worn camera: a. Where the member reasonably be- lieves that recording would com- promise officer or civilian safety; b. Where, in an interaction with a con- fidential police informant, the in- formant objects to the recording;128 c. Where the member is present at a “strip search” of a civilian;129 d. Where, in an interaction with an apparent victim of an offense, the victim objects to the recording;130 126 This provision is based on id. § 4. 127 This Note has not focused on specific exceptions to mandatory recording requirements. Nevertheless, it offers a sampling of commonly endorsed exceptions for illustrative purposes. Such exceptions should ul- timately be formulated with input from affected community members. 128 Cf. NYPD, supra note 71, § 10(e) (prohibiting recording when “[i]nterviewing a current or potential confidential informant”). 129 This provision is based on CHI. POLICE DEP’T, supra note 75, § IV(A)(4), and PHILA. POLICE DEP’T, supra note 122, § 4(C)(1)(d). 130 Cf. NYPD, supra note 71, § 10(g) (categorically prohibiting record- ing of interviews with victims of sexual offenses). No. 4:141] IS IT RECORDING? 179 e. Where a member is in a nonpublic area (including a civilian’s home) and the member does not have a reason to lawfully search that area without a warrant;131 f. Where the member reasonably be- lieves that recording the interaction may compromise patient confidenti- ality;132 g. Where the interaction takes place in a religious institution and the mem- ber does not have permission to rec- ord from the management of that institution;133 or h. Where the interaction takes place on the grounds of any public, private, or parochial elementary or secondary school.134 Deactivation135 7. Having activated their body-worn camera, a member must deactivate their camera in the following circumstances, and the following circumstances only: a. Where the member comes to rea- sonably believe that the circum- stances do not fall, or no longer fall, within the mandatory activation re- quirements of Rules 2, 3, and 4; b. Where the member comes to rea- sonably believe that the interaction falls within the circumstances de- scribed in Rules 6(a)–(h); or c. Where the civilian interaction has concluded. 131 This provision is based on CHI. POLICE DEP’T, supra note 75, § IV(A)(1). 132 Cf. NYPD, supra note 71, § 10(j) (prohibiting recording “inside of a medical facility unless engaging in a police action”). 133 This provision is based on PHILA. POLICE DEP’T, supra note 122, § 4(B)(4). 134 This provision is based on ACLU, supra note 121, § 1(h). 135 Cf. Kampfe, supra note 17, at 1196–97 (affording police discretion in the decision to deactivate their cameras). 180 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 8. Immediately prior to deactivating their camera, a member must, if practicable, an- nounce the reason that they are deactivating their camera.136 9. Having deactivated their camera, a mem- ber must again activate their camera if cir- cumstances change such that the reason for deactivation is no longer operative and the situation falls within the mandatory acti- vation requirements of Rules 2, 3, and 4. Notification to Superiors 10. A member will report any and all unre- corded civilian interactions, and partially recorded civilian interactions, to their su- pervisor. This report must include the reason that the interaction was not recorded or partially recorded.137 The above policy is not intended to be a one-size-fits-all solution. Rather, this proposal provides police departments and advocacy bodies with a concrete and realizable model of how to best minimize racially biased police practices through the use of body-worn cameras. Importantly, this proposal is only a starting point for what is necessarily a local, context- specific discussion. Any efforts to adopt this model ought to be preceded by meaningful community consultation, especially with regard to the countervailing privacy interests. VII. CONCLUSION In order to use body-worn cameras most effectively to curb racially biased policing, police departments ought to require recording of all police-civilian interactions, including “consensual” interactions. Previous research has reaffirmed the positive impact of body-worn cameras on police behavior when officers are required to record all police-civilian 136 This provision is based on CHI. POLICE DEP’T, supra note 75, § III(B)(4). 137 This provision is based on NYPD, supra note 71, § 6, and CHI. POLICE DEP’T, supra note 75, § III(B)(4). No. 4:141] IS IT RECORDING? 