Investigation of the Aurora Police Department and Aurora Fire Rescue, 2021
Download original document:
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
Investigation of the Aurora Police Department and Aurora Fire Rescue September 15, 2021 Table of Contents 1. Executive Summary .................................................................................................................. 1 2. Scope of Investigation and Methodology ............................................................................. 4 3. 2.1. Basis for Investigative Authority and Reason for Investigation ......................... 4 2.2. Scope of Investigative Mandate .............................................................................. 4 2.3. Our Work in the Investigation ................................................................................. 4 Background ................................................................................................................................ 6 3.1. Aurora Background ................................................................................................... 6 3.2. Aurora Police .............................................................................................................. 6 3.2.1. Notable Recent Incidents............................................................................ 6 3.2.2. Morale and Retention Issues ...................................................................... 9 3.2.3. Chief of Police History............................................................................. 10 3.2.4. Structure of the Organization.................................................................. 10 3.2.5. Training—Changes at the Police Academy ............................................ 11 3.2.6. Training—Changes to In-Service Ttraining ........................................... 12 3.2.7. Accountability Mechanisms ...................................................................... 12 3.3. Aurora Fire Rescue .................................................................................................. 15 3.3.1. Falck Rocky Mountain ............................................................................... 15 3.3.2. Ketamine Use by Aurora Fire .................................................................. 16 3.4. Aurora City Council ................................................................................................. 16 3.4.1. Recent Changes .......................................................................................... 16 3.4.2. Report and Recommendations from 21CP Solutions .......................... 18 3.5. 4. 5. Aurora Civil Service Commission ......................................................................... 19 Successes of Aurora Police and Fire .................................................................................... 21 4.1. Early Adoption and Extensive Use of Body-Worn Cameras ........................... 21 4.2. The Force Review Board to Review Use-of-Force Incidents ........................... 21 4.3. The Creation of a Force Investigation Unit to Allow Affirmative Investigations of Uses of Force ............................................................................ 22 4.4. Change in Training Philosophy and Culture at Academy .................................. 22 4.5. Expanded Community Involvement..................................................................... 23 4.6. Recognition of Need for Improvement and Change by Senior Management ............................................................................................................. 24 The Effect of This Investigation .......................................................................................... 25 5.1. Increased Scrutiny of Police Does Not Increase Crime Rates ......................... 25 i 6. 7. 5.2. Policy Restrictions on Uses of Force Are Associated With Improved Officer Safety............................................................................................................ 26 5.3. Policy Restrictions on Uses of Force Are Associated With Increased Safety for Justice-Involved Residents ................................................................... 27 5.4. Policy Restrictions on Police Are Not Associated With Less Engaged Policing ...................................................................................................................... 27 5.5. Successful Reform Efforts Are Comprehensive ................................................. 28 Pattern and Practice Findings: Legal Background ............................................................. 30 6.1. Limitations on Discriminatory Policing Practices .............................................. 30 6.2. Legal Requirements for Stops ................................................................................ 31 6.3. Legal Requirements for Use of Force and Arrests............................................. 32 6.4. Legal Requirements for Use of Ketamine ........................................................... 34 6.5. Recent Changes to Colorado Law ......................................................................... 34 Finding #1: Aurora Police Has a Pattern and Practice of Engaging in Racially Biased Policing Against People of Color as a Whole and Black People in Particular ................................................................................................................................... 37 7.1. Background ............................................................................................................... 37 7.2. The Team’s Investigation of Aurora Police’s Data ............................................. 39 7.2.1. Assembling the Data Analytics Team ..................................................... 39 7.2.2. Data Collection ........................................................................................... 40 7.3. Data Analysis ............................................................................................................ 41 7.3.1. Interactions.................................................................................................. 41 7.3.2. Arrests .......................................................................................................... 45 7.3.3. Reported Uses of Force ............................................................................ 48 7.4. Explanations for the Data ...................................................................................... 51 7.4.1. Geography, Gender, and Age and Use of Force................................... 51 7.4.2. Type of Offense or Level of Force Used .............................................. 52 7.4.3. Income ......................................................................................................... 52 7.4.4. Officer Discretion ...................................................................................... 54 7.4.5. Use of Population-Level Demographics ................................................ 55 7.4.6. Excluding Incidents Where Race Information is Unknown or Unavailable .................................................................................................. 56 7.5. The Data Combined With Other Evidence Shows a Pattern and Practice of Race-Based Policing ............................................................................ 57 7.6. Changes Needed ...................................................................................................... 57 7.6.1. Improved and More Detailed Policies and Guidance to Prevent Racially Biased Policing .............................................................. 57 ii 7.6.2. More Specific Standards and Expectations on When Officers Should Stop, Arrest, or Use Force........................................................... 58 7.6.3. Track Outcomes for Those Arrested for Misdemeanors .................... 61 7.6.4. Improved Academy and In-Service Training ......................................... 61 7.6.5. Better Timely Recordkeeping ................................................................... 62 7.6.6. Improved Hiring and Recruiting Procedures, Including Major Changes to the Civil Service Commission Standards ........................... 63 8. Finding #2: Aurora Police Has a Pattern and Practice of Using Force Excessively ................................................................................................................................ 67 8.1. Law and Aurora Policies ......................................................................................... 67 8.2. Data and Information Reviewed ........................................................................... 69 8.2.1. Documents Relating to the Use of Force .............................................. 69 8.2.2. Force Review Board and Other Meetings .............................................. 69 8.2.3. Ride-Alongs ................................................................................................. 69 8.2.4. Community Concerns Based on Interviews .......................................... 70 8.3. Findings ..................................................................................................................... 71 8.3.1. Force and Threatened Force Used Against Those in Mental Health Crises ............................................................................................... 71 8.3.2. Using Force When the Only Criminal Violation is Failure to Obey ............................................................................................................. 76 8.4. Causes of Violations ............................................................................................... 80 8.4.1. A Culture at Aurora Police That Emphasizes Justification for Force, Rather Than Whether Force Was Lawful and Appropriate ................................................................................................. 80 8.4.2. Use-of-Force Policy and Implementation Focuses on Maximum Force Permitted Under Law .................................................. 83 8.4.3. De-Escalation is Misunderstood and Not Consistently Practiced ...................................................................................................... 83 8.4.4. Mental Health Issues Are Not Adequately Addressed ......................... 85 8.4.5. Force is Used Disproportionately Against the Black Community ................................................................................................. 85 8.4.6. Aurora Police Inadequately Documents the Use of Force ................. 86 8.4.7. Aurora Fire and Aurora Police Do Not Adequately Coordinate ................................................................................................... 86 8.5. Changes Needed ...................................................................................................... 87 8.5.1. Improve Use-of-Force Policies to Give More Specific Guidance to Support Officers ................................................................. 87 8.5.2. Enhance Training on New Policies ......................................................... 87 iii 8.5.3. Change the Focus and Purpose of the Force Review Board and the Force Investigation Unit ............................................................. 88 8.5.4. Develop Measurable Goals to Improve How Officers Engage With Those Experiencing Mental Health Crises................................... 88 8.5.5. Develop a Joint Policy for Aurora Police and Fire to Coordinate on Scenes and Conduct Joint Trainings............................. 88 9. Finding #3: Aurora Police Has a Pattern and Practice of Failing to Document Stops as Required by Law ...................................................................................................... 89 9.1. Legal Background .................................................................................................... 89 9.2. Aurora Police Fails to Document All Stops as Required by Senate Bill 217 .............................................................................................................................. 89 9.3. Aurora Police Reports Do Not Include Required Detail Describing the Reasons Supporting the Contact or Stop ...................................................... 90 9.4. Changes Needed ...................................................................................................... 91 9.4.1. Develop Policies and Create Systems to Comply With Colorado Law on Documenting Stops ................................................... 91 9.4.2. Amend Aurora Police Policies to Include a Separate Directive Addressing Stops That Complies With both Colorado and Federal Law ................................................................................................. 91 10. Finding #4: Aurora Fire Rescue Has a Pattern and Practice of Administering Ketamine Illegally .................................................................................................................... 92 10.1. Ketamine and Its Use by Aurora Paramedics ...................................................... 92 10.1.1. Ketamine Use by Paramedics ................................................................... 92 10.1.2. Changes to Colorado Law on the Use of Ketamine ............................ 94 10.1.3. Aurora’s Suspension of Ketamine Use in 2020..................................... 95 10.1.4. Structure of Aurora Emergency Medical Services ............................... 95 10.2. Findings on Aurora’s Practices of Ketamine Use............................................... 95 10.2.1. Records of Past Ketamine Use Show Disparate Dosing of Ketamine Not Based on Weight and a Failure to Follow Monitoring Protocols ................................................................................ 95 10.2.2. Officers Sometimes Requested Chemical Restraints and Failed to Document That Paramedics Chemically Restrained the Person .......................................................................................................... 99 10.3. Findings on Aurora’s Ketamine Protocols ......................................................... 103 10.3.1. Protocols Did Not Emphasize the Need for Paramedics to Conduct an Independent Examination of a Patient .......................... 103 10.3.2. National Ketamine Protocol Trends and Recommendations From Medical Associations Conflict With the High Doses Permitted by Aurora Fire’s Protocol ..................................................... 106 iv 10.4. Required Changes .................................................................................................. 107 10.4.1. Review Ketamine Dose Recommendations ......................................... 107 10.4.2. Maintain a Uniform Method to Assess Patient Agitation.................. 107 10.4.3. More Stringent Review to Ensure Protocol Compliance .................. 108 11. Facilitating Meaningful Reform .......................................................................................... 110 12. Next Steps .............................................................................................................................. 111 13. Acknowledgments ................................................................................................................. 112 v 1. Executive Summary We, an investigation team appointed by Colorado’s Attorney General, Phil Weiser, find that the Aurora Police Department and Aurora Fire Rescue have a pattern and practice of violating state and federal law. Specifically, we find that Aurora Police1 has a pattern and practice of racially biased policing, using excessive force, and failing to record required information when it interacts with the community. Aurora Fire has engaged in a pattern and practice of administering ketamine in violation of the law. Our 14-month investigation included extensive data analysis and direct observation of Aurora Police during over 220 hours of in-person ride-alongs with officers on patrol and firefighters on duty. We assembled and examined a dataset of over three million records from Aurora Police’s internal systems. We attended nine months of weekly Force Review Board meetings, where we reviewed body-worn camera footage of officers using force, and observed how Aurora Police evaluates the conduct of its officers. We read 2,800 reports and associated documents from the last five years about the use of force by Aurora Police officers. We evaluated Aurora Police’s use of force and related policies put in place after legislation changed in 2020,2 heard community feedback from scores of community members, and conducted dozens of interviews with Aurora Police and Fire employees. Over the course of our investigation, we saw consistent patterns of unlawful behavior by Aurora Police and Aurora Fire. Aurora Police has a pattern and practice of racially biased policing. With the help of a team of professional statisticians and econometricians, we conducted a detailed analysis of internal Aurora Police data reflecting police activity since 2018. We observed statistically significant racial disparities—especially with respect to Black individuals—in nearly every important type of police contact with the community, from interactions to arrests to uses of force. These disparities persisted across income, gender, and geographic boundaries. Together with the other information we reviewed, we find that Aurora Police engages in racially biased policing, treating people of color (and Black people in particular) differently from their white counterparts. Aurora Police has a pattern and practice of using excessive force. We found that Aurora Police has repeatedly engaged in unlawful and unconstitutional uses of force, regularly applying greater force than reasonably warranted by the situation. For example, we observed officers using force to take people to the ground without first giving them adequate time to respond to officer commands, or generically reciting “stop resisting” when trying to control subjects, even though it appeared from other available evidence that the subject was not resisting. We observed officers immediately escalating in circumstances where the subject was in obvious mental health distress but not presenting an imminent risk of harm to themselves or others. We observed officers using force on individuals who had not committed any crime and presented no danger but who simply refused to comply For readability, we refer to the Aurora Police Department as Aurora Police and Aurora Fire Rescue as Aurora Fire throughout this report. 2 We did not review Aurora’s historical policies, including those related to stops or use of force in place prior to the major legislative change in 2020 in Senate Bill 217. 1 1 with orders. In our review of Force Review Board meetings, we saw Aurora Police continuing to analyze force only under pre-existing federal requirements rather than under the more stringent requirements of Colorado law. In addition, our investigation found that Aurora Police has a misplaced view of de-escalation, focusing on whether officers calm down after using force rather than avoiding unnecessary escalation in the first place. Aurora Police has a pattern and practice of failing to document stops as required by law. Colorado’s landmark law enforcement reform law, Senate Bill 20-217, requires that officers have a legal basis for making any contact with a member of the public, and also imposes strict recordkeeping requirements whenever any such contact is made. We found that under policies that have been in place since 2020 (after Senate Bill 217 was enacted), members of Aurora Police conduct resident stops without documenting those stops as the law requires. This leads to an entire category of police interactions (so-called investigative or Terry stops) for which there is little to no documentation, and as a result, even less scrutiny. Compounding this problem, Aurora Police policies in effect since 2020 do not provide adequate guidance to officers on when a Terry stop is appropriate under the law. While it administered ketamine, Aurora Fire had a pattern and practice of using ketamine in violation of the law. Through a comprehensive review of Aurora Fire’s ketamine and care reports, we saw a consistent pattern of illegal ketamine administrations. Aurora Fire relied on a ketamine review process that did not adequately ensure that paramedics followed the legal requirements for administering ketamine. Its review process failed to identify problems when ketamine was inappropriately administered or was administered at the request of police, and therefore did not improve its processes and training to prevent future violations. We then determined why Aurora Police and Fire violate the law. Most failures with Aurora Police relate to systematic and severe culture problems, created by several factors. First, Aurora Police’s culture leads to the frequent use of force, often in excess of what the law permits. Training is ad hoc and does not address specific needs of the organization as shown by officer behavior. Policies are short on detail or practical guidance, often doing little more than reciting the legal requirements set forth in court cases and applicable statutes or regulations. In short, Aurora Police has failed to create and oversee appropriate expectations for responsible behavior. Second, Aurora Police does not meaningfully review officers’ uses of force, relying on formal and informal systems that favor findings that officers followed policy and that hamper candid feedback on how to improve. Third, the Aurora City Charter gives the Civil Service Commission total control over entry-level hiring and the right to reverse all meaningful discipline that the Chief of Police can impose. The Commission overturns discipline in high-profile cases in a way that undermines the Chief of Police’s authority, such that as a result, Aurora Police officers who violate law or policy often remain on the force. And because of the Commission’s hiring practices, Aurora Police officers do not reflect the diversity of the city. 2 To address these problems, the Attorney General will require Aurora to fix its organizations in several ways; some of which directly relate to practices that violate the law, others of which focus on culture, leadership, and structures that lead to or enable the illegal conduct. Aurora must confront both the violations and their underlying causes to stop these patterns and practices of violating the law. We strongly encourage Aurora to enter a consent decree that will require specific changes and ongoing independent oversight. Aurora has committed to continue to cooperate with our office to try and develop a consent decree. If this effort is unsuccessful, we will seek a court-imposed order correcting these problems. To enable Aurora Police and Fire to make these necessary changes, we will—if they choose to cooperate with our team—work with them to develop specific and tailored solutions with their input, guided by experts on our team, and reflecting input from the community. Our team approached this investigation with an open mind and did not pre-judge any outcome as we carefully considered the evidence. We appreciate that the City of Aurora, Aurora Police, and Aurora Fire cooperated fully with our investigation. The conclusions we reach here are the result of a fair, independent, and objective investigation. Our team brought diverse backgrounds and experience to this work. Several members of our team worked as law enforcement officers for urban police and sheriff ’s departments, totaling 74 collective years of service. One member of our team was a former patrol officer, Chief of Police, and City Manager for Arlington, Texas. Other team members have extensive prosecution or public defender experience, including one team member who was the United States Attorney for Idaho and who worked for several years in the Civil Rights Division at the Department of Justice and another who served as a Deputy Assistant Attorney General for Civil Rights at the Department of Justice. Another team member is a former public defender and served as Deputy Chief Counsel to the President’s Commission on the BP Deepwater Horizon Oil Spill. We appreciate that the risks that first responders take must stand equally alongside a deep commitment to the rule of law. By working to elevate law enforcement, our community can both be safe from danger and free from racial discrimination. Indeed, it is critical that community members can trust that law enforcement operates fairly and is worthy of their trust. In issuing this report, and working to implement a consent decree to address Aurora’s documented failings, we will remain focused on how we can work together to elevate and improve law enforcement in Aurora. 3 2. Scope of Investigation and Methodology 2.1. Basis for Investigative Authority and Reason for Investigation Senate Bill 217, a law enforcement accountability bill enacted in Colorado in 2020, provides the basis for our investigative authority. The law authorizes Colorado’s Attorney General to open a civil investigation of any governmental authority for engaging in a pattern or practice of conduct that violates state or federal constitutions or laws.3 A pattern or practice investigation looks at whether members of a government agency have a pattern of misconduct related to the rights, privileges, or immunities of people that those members interact with. Shortly after the bill passed, Attorney General Phil Weiser began to investigate whether Aurora Police and Aurora Fire have a pattern or practice of violating the United States or Colorado constitutions or federal or state statutes. We find that both Aurora Police and Aurora Fire have a pattern and practice of violating these constitutions and laws. We discuss our findings in more detail below. The death of Elijah McClain in August 2019 and subsequent protests focused attention on Aurora Police and Aurora Fire. Our focus in this report is on the government agencies— Aurora Police and Aurora Fire—and not on any particular incident. Because of ongoing criminal proceedings, we do not address Mr. McClain’s death in this report. 2.2. Scope of Investigative Mandate Our investigation of Aurora Police focused on stops and arrests, uses of force, and possible discrimination in enforcement activities to understand whether Aurora Police had a pattern or practice of violating the law. Once we determined they did, we then looked to understand the causes of those violations, focusing on Aurora Police’s policies and procedures, hiring processes, training, discipline, record-keeping, and culture. As for Aurora Fire, our investigation focused on the use of ketamine. We reviewed the records for each ketamine administration, the protocols and procedures for administering ketamine, situations in which Aurora Fire administered ketamine to a subject after a use of force by the police, and Aurora Fire’s process for reviewing ketamine administration. We also looked at coordination between the police and fire departments when they jointly respond to an emergency. 2.3. Our Work in the Investigation During our investigation, we spoke to current and former personnel at Aurora Police, including command staff, union representatives, training academy personnel, and patrol officers. We participated in over 190 hours of ride-alongs, where we rode with patrol officers and sergeants on duty while they patrolled all three of Aurora’s policing districts, focused on the swing shift, from afternoon to early morning. We visited police 3 § 24-31-113, C.R.S. 4 headquarters and the training academy; observed most Force Review Board meetings from December 2020 through September 2021; and attended several community-police meetings, including Internal Review Board meetings and Citizens’ Police Advisory Team meetings. Finally, we spoke to many interested community members and leaders, including faith leaders, immigrant community leaders, members of various police advisory boards, and many others who expressed interest in sharing their experiences. In each community conversation, we asked community members to refer any others to us who may wish to share their experiences. During the many months of our investigation, we also hosted an intake form on our website for community members to report any concerns related to Aurora Police or Fire to our office. We spoke to leadership, training, union, and other firefighting personnel affiliated with Aurora Fire. We participated in over 30 hours of ride-alongs, covering multiple fire stations. We also observed Aurora Fire on our police ride-alongs. We interviewed others working for the City of Aurora and the Civil Service Commission to learn how those organizations functioned and interacted with Aurora Police and Aurora Fire. Our team reviewed documents and data from both the police and fire departments, and the Colorado Department of Public Health and Environment. We collected and reviewed over 2,800 use-of-force reports from 2016 to 2020. We retained a group of statisticians and economists to conduct a statistical analysis of Aurora Police data on police activity including arrests, use-of-force incidents, and number of resident interactions. And we reviewed Aurora Fire’s procedures for ketamine administration and past incidents where Aurora Fire administered ketamine. 5 3. Background 3.1. Aurora Background Aurora is Colorado’s third-largest city with a population in 2019 of 369,111.4 It is diverse— the four largest ethnic groups in Aurora are white (non-Hispanic) (44.4%), Hispanic/Latino (28.6%), Black (16.5%), and Asian (6.5%).5 About 20% of Aurora residents are foreign-born, with the majority emigrating from Latin America (38,185), Asia (18,857), and Africa (10,536).6 Aurora has grown rapidly, more than doubling in size since the 1970s.7 Aurora Police classifies the city into three districts. District 1 includes most of western Aurora and is subdivided into eleven patrol areas or “beats.”8 District 1 encompasses many of the most diverse areas of Aurora and includes many residents who speak a language other than English. District 2 covers most of northeast Aurora, with eight beats.9 District 3 includes most of southeast Aurora, also with eight beats.10 A Police Area Representative (PAR) Team in each district has responsibility for establishing relationships with community members and responding to resident concerns about their neighborhoods.11 3.2. Aurora Police 3.2.1. Notable Recent Incidents Recent incidents involving Aurora Police increased scrutiny of Aurora Police and heightened community tensions. Shortly after Mr. McClain’s death, three Aurora Police officers took photos of themselves near the scene of the police encounter with a memorial for Mr. McClain in the background. The officers texted the photos to other officers, including one of the officers involved in Mr. McClain’s death.12 Among other things, the photos depicted the officers reenacting a carotid hold and grinning. The Police Chief fired three of the officers, another officer resigned, and the three fired officers appealed their terminations to the Civil Service Commission. The Commission upheld the three terminations. Additionally, nearly a year after Mr. McClain’s death, officers in riot gear broke up a violin vigil honoring Mr. McClain’s life and love for the violin by deploying chemical agents including pepper spray after reporting that some protestors were throwing rocks and the officers ordered the vigil 2019 American Community Survey 5-Year Data Profile, Demographic and Housing Estimates, U.S. Census Bureau. 