181 interactions.138 Such a policy would also better align with a growing public preference for policies requiring recording of “all interactions with the public,”139 and the original aims of civil rights groups who, in the wake of Ferguson, pushed for mandatory recording of “every police-civilian encounter.”140 Furthermore, such a policy would duly pay heed to the fact that Black and Latinx people are more likely than White people, and much more likely than police officers, to favor such a policy.141 It is appropriate to sound two cautionary notes. Police policies ought to be shaped by the communities who will be subjected to them. This is especially true in the case of body-worn cameras, given the extraordinary power they have to erode civilian privacy.142 As Barry Friedman and Maria Ponomarenko have written: In a nation that . . . speaks endlessly of democratic engagement and the popular will, policing is a distinct outlier. Of all the agen- cies of executive government, those that “police” . . . are the most threatening to the liberties of the American people. Yet, from the standpoint 138 See Ariel et al., supra note 12; see also Ariel et al., supra note 45, at 461 (“[Body-worn cameras can deter police from using force] but only in situations where police relinquish some discretion on activating these devices.”). 139 THE POLICING PROJECT, supra note 17, at 12 (“Some 64% of respondents [from the public] favored officers recording all interactions with the public . . . .”). 140 See, e.g., Press Release, Lawyers’ Comm. for Civil Rights Under Law et al., supra note 44. 141 Compare THE POLICING PROJECT, supra note 17, at 8 (recording survey results suggesting that Blacks and Latinxs are more likely than Whites to support a policy requiring officers to record all interactions with the public), with NYPD, supra note 91, app. A (question 6) (reporting that only twenty-six percent of surveyed officers believed that they should be required to record all interactions with members of the public). 142 See Christopher Slobogin, Community Control over Camera Surveillance: A Response to Bennett Capers’s Crime, Surveillance, and Communities, 40 FORDHAM URB. L.J. 993, 996–98 (2013) (arguing for the importance of community authorization of camera surveillance systems); see, e.g., WASH. REV. CODE § 10.109.020 (2018) (“Any ordinance or resolution authorizing the use of body worn cameras should identify a community involvement process for providing input into the development of operational policies governing the use of body worn cameras.”). 182 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 of democratic governance, they are the least regulated.143 The body-worn camera policy space, where community con- sultation is already happening and should continue, is slowly correcting this democratic deficit.144 The model that this Note has suggested is thus not prescriptive, but submitted as a useful starting point for discussions between police de- partments and affected communities across the country as the body-worn camera revolution continues. Though body-worn cameras may go some way to curb- ing racial bias in policing, they are almost certainly not the panacea that some people initially hoped they would be.145 In order to best reduce racial bias in policing, what is needed is change in recruiting, organizational culture, training, and accountability mechanisms.146 This will take time. In the meantime, body-worn cameras can complement the broader change strategy. 143 Barry Friedman & Maria Ponomarenko, Democratic Policing, 90 N.Y.U. L. REV. 1827, 1830–31 (2015) (footnote omitted). 144 See Fan, supra note 16, at 402 n.26 (providing examples of different methods of community consultation on body-worn camera policies). 145 Cf. Ariel et al., supra note 79, at 294 (“[E]nthusiasm for a technological ‘fix’ to the perceived crisis in police legitimacy is unsurprising, as it is far cheaper to implement technology than [to] retrain officers or solve more endemic social problems.”); Developments in the Law¾Policing, supra note 58, at 1797 (“Their adoption should also not be used as an excuse to stifle continued conversation about the root causes of police violence and fractured community relations, as body cameras alone will never be the hoped-for cure-all.”). 146 Cf. Katheryn Russell-Brown, Body Cameras, Police Violence, and Racial Credibility, 67 FLA. L. REV. F. 207, 213 (2016) (“Requiring police to wear body cameras may do some good. However, we must address why Black claims of state violence have historically been dismissed as incredible, non-critical, and rare. We cannot look to cameras (or other technologies), to solve a problem that has historical roots in racial discrimination.”). No. 4:141] IS IT RECORDING? 183 VIII. APPENDIX A: STATED OBJECTIVES OF POLICE DEPARTMENT BODY-WORN CAMERA POLICIES Accountability Transparency Training New York Police (NY) X - - Chicago Police (IL) - X X Los Angeles Police (CA) X - X Philadelphia Police (PA) - - X Houston Police (TX) X X X Washington Metropolitan Police (DC) - - X Dallas Police (TX) - X X Phoenix Police (AZ) X - X Baltimore Police (MD) X X X Miami-Dade Police (FL) X X X TOTAL 6 5 9 184 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 IX. APPENDIX B: DEGREE OF DISCRETION AFFORDED TO OFFICERS IN ACTIVATION POLICIES Limited Discretion Minimal Discretion New York Police (NY) X - Chicago Police (IL) - X Los Angeles Police (CA) - X Philadelphia Police (PA) - X Houston Police (TX) - X Washington Metropolitan Police (DC) X - Dallas Police (TX) - X Phoenix Police (AZ) - X Baltimore Police (MD) - X Miami-Dade Police (FL) - X TOTAL 2 8 No. 4:141] IS IT RECORDING? 