5 Id. 6 Id. Selected Social Characteristics. 7 City of Aurora, Aurora History. 8 City of Aurora, District 1. 9 City of Aurora, District 2. 10 City of Aurora, District 3. 11 City of Aurora, Police Area Representatives (PAR). 12 3 Officers Fired Over Photos Taken Near Elijah McClain Memorial, The New York Times, July 4, 2020. 4 6 to end.13 This conduct led to a class-action lawsuit against the City of Aurora and various officers by the rally attendees.14 Recent incidents unrelated to Mr. McClain’s death raised further concerns about officer accountability, training, and decision-making. In December 2018, a man died after fighting with Aurora Police officers after they found him choking another person.15 In the course of subduing the man, officers tased him nearly a dozen times, punched and struck him with batons repeatedly, and ultimately restrained him face-down using four sets of handcuffs linked together. He was declared dead about an hour after he was taken into custody. Although the coroner’s office found that the man died of “restraint asphyxia” and ruled the death a homicide, prosecutors did not seek criminal charges against the involved officers. The man’s widow filed a wrongful death lawsuit against Aurora alleging excessive force. In March 2019, two residents found an on-duty Aurora Police officer unconscious, unresponsive, and armed in the driver’s seat of his running police vehicle.16 Although an Internal Affairs investigation found that the officer was alcohol-impaired and had violated multiple policies, he remained employed by Aurora Police and never faced charges for driving under the influence from the incident. An independent report by former United States Attorney John Walsh found Aurora Police’s response deficient in many ways.17 Also that month, an officer shot and killed a man experiencing a mental health crisis in his sister’s apartment, who was alone when officers responded.18 The man had called 911 claiming that he was holding hostages and that two people were dead. After his death, the man’s family filed a lawsuit in which they asserted that the sister told officers that the man was alone and unarmed. Aurora Police did not discipline the involved officers, who alleged that the man had charged them with a machete, and were cleared of criminal wrongdoing. In August 2019, an officer left a female prisoner upside down in the back of his patrol car for more than twenty minutes after her arrest.19 The officer chose to restrain the woman by tying her handcuffed hands to her feet behind her back and she slipped off the back seat as the officer took her to jail. Aurora Police publicly released the officer’s body-worn camera video, which shows the woman with her head on the floor of the car and her legs in the air as she asks the officer for help, says that she cannot breathe, and calls the officer “master.” The officer, who ignored the woman’s requests, was fired. Thousands Call For Justice For Elijah McClain in a Day of Music and Marching, CPR News, June 27, 2020. Inside Lawsuit Over Police Attack on Elijah McClain Violin Tribute, Westword, July 24, 2020. 15 Widow of Man Who Died After Fight With Aurora Police Files Wrongful Death Lawsuit, Sentinel, Dec. 11, 2020. 16 Aurora Cop Drives Drunk On Duty: Keeps Job, No Arrest, CBS Denver, Dec. 10, 2019. 17 City of Aurora, Report on Independent Review, Aurora Police Department’s Response to March 29, 2019 Incident Involving an Aurora Police Officer, March 2020. 18 Federal Judge Denies Immunity to Aurora Officers in Shooting Death Lawsuit, The Gazette, Sept. 8, 2021. 19 ‘Please Don’t Let Me Die Back Here’: Aurora Police Video Released of Woman Cuffed, Hobbled in Back of Patrol Car, CBS Denver, Sept. 29, 2020. 13 14 7 In March 2020, an officer held an Indian-American doctor at gunpoint as the doctor parked in the parking structure of a refugee center that he owned and operated.20 After pulling his car into the structure, the doctor honked at a police cruiser that was blocking his car’s path. An officer got out, approached the doctor’s car, drew his gun, questioned the doctor at gunpoint about what he was doing, and demanded proof that he owned the building. The doctor recorded the incident on his cell phone. The Chief of Police suspended the officer for 40 hours without pay and mandated that he attend de-escalation training. After publicly criticizing this discipline as inadequate, the doctor filed a civil lawsuit alleging excessive force. In August 2020, officers held a Black family at gunpoint and ordered the family members—including four children aged six, twelve, fourteen, and seventeen—out of their SUV, made them lay face down on the hot summer pavement, and handcuffed them.21 Officers later said they confused the family’s car for a stolen vehicle—the vehicles had the same license plate numbers but were from different states, and the stolen vehicle was reportedly a motorcycle, not an SUV. The stop led to a lawsuit alleging inadequate training and lack of probable cause in stopping the family.22 That same month, an officer responded to a man trespassing at a grocery store, who was lying down on the ground with his arms underneath him as a different officer tried to arrest him. The responding officer—who did not activate his body-worn camera— punched the man multiple times in the ribs and then tased him five times in a two-minute period. Aurora Police fired the officer for using excessive force in February 2021.23 And in July 2021, a police officer responding to a trespass call repeatedly struck an unarmed man in the head with his gun, choked him for nearly a minute, and threatened to shoot him.24 A second police officer stood by and did not intervene. Both officers were arrested—the first for attempted first-degree assault, second-degree assault, felony menacing, official oppression, and first-degree official misconduct; the second for misdemeanor charges of failure to intervene and failure to report use of force by an officer. The first officer resigned shortly after the incident, and the Chief of Police fired the second officer for violating Aurora Police’s policy directives and failing to intervene,25 though appeals of that decision to the Civil Service Commission remain active. Reports Colorado Officer Who Pointed Gun at Doctor’s Head on His Own Property Suspended 1 Week, NBC News, Sept. 17, 2020. 21 Family Sues Aurora Police Over Botched Stolen-Car Response That Left Children Handcuffed, Held at Gunpoint, The Denver Post, Jan. 26, 2021. 22 Gilliam v. City of Aurora, Complaint and Jury Demand, Jan. 25, 2021. 23 Aurora Police Officer Fired for Excessive Use of Force, 9 News Denver, Feb. 11, 2021. 24 Police Body Cam Video Shows Aurora Cop Strangling, Pistol-Whipping Trespassing Suspect, Netting Criminal Charges For 2 Cops, Sentinel, July 27, 2021. 25 Aurora Police Officer Involved in Pistol-Whipping Arrest Fired by Police Chief for Failing to Intervene, The Denver Channel, Aug. 12, 2021. 20 8 described the Civil Service Commission hiring the first officer despite a prior arrest for pointing a gun at a roommate while drinking.26 This list of recent incidents is not comprehensive. Aurora Police has been in the news recently for many more issues, including an officer who was recorded asking paramedics to sedate a subject,27 and several officers arrested for DUI, some of whom were fired.28 Each of these incidents renewed media attention and community scrutiny of Aurora Police, both locally and nationwide. 3.2.2. Morale and Retention Issues The recent incidents involving Aurora Police have impacted not only community trust, but also officer retention and morale. Aurora Police saw 150 officers leave or be fired from around January 2020 to July 2021, a turnover rate of nearly 20%.29 In contrast, Aurora Police’s annual turnover rate from 2014 to 2019 averaged just 6.5%. Aurora faces more challenging police retention issues than other police departments—departments in comparable Colorado jurisdictions, including Denver and Colorado Springs, saw much smaller changes in the number of departing officers.30 In interviews and ride-alongs, some officers shared that they felt unsupported by the community, Aurora Police leadership, and local elected officials and leaders, particularly after Mr. McClain’s death. Some expressed frustration at what they perceived to be a lack of accountability, believing that, for example, Aurora Police handled the incident involving the on-duty officer found drunk in his running vehicle poorly, which hurt morale. Others cited more reasons for officers leaving Aurora Police—tensions with the community, confusion over recent legislative police reforms in Colorado, lack of direction and support from police leadership, conflicting orders from various levels of leadership, the multiple ongoing investigations, and general concerns about the future of policing. Aurora Police personnel also expressed concerns that it has not been losing bad officers, but some of its most experienced officers and officers of color. The retention issues have impacted staffing capabilities, the level of experience within Aurora Police, and the composition of the executive leadership team. Although officer Criminal History of Aurora Officer John Haubert Sounds Alarm on Hiring Process, CBS 4 Denver, July 28, 2021. 27 Aurora Body Camera Video Shows Police Asking Medics to Give Powerful Drugs to Suspect to Calm Him, The Denver Channel, Dec. 3, 2020. 28 Aurora Police Sergeant Arrested for DUI in Denver, KDVR News, July 7, 2021; Aurora Police Officer Fired Nearly a Month After Drunk Driving Crash in Colorado Springs, The Denver Channel, Apr. 6, 2020. 29 Aurora Tries to Fight Uptick in Violent Crime, While Losing Officers at Alarming Rate, The Denver Channel, July 2, 2021; ‘Just Unprecedented’: Aurora Police Saw 60% Rise in Officers Leaving Department Last Year, Sentinel, Feb. 19, 2021. 30 Did Colorado Law Enforcement Flee the Profession in 2020? Depends on the Department, Denver Post, Mar. 8, 2021 (noting that the Denver Police Department saw 81 officers leave in 2020, down from 100 the year before, while the Colorado Springs Police Department saw 73 officers leave in 2020, up from a four-year average of 53 officers leaving annually). 26 9 recruitment has remained steady, Aurora Police personnel expressed concern that the eventual effect of the high turnover rate will be a much less experienced department. 3.2.3. Chief of Police History The challenges faced by Aurora Police came in the midst of a leadership transition. Chief Nick Metz, who joined Aurora Police as its chief in 2015, retired at the end of 2019.31 Chief Vanessa Wilson then spent seven months as Interim Chief, and the City Manager and the majority of City Council officially appointed her as Chief in August 2020.32 Chief Wilson became the first female and openly LGBTQ+ individual to lead the police department and remains its chief today. 3.2.4. Structure of the Organization Aurora Police currently employs about 744 sworn police officers and 237 civilians.33 In 2020, out of 758 police officers and recruits, Aurora Police employed 600 white officers (79.2%), 79 Hispanic officers (10.4%), 31 officers of two or more races (4.1%), 28 Black officers (3.7%), 15 Asian officers (2.0%), and 1 Hawaiian/Pacific Islander officer (0.1%).34 Of those officers, 662 were male (87.3%), and 96 were female (12.7%). The Chief of Police oversees an executive leadership team comprised of a deputy chief of police; an executive officer; three division chiefs overseeing the Professional Accountability, Operations, and Metro Divisions; and two managers overseeing the Community Relations and Business Services Divisions.35 The Professional Accountability Division is responsible for professional standards, training, and electronic support. The Operations Division houses most police officers and oversees the three districts, patrol sections, and police area representatives assigned to the three districts. The Metro Division covers the Investigations Bureau and Special Operations Bureau, which include specialized units such as Major Investigations, SWAT, K-9, the Crisis Response Team, and the Traffic Section. The Community Relations section oversees recruiting, Aurora for Youth, and other communityoriented programs. Finally, the Business Services Division handles administrative services, equipment and facilities, records and support, and property and technical services.36 The leadership team also includes six commanders who oversee the three districts, the Internal Affairs Bureau, the Investigations Bureau, and the Special Operations Bureau. The district and Special Operations commanders are supported by four captains, also called deputy commanders. Under the captains are lieutenants and then sergeants, who directly oversee patrol officers. Aurora Police Chief Nick Metz Announces Retirement, to Leave Department at End of 2019, The Denver Channel, Sept. 27, 2019. 32 Vanessa Wilson Appointed Chief to Lead Aurora’s Beleaguered Police Department, Sentinel, Aug. 4, 2020. 33 City of Aurora, Police. 34 Aurora Police Department, 2020 4th Quarter Affirmative Action Report, Jan. 11, 2021. 35 City of Aurora, Executive Leadership Team. 36 City of Aurora, Organizational Chart. 31 10 Two unions compete for membership of Aurora Police officers—the Aurora Police Association and the Fraternal Order of Police, Lodge 49. In 2019, FOP Lodge 49 was certified by majority vote to serve as the collective-bargaining unit for police officers beginning in 2021, replacing the Aurora Police Association.37 The contract between Aurora and the FOP Lodge 49 runs through the end of 2022.38 3.2.5. Training—Changes at the Police Academy Aurora Police operates its own police academy that, as of 2021, trains about twenty cadets in each of five sessions per year. All officers hired into Aurora Police—both entry-level and lateral—must go through programs at the academy. Before 2021, the entry-level academies occurred twice a year, lasted for 26 weeks, and trained about 40 recruits per academy. The Training Section changed the frequency of the academies due to concern about losing qualified applicants to other police agencies with more frequent academies and more flexible hiring schedules. The entry-level academies train recruit officers on the law, firearm usage, law enforcement driving, custodial arrest techniques, self-defense, and other police tactics.39 Police leadership described major changes being made to the culture and curriculum of the training academy as of mid-to-late 2020. Because our investigation began in late 2020, we did not review training curriculums or content that the training academy used before 2020. Leadership told us that it has tried to make the academy less militaristic than it has historically been—by, for example, stopping the use of military terminology, no longer assigning officers “battle buddies,” and ending the practice of yelling at recruits and using public embarrassment as punishment for mistakes. Leadership also reported that the academy is seeking to shift to a guardian mindset focused on officers’ roles as guardians of the community.40 Leadership has prioritized improving the diversity of academy staff, so the officers working there are more reflective of the demographics of the recruit class. The Training Section has also changed the academy’s curriculum. The academy has increased community involvement by, for example, bringing in Black members of the community to speak to recruits about their experiences with police. Asian academy instructors spoke to recruits about the increase in anti-Asian violence in the wake of COVID-19. And the academy has incorporated critical incidents involving police in other jurisdictions into its training—for example, training recruits on distinguishing between drawing their Taser versus firearm after the police killing of Daunte Wright in Minneapolis.41 Training personnel shared additional plans to reshape the academy— continuing to incorporate more community groups into training, such as members of the LGBTQ+ community; working with a local acting group for more realistic in-person Aurora FOP Lodge 49, Who We Are. Agreement Between the City of Aurora and Fraternal Order of Police, Lodge # 49, January 1, 2021 through December 31, 2022. 39 City of Aurora, Academy Life. 40 For a discussion of the different approaches to policing, see Sue Rahr & Stephen K. Rice, From Warriors to Guardians: Recommitting American Police Culture to Democratic Ideals, New Perspectives in Policing, April 2015; Samuel E. Walker & Carol A. Archbold, The New World of Police Accountability 17-19 (3d ed. 2020). 41 What to Know About the Death of Daunte Wright, The New York Times, Apr. 23, 2021. 37 38 11 scenario-based training; using a new simulator for scenario-based training that includes situations in which de-escalation rather than using force is the right outcome; and implementing joint scenario-based training with both the police and fire academies. 3.2.6. Training—Changes to In-Service Training Aurora Police’s in-service training provides mandatory recurring training for patrol officers and other police personnel after graduating from the academy. It “focuses on those skills pertinent to the performance of functions crucial to the delivery of law enforcement services to the community.”42 It includes training on uses of force, arrest control, firearms, legal updates, anti-bias, and other topics. As with the academy, we did not review in-service training used prior to 2020. Although we heard from some officers that the quality of training offered by Aurora surpasses that of training offered at other Colorado police departments, we also heard that Aurora Police has historically neglected meaningful in-service training, with one sergeant describing its treatment as a “check-the-box” exercise. On ride-alongs and during interviews, officers expressed a desire for fewer online trainings and more scenario- and skills-based trainings on community interactions, night-time decision-making, firearm usage, and other tactics. Training personnel similarly described a desire for better in-person scenario-based training and highlighted several new areas of focus for in-service training: de-escalation, use-of-force report writing, and mental health issues in both civilians and officers. By the end of 2021, officers will have participated in three in-service training blocks.43 The first is online-only and includes training on anti-bias, community policing, de-escalation, mental illness, and excited delirium. The second training will be in-person, covering topics including firearm usage, emergency vehicle operations, and arrest control tactics. The third block of training is new for 2021 and will require patrol officers to train in-person on deescalation through verbal judo, use-of-force report writing, and high-risk stops, among other topics. We did not have an opportunity to observe the new block of training, which is scheduled for the fall of 2021. The Training Section shared its plans to implement some of the same changes planned for the academy in that block, including having community members participate and using actors for more realistic scenario-based training. Training personnel also plan to use incidents captured by body-worn cameras of Aurora Police officers during in-service training. 3.2.7. Accountability Mechanisms At least four potential accountability mechanisms govern police departments—internal discipline (including termination), lawsuits, community feedback, and City Council oversight of the City Manager and Chief of Police. Discipline and termination are governed by the Internal Affairs Bureau with supervision by the Chief of Police and 42 43 Aurora Police Department Directive 07.02 (revised Aug. 20, 2020). 2021 APD In-Service Schedule. 12 review before the Aurora Civil Service Commission.44 Internal Affairs handles the process for all allegations of misconduct, including community complaints, internal complaints, and potential policy violations. Resident complaints raised over 1,500 allegations against sworn officers from 2018 to 2020, and 951 of those complaints were sustained, meaning Aurora Police found that officers violated policies. Only 53 allegations were about excessive use of force. Aurora Police representatives explained that a single complaint may include more than one allegation against more than one officer. But Internal Affairs tracks the number of cases it investigates (not allegations) and the number of officers with complaints sustained against them or not. Because the number of allegations does not correspond to either the number of cases referred for investigation or the number of officers investigated, it was impossible for our team to determine how many officers were the subject of complaints. This accounting disconnect obscures the complaint process and makes it difficult for the community to know if complaints are a useful accountability tool. However, when a preliminary investigation reveals a serious complaint that if sustained could result in more than 40 hours suspension without pay, it should be referred for further Internal Affairs investigation. Those cases generally resulted in sustained complaints. In the 122 serious cases investigated by Internal Affairs between 2018 and 2020, it sustained allegations against 122 officers and did not sustain allegations against 35 officers. Supervisors in each district investigated another 112 less-serious cases. Our team was not provided the outcomes for these district-level investigations. After the Internal Affairs process, the Chief of Police may impose discipline, subject to review by the Civil Service Commission.45 An officer may appeal any disciplinary order other than an oral or written reprimand.46 Our team could not determine how many of the 122 officers received discipline as a result of the sustained complaints, or how many of the 112 less-serious cases led to officer discipline. While Internal Affairs posts generic descriptions of investigation outcomes on bulletin boards in each police station, the overall lack of transparency makes it difficult to know how Aurora Police uses resident complaints to spot recurring problem behaviors, address policy gaps, or discipline officers for misconduct. Lawsuits likewise have not created accountability for Aurora Police officers. Aurora has recently settled lawsuits involving excessive use of force in response to a noise complaint47 and a K-9 officer ordering a police dog to bite a restrained woman.48 City representatives reported that in recent years, settlements included requirements that Aurora Police make Aurora Police Department Directive 10.02. (revised May 19, 2020). Aurora Police Department Directive 10.02.15 (revised May 19, 2020). 46 Aurora Municipal Code, § 3-16(8). 47 Aurora Police to Pay Man $285,000 Settlement for Ordering Him out of Home, Slamming Him on Ground, 9 News Denver, Dec. 10, 2020. 48 Aurora Pays $80,000 Settlement After Police Ordered Dog to Bite Woman Being Restrained by Officers, The Denver Post, Feb. 24, 2021. 44 45 13 certain changes, such as adopting a new SWAT policy, adding trainings on vehicle stops, and responding to persons with epilepsy. However, we saw no evidence of a coordinated approach by the city to provide direct feedback to officers and their supervisors whose conduct prompted lawsuits. Allegations from the lawsuits do not launch a formal process to help Aurora Police identify patterns of rights violations or other opportunities to improve. Complaints from the lawsuits are not included in an early warning system database that could notify a supervisor and trigger interventions. Instead, the City Attorney’s office states that they maintain a dialogue with each city department to discuss allegations and their implications. Those informal conversations, while necessary, do not systematically examine the costs imposed by officer conduct and lead to a less comprehensive review of any patterns that a police department’s early warning system is designed to catch. Aurora reports paying over $7 million in settlements for claims of excessive force and constitutional violations by Aurora Police over the ten-year period from 2008 to 2018. Of course, police departments settle lawsuits for many reasons, and a settlement does not necessarily mean that a legal violation occurred. But generally, significant settlements occur in cases which pose risk that the city, or officers that it indemnifies, will be found to have violated the law. The payouts by Aurora to settle claims of excessive force and constitutional violations do not directly impact Aurora Police’s budget or create other direct consequences for the organization. Rather, the city handles all liability in an aggregate Risk Management Fund that provides self-insurance for most claims and purchases excess insurance coverage for large liabilities. This Fund is a separate line item in the city’s budget. It aggregates all liabilities facing the city, whether they arise from police conduct, vehicle accidents, or other municipal operations in various other city departments.49 Our discussions with Aurora personnel reinforced the lack of a direct connection between payments made to resolve cases against the police or fire department and any consequences to the police or fire departments. Without such direct connection, lawsuits provide inadequate incentives to change police behavior.50 Senate Bill 217 waives qualified immunity in some specific circumstances and adds a provision which states that officers themselves may face personal liability up to $25,000 in lawsuits for police misconduct when their employer finds that “the officer did not act upon a good faith and reasonable belief that the action was lawful.”51 This provision has not City of Aurora, 2021 Operating and Capital Improvement Budget. Samuel E. Walker & Carol A. Archbold, The New World of Police Accountability 47-49 (3d ed. 2020) (“The basic flaw in the strategy of seeking police reform through civil suits is the assumption that public officials will react in a rational and coordinated manner. That is, they will see a problem (the costs of police misconduct cases) and will take the necessary steps to reduce those costs. The evidence, however, indicates that public officials respond indifferently and with no coordinated plan. In practice, one agency of local government (the police) perpetuates the harm, another agency defends it in court (the law department), and a third agency writes the check (the treasurer).”); Joanna C. Schwartz, Myths and Mechanics of Deterrence: The Role of Lawsuits in Law Enforcement Decisionmaking, 57 UCLA L. REV. 1023, 1067 (2010) (finding “most law enforcement officials know little about lawsuits alleging misconduct by their officers”). 51 § 13-21-131(1) & (4), C.R.S. 49 50 14 been in effect long enough for us to determine how it will play out in practice. As we discuss below, many officers are concerned about the uncertainty created by this bill. Community feedback provides another avenue to ensure police accountability. As discussed in later sections, Aurora has several advisory teams comprised of community members who give feedback on police policies, disciplinary decisions, and recent critical events. These feedback mechanisms make community perspectives available to the police department, but advisory teams do not have any decision-making or policy authority. Finally, political accountability for appointment of the City Manager and the Chief of Police creates some potential feedback for Aurora Police. But this feedback focuses on consequences for past decisions, rather than specific input on forward-looking decisions. We find that these accountability mechanisms do not work to provide sufficient feedback for Aurora Police on the behavior of its officers. The feedback that does occur is too cumbersome, ad hoc, and diffuse to provide meaningful input to the organization. To follow the law, Aurora Police must improve its accountability mechanisms. 3.3. Aurora Fire Rescue Aurora Fire Rescue is led by Chief Fernando Gray, who joined Aurora Fire in June 2017.52 It employs about 440 full-time personnel and has seventeen fire stations located throughout Aurora.53 Within four years of starting at Aurora Fire, all firefighters receive training to become registered Emergency Medical Technician Paramedics.54 Interviews with fire personnel revealed that Aurora Fire has not seen the same turnover, morale, and community relations issues as Aurora Police. But many raised similar concerns over the effect of recent legislation, potential liability, and changes to Aurora Police policies that impact their cooperation. Aurora Fire has also faced recent controversy, including the termination of the second-in-command Deputy Chief following an investigation into his misuse of time off, retaliation against a subordinate, and use of racially derogatory language.55 3.3.1. Falck Rocky Mountain Falck Rocky Mountain is the sole emergency ambulance provider for Aurora. Falck is a Danish-owned private company, whose business unit Falck Rocky Mountain has contracted with Aurora Fire since 2015.56 Through this partnership, Falck responds to over 30,000 medical services requests each year in Aurora. On most calls for medical service, at least one Falck ambulance responds alongside Aurora Fire. Aurora Fire typically maintains Aurora Names Dallas Firefighter Fernando Gray New City Fire Chief, Sentinel, Apr. 11, 2017. City of Aurora, Fire Stations. 