185 X. APPENDIX C: ACTIVATION POLICIES REQUIRING MANDATORY RECORDING OF “CONSENSUAL” INTERACTIONS “Consensual” Stops “Consensual” Searches Requests for Consent New York Police (NY) - X - Chicago Police (IL) - X X Los Angeles Police (CA) X X X Philadelphia Police (PA) - X - Houston Police (TX) X X X Washington Metropolitan Police (DC) X X X Dallas Police (TX) - X X Phoenix Police (AZ) - X X Baltimore Police (MD) - X X Miami-Dade Police (FL) X X X TOTAL 4 10 8 186 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 XI. APPENDIX D: METHODOLOGY A. Choice of Sample My sample is made up of the ten largest metropolitan police departments in the United States (as measured by the number of full-time sworn personnel). I chose to study metropolitan police departments, rather than specialized agencies also using body-worn cameras, in order to capture the widest variety of police activities and the highest number of police-civilian interactions.147 Similarly, I focused on the largest police departments because these are likely to serve the largest number of civilians and arguably are more likely than smaller departments to have expended considerable time and energy in formulating body-worn camera policies. To determine which departments fell within my pa- rameters, I relied upon data from the most recent Law Enforcement Management and Administrative Statistics (LEMAS) Survey.148 This survey is sponsored by the Bureau of Justice Statistics, a subsidiary body within the Depart- ment of Justice.149 The most recent survey data is from 2013 and was reported in May 2015.150 The police departments in my study were, from largest to smallest: New York (NY), Chicago (IL), Los Angeles (CA), Philadelphia (PA), Houston (TX), Washington (DC), Dallas (TX), Phoenix (AZ), Baltimore (MD), and Miami-Dade (FL).151 147 Cf. Fan, supra note 16, at 424 (“The data collection focused on the primary police department serving each city because the portfolio of law enforcement activities by the municipal police department is broader than specialized agencies. Moreover, the primary municipal police agency typically serves the greater portion of the city area and more people.” (footnote omitted)). 148 BRIAN A. REAVES, U.S. DEP’T OF JUSTICE, BUREAU OF JUSTICE STATISTICS, LOCAL POLICE DEPARTMENTS, 2013: PERSONNEL, POLICIES, AND PRACTICES (Irene Cooperman et al. eds., 2015), https://www.bjs.gov/content/ pub/pdf/lpd13ppp.pdf [https://perma.cc/KN94-X95U]. 149 See id. at 1 (“The LEMAS Survey, conducted periodically since 1987, collects data on a range of topics from a nationally representative sample of state and local law enforcement agencies.”). 150 See id. 151 Id. at 14 app., tbl.2. No. 4:141] IS IT RECORDING? 187 My relatively small sample number (n=10) was in- formed by my study’s time and resource constraints and constitutes an admittedly significant study limitation. Nev- ertheless, my sample size includes geographically diverse cities and populations of dramatically different sizes (from New York’s population of approximately 8,500,000152 to Miami’s population of approximately 500,000153). Fur- thermore, the policies of these ten departments are worthy of close study because the policy choices made by these large departments are most likely to be imitated by other, smaller departments.154 B. Obtaining the Policies All of the police departments included in my study had publicly available policies on their officers’ use of body- worn cameras. Two invaluable resources in locating the poli- cies were the websites of the Brennan Center155 and the Reporters Committee for Freedom of the Press.156 Other use- ful web-based policy and legislation aggregators were the Urban Institute157 and Arizona State University.158 I supplemented the few gaps in coverage from these websites 152 New York, New York, CITY-DATA.COM, http://www.city-data .com/city/New-York-New-York.html [https://perma.cc/CEP5-3ULP]. 153 Miami, Florida, CITY-DATA.COM, http://www.city-data.com/ city/Miami-Florida.html [https://perma.cc/L3D5-UCSC]. 154 See Fan, supra note 16, at 424 (“[P]rominent cities help set the standards for others to emulate.”); Charles R. Shipan & Craig Volden, The Mechanisms of Policy Diffusion, 52 AM. J. POL. SCI. 840, 843 (2008) (“The likelihood of a city adopting a policy increases when its nearest bigger neighbor adopts the same policy.”). 155 Police Body-Worn Camera Policies, BRENNAN CTR. FOR JUST., https://www.brennancenter.org/body-cam-city-map [https://perma.cc/DY4X- 7MFG] (last updated Sept. 26, 2016). 156 Access to Police Body-Worn Camera Video, REPORTERS COMMITTEE FOR FREEDOM PRESS, https://www.rcfp.org/bodycams [https://perma.cc/NK4 G-GADB]. 157 Police Body-Worn Cameras: Where Your State Stands, URB. INST., http://urbn.is/bodycam [https://perma.cc/L86T-PTYW]. 