54 City of Aurora, Aurora Fire Rescue Recruiting FAQs. 55 Aurora Fires Deputy Fire Chief Stephen McInerny for Policy Violations, CBS Denver, Mar. 5, 2021. 56 Falck Rocky Mountain. 52 53 15 medical control over major calls and Aurora Fire personnel provide Advanced Life Support services, often using Falck ambulances and medical supplies from Falck. 3.3.2. Ketamine Use by Aurora Fire Starting in 2018, Aurora Fire requested and received a waiver from the Colorado Department of Public Health and Environment to permit paramedics to administer ketamine outside of a hospital setting. Without such a waiver, paramedics may not administer ketamine in Colorado. The waiver required, among other things, ketaminespecific training, specific protocols for administering ketamine, and detailed reporting to the state shortly after every use. For the entire time of this waiver, Aurora Fire paramedics administered ketamine under the supervision of its medical director, Dr. Eric Hill. In 2020, Aurora City Council suspended the use of ketamine by its paramedics and the waiver expired in 2021. 3.4. Aurora City Council Aurora is a council-manager form of government, run by City Council and the City Manager. City Council includes Aurora’s mayor, one member from each of the city’s six wards, and four at-large members. City Council hires the City Manager, who oversees the day-to-day operations of the city government, including Aurora Fire and Aurora Police.57 3.4.1. Recent Changes Aurora City Council made several policy changes in response to Elijah McClain’s death and the public reaction. First, City Council unanimously approved a moratorium on the use of ketamine by Aurora Fire and Falck Rocky Mountain.58 The moratorium became effective on September 15, 2020, and required Aurora Fire to remove ketamine from all service units. It expired on March 24, 2021, but Aurora Fire has stated that it has no immediate plans to resume ketamine’s usage.59 Second, City Council funded a pilot program that would permit mental health workers to respond to some 911 calls instead of police officers.60 The program—modeled after the Support Team Assisted Response, or STAR, program from Denver, and the Crisis Assistance Helping Out On The Streets, or CAHOOTS, program from Eugene, Oregon— will feature an EMT or paramedic and mental health clinician responding to mental healthrelated calls instead of police. The city started the program in September of 2021.61 Third, City Council launched an independent investigation of Mr. McClain’s death, led by Jonathan Smith, executive director of the Washington Lawyers’ Committee for Civil Rights City of Aurora, Your Aurora Government 2021. City of Aurora, Ketamine Moratorium. 59 Aurora Has No Immediate Plans to Re-Implement Ketamine After Elijah McClain Review, KDVR News, Mar. 19, 2021. 60 Aurora Funds New Program to Replace Cops With Clinicians to Some 911 Calls, Sentinel, Sept. 21, 2020. 61 Denver’s STAR Program Will Be Replicated in Aurora Starting This Week, Denverite, Sept. 6, 2021. 57 58 16 and Urban Affairs.62 City Council set the scope of the investigation, which focused on the timeline of events leading to Mr. McClain’s death, the actions of Aurora Police and Fire during the encounter, and the policies and procedures associated with use of force, ketamine use, and administrative incident reviews, among other issues.63 The investigation team released its findings on February 22, 2021.64 Fourth, City Council established a Community Police Task Force composed of twelve community members and one ex officio police member.65 City Council asked the task force to review Aurora Police policies and procedures and to provide recommendations on improving police interactions with community members. The Community Police Task Force presented its recommendations to City Council on April 5, 2021.66 Its primary recommendations included: • The creation of an independent citizen’s oversight office with investigatory and subpoena powers, responsible for: o reviewing resident complaints, critical incidents, and Aurora Police policies and practices, o reviewing all Aurora Police disciplinary actions, with the power to overturn or refer such cases, o contemporaneously receiving all body-worn camera footage, and o working with the Civil Service Commission on hiring and discipline; • An increase in funding for mental health services, including a co-responder program pairing officers with mental health professionals, mandatory mental health training, and mental health screening for officers; • The creation of various programs to protect the anonymity of officers who report misconduct of other officers, provide for community member “safe to tell” tip lines, diversify the police academy classes to reflect Aurora’s demographics, and monitor the social media activity of police officers; and • An amendment to the Charter giving the independent citizen’s oversight office authority that is parallel to that of the Civil Service Commission. City Council has begun to implement some of these recommendations. Aurora City Council Unanimously Approves Scope of Elijah McClain Investigation, Sentinel, July 21, 2020. Independent Report Released in Elijah McClain Case, City of Aurora, Mar. 16, 2021. 64 City of Aurora, Investigation Report and Recommendations, Feb. 22, 2021. 65 City of Aurora, Community Police Task Force. 66 City of Aurora, Agenda and Meeting Minutes from April 5, 2021, 72-83. 62 63 17 3.4.2. Report and Recommendations from 21CP Solutions Chief Wilson and City Manager Jim Twombly hired 21CP Solutions, a group of experts focusing on the areas of civil rights and public safety, to review Aurora Police’s policies, procedures, and operations.67 21CP Solutions presented its report and recommendations to City Council on August 16, 2021. The 21CP Solutions report considered how Aurora Police could improve their policies and organization to align with recognized best practices in procedural justice and policing.68 The report identified six areas where Aurora Police could improve their operations: policing practices, community engagement, organizational structure, officer development, accountability, and data management.69 Most of the report focused on Aurora Police’s policies and practices, where 21CP Solutions recommended: • Specific changes to Aurora Police’s use-of-force, arrest, crisis intervention, and bias-free policing policies; • Commitment to a dedicated mental health crisis response model; • Guidance on how officers write use-of-force narratives and general offense reports for all stops and standardized reporting of required data for all encounters; • Improved guidance on use-of-force investigation, review, and adjudication; • Periodic analysis of data demonstrating trends in arrests, stops, uses of force, and biased-based policing; and • Improved transparency of resident complaints. The report also suggested strategies to address staffing shortages and recommended expanding the roles and collaboration between the Community Policing Task Force and Police Area Representatives. Other suggestions included: • Improving communication between command staff and officers; • Recruiting and hiring more diverse officers to better reflect the Aurora community; • Using scenario-based training on stops, searches, arrests, de-escalation strategies, critical decision-making, procedural justice, and bias-free policing; • Revising the Field Training Officer and Peer Support programs; City Undertakes Comprehensive Review of Police, City of Aurora, Aug. 25, 2020. 21CP Solutions, Recommendations for the Aurora Police Department, August 2021. 69 We have summarized these areas of focus. The 21CP report titles each area: Critical Operations; Community Engagement and Participation; Organization and Command Structure; Selection, Supervision, and Support of Personnel; Accountability; and Equipment, Technology, and Data Systems. 67 68 18 • Enhancing the Early Intervention System to improve officer development; • Overhauling the complaint, investigation, and disciplinary process for officers; • Reviewing the Civil Service Commission’s equity in hiring and discipline decisions; and • Creating data analysis and management technology solutions. 21CP Solutions released its report in August 2021, and it covers many areas where Aurora Police has already committed to making adjustments. 3.5. Aurora Civil Service Commission Aurora’s Charter creates a Civil Service Commission and gives it significant oversight of police and fire operations. Five community members selected by City Council serve on the Commission.70 The members serve three-year overlapping terms and are limited to three consecutive terms.71 Aurora gives the Civil Service Commission extraordinary control over hiring and discipline of police and fire employees. The Charter grants the Commission sole responsibility for the examination and certification of all entry-level applicants to the police and fire departments. As a result, Aurora Police and Fire cannot make the final decisions on entrylevel hiring.72 The Commission also sets the service requirements and examination procedures for internal police and fire promotions.73 Finally, the Commission can review discipline imposed by the police and fire chiefs. Any disciplinary order other than an oral or written reprimand may be appealed to the Commission, and the Commission can affirm, reverse, or modify the disciplinary order by majority vote.74 This oversight over disciplinary decisions has been controversial—for example, the Commission overturned a decision by former Chief of Police Nick Metz to terminate a lieutenant who referred to a group of Black residents as “Alabama porch monkeys.”75 We heard concerns from community members and police officers about the scope of the Commission’s power, particularly as it relates to hiring and discipline of police officers. A recent analysis of the 2018 to 2020 police academies found that the Commission’s hiring process leads to disparate outcomes for applicants of color—of 2,809 white applicants who met minimum qualifications, the Commission hired 119, for a hiring rate of 4.24%.76 In contrast, the Commission approved the hiring of only 5 of 454 qualified Black City of Aurora, Civil Service Commission. Aurora Charter, § 3.17. 72 Aurora Charter, §§ 3.16(10), 3.17(3). 73 Aurora Charter, §§ 3.16(6), 3.17(3). 74 Aurora Charter, § 3.16(8)(e)-(h). 75 Aurora Officer Fired for Racial Slur Reinstated by Civil Service Commission, KDVR News, July 10, 2018. 76 City of Aurora, Aurora Civil Service Commission 2020 Overview and Recent Entry-Level Hiring Summary, p. 57-60. 70 71 19 applicants (1.1%), 32 of 1,073 qualified Hispanic or Latino applicants (3%), and 5 of 139 qualified Asian applicants (3.6%). We also heard concerns about the Commission’s minimal transparency—meeting recordings or transcripts are unavailable online, and community members have little insight into hiring and disciplinary decisions. As for disciplinary decisions, the Commission does not publicly release any documents or information except for the formal appeal filed by the member and the Commission’s ultimate disciplinary findings. And the Commission only produces these documents in response to requests under the Colorado Open Records Act—it does not make this information easily accessible online or otherwise. The federal government investigated the entry-level hiring practices of the Civil Service Commission in 2009. The government began the investigation because of information suggesting that Aurora Police and Fire had far lower percentages of Black and Hispanic personnel than would be expected based on comparable agencies. The investigation covered all phases of the hiring process but at first focused on Aurora’s use of written examinations to screen Police and Fire applicants. In 2012, the government expanded the investigation to determine whether Aurora’s use of a physical abilities test during the entrylevel hiring process discriminated against female candidates for Aurora Police. This investigation closed in 2013 and did not result in a formal consent decree. Instead, Aurora voluntarily made changes to its hiring process in response to the government’s concerns, including awarding “preference points” to an applicant’s overall ranking for hire if they were proficient in a second language; replacing Police and Fire’s written examination and interview process with a more objective video-based examination; moving the Police and Fire physical abilities test from the beginning to the end of the testing process; and giving the Civil Service Commission responsibility for conducting Police and Fire entry-level background investigations. City Council recently appointed several females and people of color to serve on the Civil Service Commission. Community members expressed optimism about the Commission’s efforts to diversify its membership, but shared concern that those efforts would be insufficient to create actual change. 20 4. Successes of Aurora Police and Fire While we find that Aurora Police and Fire have a pattern and practice of violating the law, as explained in further detail below, we did observe some aspects of Aurora Police and Fire that embraced best practices for public safety agencies. Though some of these practices must improve as a result of our findings, Aurora’s decision to adopt many of these practices into their departments represents noteworthy progress towards improving their agencies. 4.1. Early Adoption and Extensive Use of Body-Worn Cameras Aurora Police began deploying body-worn cameras in 2016.77 Senate Bill 217 requires all law enforcement agencies throughout Colorado to use body-worn cameras by 2023.78 Aurora’s decision to do so several years earlier has enabled better outside review of their officers’ conduct and more accurate review of use-of-force incidents. Aurora recently provided officers with new camera systems that have better video quality. Although not all features are operational on the new system, the cameras will activate automatically when an officer leaves their vehicle or unholsters their weapon, and doing so will activate the cameras of any nearby officers. Of course, opportunities to improve how the body-worn camera program works remain, including improved training to ensure compliance with policies. For example, Aurora should ensure that officers understand the new camera system’s automation and manual overrides and always mount the cameras on their outermost garment. As one example, officers activated the old camera system by touching the front of it. This is now how officers turn off the new cameras. This means that officers in the habit of touching their camera to turn it on are now inadvertently turning off the automatic system. In addition, we frequently saw cameras dislodged during uses of force, which officers have said is still a problem with the new cameras. 4.2. The Force Review Board to Review Use-of-Force Incidents Aurora established a Force Review Board in 2016 to review significant uses of force by its officers. This Board has a mix of department members, including representatives from the Training Section, the Professional Standards Section, and a sergeant and officer.79 Many, and sometimes all, of the members of the Force Review Board are outside the chain of command of officers involved in a particular incident. The Board typically meets once a week and reviews the reports and body-worn camera videos of significant uses of force. We observed the Force Review Board meetings for over nine months during this investigation. Reviewing uses of force outside the chain of command can improve consistency in how officers use force, allow for Aurora Police to learn how to reduce the need for using force in the future, and ensure that Aurora Police’s training on interacting with the community City of Aurora, Internal Audit Report—Aurora Police Department Body-Worn Camera Compliance, p 8. § 24-31-902, C.R.S. 79 Aurora Police Department Directive 05.04.7 (revised Aug. 12, 2021). 77 78 21 and using force reflects what the Force Review Board learns about the actual experiences of Aurora Police.80 Under Aurora Policy, the Force Review Board must review “Use of Force Reports for compliance with Standard Operating Procedures, Department Directives and applicable law.”81 Use-of-force reports are created by sergeants and others in the chain of command above the officers who used force. While we commend Aurora for the establishment of the Force Review Board, as we discuss below, its operation falls far short of its potential, particularly in its focus simply on what can be justified under the law, rather than what is legal and appropriate in the circumstances. 4.3. The Creation of a Force Investigation Unit to Allow Affirmative Investigations of Uses of Force Chief Wilson recently created the Force Investigation Unit to support the Force Review Board. Until the creation of the Force Investigation Unit, the Force Review Board could only review information from the body-worn cameras or the reports created by those in the chain of command. This limitation prevented full internal investigations of uses of force, as the Force Review Board was limited to only those reports created for it by those supervising the officer. We had the opportunity to review the operation of the Force Investigation Unit for several months. Based on our limited review, the Force Investigation Unit brings some level of objectivity to the review, but we still see similar patterns in the review of force—a focus on justifying the force used rather than evaluating whether it was both legal and appropriate. And the operation of the Force Investigation Unit does not yet meet its potential. Members of the unit said that it was formed in part to help identify trends and patterns in uses of force by officers, but no structured method of doing this exists. However, the Force Investigation Unit has begun to standardize the presentation and review of uses of force, lessening the prior inconsistency in use-of-force review that resulted from sergeants and others in the chain of command evaluating lower-level uses of force. 4.4. Change in Training Philosophy and Culture at Academy In 2020, Chief Wilson appointed Lieutenant Brett Parvin to lead the Training Academy with a mandate to change the philosophy and approach to training at the Academy. As discussed above, these changes—particularly for the entry-level academy programs— represent significant positive steps. The approach appears to be focused on creating a guardian culture that emphasizes protecting the community Aurora Police serves.82 In addition, the Training Academy plans to increase community involvement, including participation by community members who have encountered police uses of force. Because Samuel E. Walker & Carol A. Archbold, The New World of Police Accountability 298 (3d ed. 2020). Aurora Police Department Directive 05.04.7 (revised Aug. 12, 2021). 82 Sue Rahr & Stephen K. Rice, From Warriors to Guardians: Recommitting American Police Culture to Democratic Ideals, New Perspectives in Policing, April 2015. 80 81 22 the changes are underway, we cannot determine whether they will create the desired effect. We would expect to see meaningful changes in measures such as a decrease in excessive uses of force, community complaints, referrals to Internal Affairs, and similar measures. And Aurora Police must extend the changes from the Training Academy to In-Service Training of current officers. While we understand Aurora Police plans to extend this training, it has not fully done so. 4.5. Expanded Community Involvement Aurora Police has started several efforts to engage and receive feedback from the community. These efforts reflect Aurora Police’s desire to seek community input into its operations, needs, and discipline. • The Aurora Key Community Response Team was created in 1992. Community stakeholders representing diverse groups meet regularly to facilitate communications between the police and community. • Aurora Police created the Community Policing Advisory Team (“CPAT”) in 2017. It is a smaller group of community leaders who meet monthly and consult leadership on promotions at the commander level. CPAT members also help connect police to community events and provide input on noteworthy events. • The Aurora Community Police Task Force is new as of 2020 and members are appointed by Aurora City Council. This group makes recommendations to the police and also reviews police policies and procedures. • The Chief ’s Review Board is a blended team of law enforcement and community members that City Council appoints to review police disciplinary matters and make recommendations to Chief Wilson. To increase community input, transparency, and diversity, Chief Wilson added four community members to this Board. • The Independent Review Board is also a blended team of law enforcement and community members who review disciplinary matters and make recommendations to Chief Wilson. • The Community Relations Section was recently created by Chief Wilson and includes Aurora for Youth, Recruiting, and Community Resources. • Police Area Representatives (“PAR”) work in the three Aurora police districts and facilitate connectivity between the police and community members at major city events, school activities, and food drives. PAR officers focus on building community relationships. None of these efforts to expand community involvement give any actual decision-making authority to those outside of Aurora Police, nor do they empower community members to obtain more information beyond that provided by Aurora Police. But these efforts do create more transparency and provide more information to the community. As with other positive steps, Aurora Police can and should build on these efforts. 23 4.6. Recognition of Need for Improvement and Change by Senior Management In November 2020, Chief Wilson introduced A New Way, Aurora’s five-point plan to restore trust.83 The plan focuses on five primary concepts: operating, leadership, service, accountability, and engagement. A few key changes respond directly to concerns expressed by community members. Command staff understand the need to move from a “what can we do” perspective on arrests and uses of force to a “what should we do” perspective. Aurora Police added Accountability to its core values of Duty, Honor, and Integrity. This addition of Accountability is a needed change. Aurora Police leadership has prioritized improving diversity and gender equity at all levels of the agency, starting with internal promotions. To implement this goal, officers now must participate in diversity, equity, and anti-bias training, and must read the book Difference Matters: Communicating Social Identity by Brenda Allen about race, age, and gender identity before any promotion. Increasing accountability and transparency are critical to improving community relations. Chief Wilson created a Force Investigation Unit that will interview people who witnessed use-of-force events, to bring more perspective to the review. As discussed, this new investigation unit reports to the Force Review Board. Additionally, a Police Auditor will review Standard Operating Procedures. Internal Affairs synopses will soon be publicly available to educate the community about discipline and improve operations. Aurora Police states that it intends to work with the Crime and Justice Institute, an organization that specializes in developing effective policies for law enforcement organizations, to perform a comprehensive redraft of Aurora Police’s policies relating to the use of force. As for community engagement, Chief Wilson and the command staff have committed to continuing and improving the community task forces that are already operational. They have also conveyed that they are focused on the Explorer Program and the Chief ’s Youth Advisory Team to engage younger community members. 83 City of Aurora, A New Way: Our Plan to Restore Trust. 24 5. The Effect of This Investigation We are aware of several concerns surrounding the new legislation on police accountability and this investigation. In other cities where similar investigations took place, officers expressed concern that restrictions and lack of training on new use-of-force policies could make them hesitate at a crucial moment, putting them, other officers, and the public at risk.84 Police union representatives echoed those concerns.85 This year, two Colorado officers were killed by armed assailants, highlighting the significant risks that always accompany serving as a police officer.86 Public officials in Colorado have said they worry about how public scrutiny is affecting officer morale and public safety.87 During ride-alongs, officers told us that they feared the threat of lawsuits or losing their jobs if they made even the smallest mistakes under the recent legislation, which provides for potential individual liability. That concern, coupled with what officers describe as conflicting directions about how and whether to contact community members, has some officers reluctant to take on roles where high levels of community interaction occur. As a result, some special assignments that focus on high levels of contact designed to deter crime are often understaffed. The changes we require in this report are designed to improve public safety, officer safety, and confidence in Aurora Police and Fire. When implemented properly, appropriate accountability measures do not create higher crime rates or increase risks to officer safety. Indeed, as the discussion below shows, some police departments that worked to support officers with comprehensive training, clear guidance, fair discipline policies, and positive, engaged supervision saw crime rates and resident complaints fall dramatically while officer safety improved. 5.1. Increased Scrutiny of Police Does Not Increase Crime Rates Investigations of police departments and their resulting reforms are associated with decreases in the overall violent crime rate.88 In several major cities, for example, crime rates declined substantially following consent decrees.89 “Both property crimes (down 53%) and violent crimes (down 48%) decreased in Los Angeles,” and this drop in crime was not mirrored in several nearby communities that did not attempt reforms. Cincinnati’s violent Stephen Rushin & Griffin Edwards, De-Policing, 102 CORNELL L. REV. 721, 725-26 (2017); see also Lawrence Rosenthal, Good and Bad Ways to Address Police Violence, 48 URB. LAW. 675, 737 (2016). 85 The LAPD is Officially Suggesting a Few Things Officers Can Try to Do Before Pulling the Trigger, Los Angeles Times, Mar. 15, 2016. 86 Colorado Officer Killed in Boulder Grocery Store Remembered as Hero Who Put Others First, NBC News, Mar. 30, 2021; Police Chief Says Slain Colorado Officer Was ‘Targeted’ in Monday Shooting, CNN, June 22, 2021. 87 ‘We Need to Take Our Community Back’: Aurora Police Chief Addresses Rising Gun Violence, KDVR News, June 22, 2021; Aurora Overtakes Colorado Springs as 2nd Most Violent City in Colorado, KDVR News, Mar. 17, 2021. 88 Kenny Lo & Sarah Figgatt, Violent Crime Rates Declined in 10 Jurisdictions Following Comprehensive Police Reform, Center for American Progress, Nov. 16, 2020. 89 Paul Butler, The System Is Working the Way It Is Supposed to: The Limits of Criminal Justice Reform, 104 GEO. L.J. 1419, 1459 (2016). 84 25 crime rate also dropped 56%.90 Following Pittsburgh’s consent decree, crime rates, including homicides, decreased consistent with the national trends.91 These cities show that policing reforms are associated with improved community safety. True, some cities experienced a small increase in property crimes and “street crimes like burglary, motor vehicle theft, and robbery—i.e., criminal activity that is likely sensitive to situational deterrents like aggressive street policing.”92 But in those cities, the “apparent uptick in crime was concentrated in the years immediately after the initiation of external regulation and diminished into statistical insignificance over time.”93 After a period of adjustment, these departments were just as effective if not better at deterring both violent crime and property crime. Researchers suggest that this uptick in street crime may stem from reform “that  is virtually all stick and no carrot,” and may be avoided by recognizing and rewarding officers who engage in proactive policing of high-crime areas while adhering to reforms.94 Police departments that voluntarily implemented more restrictive use-of-force policies also saw lower crime rates.95 Cities that placed limits on officers’ use of force, without addressing training or accountability, maintained similar levels of violent crime and property crime as cities that did not change their policies.96 But the more comprehensive the reforms, the more crime rates dropped. For example, in New Jersey, Camden County’s new training program and restrictive use-of-force policies accompanied a 42% reduction in the overall violent crime rate.97 Homicides also decreased 63%, and robberies by 60%.98 In short, public safety typically increased after police departments reformed how officers used force and trained officers on effective alternatives to force. 5.2. Policy Restrictions on Uses of Force Are Associated With Improved Officer Safety Several studies show that policy reforms limiting an officer’s use of force are not associated with increased risk to officers—instead, quite the opposite. In cities that placed greater limits on when officers could use force, officers were less likely to be assaulted or killed in the line of duty.99 A recent study compared the policies large-city police departments had Id. Robert C. Davis, et al., Turning Necessity into Virtue: Pittsburgh’s Experience with a Federal Consent Decree, Vera Institute of Justice, 54-56 (2002). 92 Stephen Rushin & Griffin Edwards, De-Policing, 102 CORNELL L. REV. 721, 765-66 (2017). 93 Id. at 759. 94 Lawrence Rosenthal, Good and Bad Ways to Address Police Violence, 48 URB. LAW. 675, 724 (2016). 95 National Institute of Justice, How Police Supervisory Styles Influence Patrol Officer Behavior (2003). 96 Samuel Sinyangwe, Examining the Role of Use of Force Policies in Ending Police Violence, Fig. 10, The Police Use of Force Project, Sept. 20, 2016. 97 Josiah Bates & Karl Vick, America’s Policing System Is Broken. It’s Time to Radically Rethink Public Safety, Time, Aug. 6, 2020. The article notes that economic growth may also play a role in decreased crime rates. 98 Id. 99 Samuel Sinyangwe, Examining the Role of Use of Force Policies in Ending Police Violence, Fig. 8, The Police Use of Force Project, Sept. 20, 2016. 