158 10 Body-Worn Camera Programs: How Do They Compare?, ARIZ. ST. U. SCH. CRIMINOLOGY & CRIM. JUST., https://ccj.asu.edu/content/10- body-worn-camera-programs-how-do-they-compare [https://perma.cc/2DCT- 7QMT]. 188 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1 with independent web-based research, including reference to individual department websites.159 The policies are as in force on December 31, 2017. C. Specific Methodology: Appendix A My methodology for the results generated in Appendix A was to search the department policies for the words “accountability,” “transparency,” and “training.” In the case of “accountability” and “transparency,” I searched only the portion of the policy devoted to objectives or goals. In the case of “training,” I searched both the objectives and goals portions of the policy and also searched for any mention of the word “training” as a consequence of noncompliance with department policy. D. Specific Methodology: Appendix B My methodology for the results generated in Appendix B was to measure the department activation poli- cies against my binary definitions of the “limited discretion” model160 and the “minimal discretion” model.161 Put simply, I categorized a policy as a limited discretion model if it only imposed a mandatory recording requirement with respect to specifically enumerated events. Contrapuntally, I categorized a policy as a minimal discretion model if it imposed a mandatory recording obligation stated in general terms (for example, “[o]fficers shall . . . record all law enforcement activities”162). 159 Cf. Fan, supra note 6, at 930 n.183 (adopting similar methods for obtaining department policies). 160 See supra Section V(A) (“Limited discretion activation policies operate by presumptively providing police officers discretion as to whether or not to record their activities while on duty. This presumption of officer discretion is only displaced in specific circumstances (in which recording is mandatory).” (footnote omitted)). 161 See supra Section V(B) (“[Minimal discretion activation] policies start from a presumption that all enforcement activities will be recorded, subject only to narrow exceptions.”). 162 HOUS. POLICE DEP’T, GENERAL ORDER NO. 400-28, BODY WORN CAMERAS § 6, at 3 (2017). No. 4:141] IS IT RECORDING? 189 E. Specific Methodology: Appendix C The results presented in Appendix C represent per- haps the most contestable of my findings because they are based in part on my definition of “consensual” interactions. In order that interested readers can test my results against their own definitions of “consensual” interactions, I include in the footnotes below the specific provisions of the various policies that I adjudged to support my findings. The policies I understood to require mandatory recording of “consensual” interactions were: Los Angeles,163 Houston,164 Washington, D.C.,165 and Miami.166 I adjudged all ten policies to cover “consensual” searches.167 Finally, I characterized eight of the policies (all except the New York and Philadelphia policies) to require mandatory recording of requests for consent to search.168 163 See L.A. POLICE DEP’T, OFFICE OF THE CHIEF OF POLICE, SPECIAL ORDER NO. 12, BODY-WORN VIDEO PROCEDURES—ESTABLISHED § III (2015) (requiring recording of all “[p]edestrian stops (including officer-initiated consensual encounters)”). 164 See HOUS. POLICE DEP’T, supra note 162, at 1, 3 (requiring recording of “[a]ny event during which an officer exercises his police authority or conducts any type of investigation, whether consensual or otherwise”). 165 See METRO. POLICE DEP’T OF THE D.C., GO-SPT-302.13, BODY- WORN CAMERA PROGRAM § V(A)(3) (2016) (requiring recording of all “self- initiated police action”). 166 See MIAMI-DADE POLICE DEP’T, DIRECTIVE 16-18, BODY-WORN CAMERA SYSTEM §§ V(H), VII(B) (2016) (requiring recording of all instances “where a law enforcement officer, acting in an official capacity, comes in contact with the general public”). 167 See NYPD, supra note 71, § 5(e); CHI. POLICE DEP’T, supra note 75, § III(A)(2)(j); L.A. POLICE DEP’T, supra note 163; PHILA. POLICE DEP’T, supra note 122, § 4(A); HOUS. POLICE DEP’T, supra note 162, at 1; METRO. POLICE DEP’T OF THE D.C., supra note 165, § V(A)(4)(p)(1); DALL. POLICE DEP’T, GENERAL ORDER 332.00, BODY WORN CAMERAS § 332.04(A)(1)(g) (2015); PHX. POLICE DEP’T, OPERATIONS ORDER 4.49, BODY-WORN VIDEO TECHNOLOGY § 3(B)(3)(a) (2018); BALT. POLICE DEP’T, supra note 77; MIAMI- DADE POLICE DEP’T, supra note 166, § VII(B)(7) (requiring recording of all searches, presumably including “consensual” searches). 168 See CHI. POLICE DEP’T, supra note 75, § III(2)(l); L.A. POLICE DEP’T, supra note 163; HOUS. POLICE DEP’T, supra note 162, at 1, 3; METRO. POLICE DEP’T OF THE D.C., supra note 165; DALL. POLICE DEP’T, supra note 167; PHX. POLICE DEP’T, supra note 167; BALT. POLICE DEP’T, supra note 77; MIAMI-DADE POLICE DEP’T, supra note 166. 190 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 9:1