90 91 26 implemented to limit use of force and then looked at corresponding FBI data on officer assaults and deaths for each city. 100 In cities with more policy limits, officers experienced fewer assaults—on average, one-third fewer assaults.101 Yet in cities that did not reform their use-of-force policies, the risk of an officer being killed was nearly three times higher.102 The research did not confirm whether officer safety was increased because of the new policies or because of the training involved with implementing them. But it did confirm that reforms to use-of-force policies generally do not place officers at higher risk of injury. 5.3. Policy Restrictions on Uses of Force Are Associated With Increased Safety for Justice-Involved Residents Those same policy limits also correlate with fewer in-custody deaths without increased crime rates.103 Cities that adopted four or more identified use-of-force policy reforms saw 54% fewer deaths in police custody than departments with only one or two policies in place.104 Cities that had eight key policies in place saw 72% fewer deaths.105 The most significant decreases in resident deaths were found in departments whose use-offorce policies included de-escalation tactics, use-of-force continuums, and the requirement for officers to exhaust all other reasonable means prior to shooting, which required that officers undergo training to learn new critical decision-making tactics. Those departments also added comprehensive reporting requirements on use-of-force incidents, which required leadership to enforce the new policies and develop a culture of accountability.106 In short, when departments adopted a comprehensive approach to reform, including policy revisions, training, transparency, and accountability, fewer residents were killed by police, fewer officers were injured, and the crime rate went down. 5.4. Policy Restrictions on Police Are Not Associated With Less Engaged Policing Although officers sometimes report that they fear that reforms in departmental policies will make them more hesitant to intervene, the evidence shows that they remain highly engaged after implementing these reforms. One study found that, in the wake of two Id., at 2-3, Figs. 8-9. The study examined eight specific policies and their impact on both public safety and officer safety. The policies required officers to (1) use de-escalation, (2) refer to a use-of-force continuum, (3) exhaust all other means before shooting, (4) give a warning before shooting, and (5) intervene when other officers violated policy; restricted officers from (6) firing at moving vehicles or (7) using chokeholds or strangleholds; and (8) required comprehensive departmental reporting. 101 Id. at Fig. 9. 102 Id. The data used by the authors was averaged from multiple departments, and thankfully, not every department loses an officer in the line of duty. Therefore, we have chosen to refer to this data as representing the “risk” to officers. 103 Police Use of Force Project: How police use of force policies can help to end police violence. 104 Samuel Sinyangwe, Examining the Role of Use of Force Policies in Ending Police Violence, The Police Use of Force Project, Sept. 20, 2016. 105 Id. 106 Id. 100 27 consent decrees, “sick time—one indicator of low officer morale—had actually declined.”107 Although officers “believed that discipline was frequently meted out,” the number of disciplinary actions against officers following the decree went down.108 After the consent decree in Los Angeles, officers reported that they were “more guarded in their interactions,” but arrest rates actually went up and “the rate at which arrests resulted in felony charges increased.”109 The researchers credited the increased felony clearance rate to a policy requiring that officers document their reasonable suspicion for a stop, “reducing the number of ‘weak’ cases.”110 So while officers protested the increased scrutiny, they “adjusted to the new rules on stops, arrests, and use of force and actually increased their enforcement efforts.”111 5.5. Successful Reform Efforts Are Comprehensive Finally, how much a city benefits from police reforms—through fewer resident injuries from police use of force, fewer officer injuries, and lower crime rates—depends on how comprehensive those reforms are. The most positive results come from policy changes accompanied by officer training, support from senior officers and field supervisors, and transparency of accountability measures.112 For example, training officers on how to interact with the public with dignity and respect, giving residents a voice in encounters, appearing neutral in decision-making, and communicating trustworthy motives—commonly called “procedural justice”—reduced the number of use-of-force complaints against Chicago’s officers and the number of times officers reported using force.113 “[W]hen police interact with citizens in a procedurally fair manner, such as speaking calmly and explaining the reasons for their actions to the Robert C. Davis, Nicole J. Henderson & Christopher W. Ortiz, Can Federal Intervention Bring Lasting Improvement in Local Policing?: The Pittsburgh Consent Decree, Vera Institute of Justice, 25-26 (2005). The report noted that officer morale was already low well before the pattern and practice investigation and subsequent consent decrees, but officers associated their low morale with the scrutiny, not with the police department’s culture. See also, Christopher J. Marier & Lorie A. Fridell, Demonstrations, Demoralization, and De-policing, 19 CRIMINOLOGY & PUB. POL’Y 693, 714 (2020) (“High officer cynicism is significantly associated with withdrawal from police work, but cynicism on the part of police is high irrespective of periods of increased public antipathy. Thus, although public demonstrations against the police do not present threats to the police institution, some entrenched elements of police culture and morale may continue to challenge public administrators.”) 108 Robert C. Davis, Nicole J. Henderson & Christopher W. Ortiz, Can Federal Intervention Bring Lasting Improvement in Local Policing?: The Pittsburgh Consent Decree, Vera Institute of Justice, 25 (2005). 109 Samuel Walker, “Not Dead Yet”: The National Police Crisis, A New Conversation About Policing, and the Prospects for Accountability Related Police Reform, 2018 U. ILL. L. REV. 1777, 1836 (2018). 110 Id. 111 Id. at 1836-37. 112 Samuel Sinyangwe, Examining the Role of Use of Force Policies in Ending Police Violence, The Police Use of Force Project, Sept. 20, 2016. 113 George Woods, et al., Procedural justice training reduces police use of force and complaints against officers, Proceedings of the National Academy of Sciences of the United States of America (May 5, 2020) (corrected July 6, 2021). 107 28 individual . . . , citizens are more likely to comply and be respectful toward police officers, thereby reducing the need for officers to resort to force.”114 On the other hand, when departments provided Crisis Intervention Training (CIT)—a well-regarded de-escalation model—the rate of officer injuries and use-of-force incidents remained unchanged.115 CIT reduced the number of arrests, mainly because officers transported subjects to hospitals rather than jail, but use-of-force incidents and levels of force remained the same.116 Most departments train only a select number of officers in CIT, so the majority of officers have not been taught the same crisis de-escalation approach. And CIT training is largely focused on communication, “but does not provide guidance on how officers should combine communications with tactics.”117 When situations are evolving, or communication is initially unsuccessful, officers may resort to defensive tactics such as the use of force. Our own observation of CIT-trained officers supports this assessment. When confronted with people in a mental health crisis, we observed CIT-trained officers either disregard their CIT tactics or fail to intervene while other officers employed traditional, defensive tactics that escalated the crisis, even when the situation was not rapidly evolving or potentially violent. The lack of demonstrated commitment to CIT tactics meant that Aurora Police’s investment in the training had little effect on resident encounters. Successfully addressing the concerns raised in our investigation will require a comprehensive approach that includes policy reform, supportive training of new and existing officers, invested supervision, and accountability to sustain constitutional policing that minimizes the need for force. Combining these strategies can create safer streets for Aurora’s residents and officers alike. Mawia Khogali, Redefining Standards of Excessive Force: Implications for Policy and Practice, 12 S. J. POL’Y & JUST. 105, 128 (2018). 115 Robert E. Worden, et al., The Impacts of Implicit Bias Awareness Training in the NYPD, IACP/UC Center for Police Research and Policy & John F. Finn Institute for Public Safety, 145 (2020). 116 Id. 117 Critical Issues in Policing Series: Guiding Principles on Use of Force, Police Executive Research Forum, 26-27 (2016). 114 29 6. Pattern and Practice Findings: Legal Background In conducting our investigation, we focused on federal and state constitutions and laws relating to four areas: (1) limitations on discriminatory policing practices; (2) the standards for stopping people on the street in Colorado, often called a Terry stop (after the U.S. Supreme Court case that originally established the constitutional requirements for such stops); (3) the law governing use of force and arrests; and (4) the requirements for administering ketamine outside of a hospital. 6.1. Limitations on Discriminatory Policing Practices Colorado and federal constitutions and laws prohibit police officers from engaging in biased policing, whether bias manifests through stops, arrests, uses of force, or other policing activities. The Equal Protection Clause of the United States Constitution provides that the government shall not “deny to any person within its jurisdiction the equal protection of the laws,” and article II, section 25 of the Colorado Constitution requires that all persons receive “equal treatment under the laws.”118 Colorado law has, since 2001, barred the use of racial and certain other forms of profiling by law enforcement officers.119 In its current form, Colorado law prevents the practice of “relying solely on race, ethnicity, gender, national origin, language, religion, sexual orientation, gender identity, age, or disability” to determine “the existence of probable cause,” “reasonable and articulable suspicion that an offense has been or is being committed,” or “the scope, substance, or duration of an investigation or law enforcement activity to which a person will be subjected.”120 Colorado’s Anti-Discrimination Act states that it is a “discriminatory practice and unlawful” for a person “to refuse, withhold from, or deny to an individual or a group, because of [protected characteristics, including race], the full and equal enjoyment of the . . . services . . . of a place of public accommodation.”121 A “place of public accommodation” includes “any place offering services, facilities, privileges, advantages, or accommodations to the public.”122 This definition logically covers police and fire departments. A recent Colorado federal court decision held that these statutory provisions “appl[y] to organizations that conduct activities in facilities . . . that are open to the public” E.g., Flowers v. Mississippi, 139 S. Ct. 2228, 2241-42 (2019); Peña-Rodriguez v. Colorado, 137 S. Ct. 855, 867 (2017); Mayo v. Nat’l Farmers Union Prop. & Cas. Co., 833 P.2d 54, 56 n.4 (Colo. 1992). 119 § 24-31-309(3), C.R.S. (effective from 2001 through 2016) (“Any peace officer certified pursuant to this part 3 shall not engage in profiling.”). 120 § 24-31-309(2)(b), (3), C.R.S. 121 § 24-34-601(2)(a), C.R.S. 122 § 24-34-601(1), C.R.S. 118 30 and that the law should be applied broadly.123 Members of Aurora Police and Fire may not discriminate against members of the community while conducting policing activities.124 In addition, several federal statutes proscribe Aurora Police and Fire from racially discriminating against members of the public. Title VI of the Civil Rights Act of 1964 and its implementing regulations prohibit entities that receive federal funds from discriminating against individuals on the basis of race, whether intentionally or through practices that have the effect of discrimination.125 Similarly, the Safe Streets Act prohibits law enforcement practices that intentionally discriminate against or that impose an unjustified disparate impact against a particular racial group.126 These laws govern the following discussion of stops, arrests, and uses of force. 6.2. Legal Requirements for Stops Under the Fourth Amendment of the U.S. Constitution, when an officer stops a resident to investigate a potential crime—known as a Terry stop—the officer must reasonably suspect that the person is involved in criminal activity.127 And Colorado imposes similar requirements for a lawful stop: the purpose of the stop must be reasonable, and the character and scope of the stop must be reasonably related to its purpose.128 Officers must be able to point to “specific and articulable facts” in support of their suspicion that the person is involved in criminal activity; simply seeing someone walk away from police officers in a high-crime area, for example, is insufficient.129 Further, the investigatory stop must be limited to “a brief detention to confirm or dispel” the officer’s suspicion that a crime was, is being, or will be committed.130 And under Colorado law, “reasonable suspicion” cannot turn on racial or other profiling.131 Creek Red Nation, LLC v. Jeffco Midget Football Ass’n, Inc., 175 F. Supp. 3d 1290, 1298 (D. Colo. 2016). In line with this interpretation, another state court, interpreting a public accommodations statute like Colorado’s, found that police departments are places of public accommodation and that police departments and their officers must comply with their state’s anti-discrimination provisions. See Ptaszynski v. Uwaneme, 853 A.2d 288, 297 (N.J. Super. Ct. App. Div. 2004) (“[W]e conclude that the Township police department—both the building and the individual officers—is a place of public accommodation. A municipal police force is nothing more than ‘an executive and enforcement function of municipal government[.]’ As a public entity, by its very nature a police force is a place of public accommodation.” (citation omitted)). And at least two more states’ courts have analyzed police conduct under their states’ public accommodations and anti-discrimination laws. See Gazette v. City of Pontiac, 536 N.W.2d 854 (Mich. Ct. App. 1995); McKinney v. City of Tukwila, 102 Wash. App. 1033 (2000). 125 42 U.S.C. § 2000d (no person shall “be excluded from participating in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving [f]ederal financial assistance” based on race); 28 C.F.R. § 42.104(b)(2) (recipients of federal funds “may not . . . utilize criteria or methods of administration which have the effect of subjecting individuals to discrimination because of their race”); see also Alexander v. Sandoval, 532 U.S. 275, 281-82 (2001). 126 28 C.F.R. § 42.203. 127 Terry v. Ohio, 392 U.S. 1 (1968). 128 People v. Threlkel, 2019 CO 18, ¶ 18. 129 People in Interest of K.D.W., 2020 COA 110, ¶¶ 16, 26. 130 Threlkel, ¶ 17. 131 § 24-31-309, C.R.S. 123 124 31 Starting in mid-2020, Senate Bill 217 required officers to report information about all stops, including the stopped person’s perceived demographic information, reason for contact, and result of contact.132 This provision is discussed further in Section 9. 6.3. Legal Requirements for Use of Force and Arrests If the stop progresses to an arrest, the law places additional limits on officers. When making an arrest, the officer must have a warrant, probable cause to believe a person committed a crime, or have personally witnessed the crime.133 As with reasonable suspicion in stopping someone, probable cause for an arrest cannot be supported by profiling based on race or other protected characteristics, though a physical description of the subject may, of course, include the subject’s race.134 And an officer cannot use excessive force when making an arrest or bringing a person into submission.135 If an officer uses excessive force, he or she is “subject to the criminal laws of [Colorado] to the same degree as any other citizen,” including the laws on homicide and assaults.136 Force is excessive if it is greater than that necessary to carry out an arrest, prevent escape, or prevent injury to the officer or others.137 And Senate Bill 217 now requires that officers use nonviolent means instead of force if possible.138 Colorado law now also requires officers to intervene when another officer is using excessive force and to report the excessive use of force.139 In addition, the “mentally ill or disturbed condition of the suspect is a relevant factor in determining reasonableness of an officer’s responses to a situation.”140 When a subject exhibits diminished capacity, an officer’s conduct may be unreasonable if the officer was not “in danger at the precise moment” that his or her conduct “unreasonably created the need to use such force.”141 In other words, if an officer’s tactics cause a subject in crisis to react aggressively, those actions may be “immediately connected to the suspect’s threat of force,” and the officer’s defensive use of lethal force may be considered unreasonable.142 Officers, therefore, should avoid tactics that escalate an encounter with a person they suspect is experiencing a mental health crisis. § 24-31-309(3.5), C.R.S. U.S. Const. Am. IV; § 16-3-102, C.R.S. 134 § 24-31-309, C.R.S. 135 McDaniel v. People, 499 P.2d 613, 615 (Colo. 1972). 136 § 18-8-803(1), C.R.S.; see also Campbell v. People, 133 P. 1043, 1046 (Colo. 1913) (explaining that an officer “stands upon the same plane” as a private citizen when committing criminal acts); Bourie v. Dep’t of Higher Educ., 929 P.2d 18, 21 (Colo. App. 1996) (recognizing clear “public policy that law enforcement officers have no immunity from criminal prosecution nor are they accorded any special status with respect to the use of force except in making an arrest”). 137 See § 18-8-803(2), C.R.S.; § 18-1-707, C.R.S. 138 § 18-1-707(1)-(4.5), C.R.S. 139 § 18-8-802(1), C.R.S.; § 18-8-802(1.5), C.R.S. 140 Estate of Ceballos v. Husk, 919 F.3d 1204, 1214 (10th Cir. 2019). 141 Id. at 1215. 142 Allen v. Muskogee, Okl., 119 F.3d 837, 841 (10th Cir. 1997). 132 133 32 The Tenth Circuit Court of Appeals, in four cases, has explained what types of tactics might be viewed as recklessly and deliberately escalating an encounter where a subject exhibits diminished mental capacity.143 In each case, the officers were on scene to conduct a welfare check. The person in crisis was armed but “posed harm to no one” when officers escalated the situation.144 But the officers acted quickly, yelled commands, crowded the person and often trapped them, and did not try to de-escalate the encounter.145 Officers on scene in one incident had received Crisis Intervention Training (“CIT,” discussed above), but they did not use those tactics or “tak[e] steps to calm the situation.”146 In each case, the person in crisis responded aggressively to the officers. One raised his gun, another lunged with his knife, a third pointed a sword at the officers, and a fourth approached with a raised baseball bat. All four were shot by officers who said they acted to defend themselves. The officers’ tactics, rather than defusing the situation, contributed to the person in crisis feeling defensive and acting aggressively. Particularly because the subjects were not criminal suspects, these tactics may have “unreasonably escalated the situation to the point deadly force was required.”147 The officers’ actions were “immediately connected to the suspect’s threat of force,” and the use of lethal force was therefore unreasonable.148 In all four cases, officers had been told by the dispatcher or people on the scene that the person was experiencing a mental health crisis. For example, in the most recent case, the court noted “the responding officers knew [the man’s] capacity to reason was diminished, whatever the underlying reason might have been—mental health problems, emotional distress, drunkenness, or drugs,” and so “an objective officer … would have taken those facts into account before provoking a fatal encounter.”149 Knowing that a subject is experiencing a mental health crisis, therefore, triggers an officer’s need to use modified tactics and de-escalation to avoid provoking the need for a higher level of force. The Tenth Circuit used this same analysis in another recent case to find that officers must consider whether a person is inebriated or otherwise impaired, even if the service call was See Estate of Ceballos, 919 F.3d 1204 (incorporating the analysis of incidents in Allen v. Muskogee, Okl., 119 F.3d 837 (10th Cir. 1997); Sevier v. City of Lawrence, Kan., 60 F.3d 695 (10th Cir. 1995); Hastings v. Barnes, 252 F. App’x 197 (10th Cir. 2007)). 144 Est. of Ceballos, 919 F.3d at 1217. 145 In Hastings, 252 F. App’x at 203, the court explained how the officers’ tactics caused the man to react defensively, provoking their use of force: He was a potentially mentally ill/emotionally disturbed individual who was contemplating suicide and had called for help. Rather than attempt to help [him], [the officers] crowded themselves in [his] doorway (leaving no room for retreat), issued loud and forceful commands at him and pepper-sprayed him, causing him to become even more distressed. Hastings, an unpublished case, was referenced in Bond v. City of Tahlequah, Oklahoma, 981 F.3d 808, 817 (10th Cir. 2020) and Est. of Ceballos, 919 F.3d at 1217. 146 Id. at 1211. 147 Id. 148 Allen, 119 F.3d at 841. 149 Estate of Ceballos, 919 F.3d at 1217. 143 33 for a trespass, and not a welfare check.150 Applying this law, a September 3, 2021 district court order denied four Aurora Police officers qualified immunity when one of the officers shot a man who had falsely reported to 911 that he had killed two people and was holding two more hostage.151 According to a lawsuit filed by the man’s family, the officers were told upon arrival to the apartment that the man was alone, unarmed, and had a mental health issue. The district court found that the officer who shot the man had reason to know the man was not a danger to anyone and was possibly experiencing a mental health crisis, but proceeded to corner and confront him at gunpoint, provoking the man’s defensive response. The other three officers were denied qualified immunity for failing to intervene to prevent the shooting, which occurred within two minutes of the four officers entering the apartment. 6.4. Legal Requirements for Use of Ketamine Finally, ketamine cannot be administered unless it is (1) used for bona fide medical needs, and (2) administered by or under the direction of a person licensed or legally authorized to do so.152 The unlawful administration of drugs constitutes second-degree assault and occurs where a person “intentionally causes stupor, unconsciousness, or other physical or mental impairment or injury” by administering a drug “for a purpose other than lawful medical or therapeutic treatment.”153 6.5. Recent Changes to Colorado Law Senate Bill 217, signed into law on June 19, 2020, reformed law enforcement operations in Colorado.154 Colorado lawmakers introduced the bill shortly after George Floyd’s death in Minneapolis.155 It requires law enforcement agencies across Colorado to make significant changes to their policies, training, and reporting requirements. The Colorado legislature made additional changes in 2021: House Bill 21-1250 updated some of Senate Bill 217’s requirements and deadlines and House Bill 21-1251 limited the use of ketamine.156 Both became effective on July 6, 2021. Senate Bill 217 created more specific statutory criteria for the use of deadly force. For example, the law requires that an officer both believe and have an objectively reasonable belief that the officer or another person “is in imminent danger of being killed or of Bond, 981 F.3d at 824. Bond held that it has been clearly established law that officers who respond to an agitated person and “unreasonably escalate a non-lethal situation into a lethal one through their own deliberate or reckless conduct” violate the Constitution. The defendants in this case have filed a petition asking the U.S. Supreme Court to hear this case which the Court has not yet decided. 151 Flores v. Aurora, No 1:20-cv-0018-RBJ, 2021 WL 4033117 (D. Colo. Sept. 3, 2021). 152 § 18-13-123(3), (4)(a), C.R.S. 153 § 18-3-203(1)(e), C.R.S.; see also People v. Nygren, 696 P.2d 270, 271-72 (Colo. 1985). 154 Colorado Senate Bill 20-217, Enhance Law Enforcement Integrity. 155 Colorado Governor Signs Sweeping Police Accountability Bill Into Law. Here’s How it Will Change Law Enforcement, The Colorado Sun, June 19, 2020. 156 Colorado House Bill 21-1250, Measures to Address Law Enforcement Accountability; House Bill 21-1251, Appropriate Use of Chemical Restraints on a Person. 150 34 receiving serious bodily injury” before using deadly force.157 Additionally, law enforcement may not use deadly physical force to apprehend someone suspected only of a minor or nonviolent offense.158 And the bill limited Colorado’s “fleeing felon law,” which previously allowed officers to use deadly force to stop a person running away if officers suspected them to be armed or to have used a weapon in a crime.159 As for less-lethal force, the bill bars officers from using carotid holds or other types of chokeholds, which apply pressure to a person’s neck so that it becomes difficult for the person to breathe or blood flow is cut off to the person’s brain.160 It requires officers to use non-violent means before resorting to physical force where possible and permits the use of force “only if nonviolent means would be ineffective in effecting an arrest, preventing an escape, or preventing an imminent threat of injury to the peace officer or another person.”161 It includes protest-specific provisions, barring officers from using chemical agents such as tear gas without warning and from firing less-lethal projectiles indiscriminately.162 And it imposes a new “duty to intervene” on law enforcement officers, requiring officers to intervene if they witness another officer using excessive force against a person.163 One method to ensure compliance with these provisions will be through the bill’s body-worn camera provisions, requiring all Colorado law enforcement agencies to outfit their officers with cameras by July 1, 2023.164 Officers must activate these cameras when interacting with the public to enforce the law or investigate possible violations of the law. The bill requires law enforcement agencies to abide by new data collection and reporting requirements. Agencies must collect demographic data on officers’ contacts with the public when enforcing the law or investigating possible legal violations, including the location, circumstances, and result of the interaction; data on when an officer unholsters their weapon; and use-of-force data, including information about the type of force used, injuries, and the basis for the encounter.165 By April 1, 2022, the agencies must report this information to the state.166 Finally, Senate Bill 217 changed the law for police misconduct claims. It altered Colorado’s qualified immunity provisions, allowing officers to be sued in their individual capacities and liable for up to $25,000 in damages when their employer determines that “the officer did not act upon a good faith and reasonable belief that the action was lawful.”167 And it § 18-1-707(4.5), C.R.S. § 18-1-707(2)(a), C.R.S. 159 Polis Signs Broad Police Accountability and Reform Bill Into Law After Weeks of Protests, CPR News, June 19, 2020. 160 § 18-1-707(2.5), C.R.S. 161 § 18-1-707(1), C.R.S. 162 § 24-31-905(1), C.R.S. 163 § 18-8-802(1.5), C.R.S. 164 § 24-31-902, C.R.S. 165 § 24-31-903(2), C.R.S.; § 24-31-309(3.5), C.R.S. 166 § 24-31-903(2), C.R.S. 167 § 13-21-131(4), C.R.S. 157 158 35 authorizes the Colorado Attorney General to investigate and file a civil suit against any governmental agency when the Attorney General has reasonable cause to believe that an agency has engaged in a pattern or practice of violating state or federal law.168 The recent ketamine law prohibits officers from requesting that paramedics give any chemical sedative to a suspect and requires other officers and paramedics to report when an officer requests sedation.169 The new law restricts paramedics from administering ketamine in the presence of law enforcement except in the case of a justifiable medical emergency, and then only when trained personnel assess the patient’s weight and required medical monitoring is available.170 The law also excludes “excited delirium” as a justifiable medical emergency.171 Excited delirium, discussed in Section 10, is the diagnosis most emergency medical services’ policies cite as triggering the urgent need to administer ketamine to police-involved patients. § 24-31-113, C.R.S. § 18-8-805, C.R.S. 170 § 25-3.5-209, C.R.S. 171 § 25-3.5-103(8.6), C.R.S. 168 169 36 7. Finding #1: Aurora Police Has a Pattern and Practice of Engaging in Racially Biased Policing Against People of Color as a Whole and Black People in Particular As discussed above in Section 6, state and federal law prohibit police departments from engaging in racially biased law enforcement activities. During the course of our investigation, we examined a host of qualitative and quantitative evidence, all of which point to a singular conclusion: Since at least 2018, Aurora Police has engaged in a pattern and practice of racially biased policing against people of color as a whole and Black people in particular. The racial disparities we observed extend to nearly every significant facet of police contact with the community, from interactions to arrests to uses of force. 7.1. Background Statistics reported publicly by Aurora Police that predate our investigation already suggested the possibility that Aurora Police’s law enforcement activities have been biased against Black individuals172 and other minority groups. These statistics showed that Aurora Police arrested and used force against Black individuals to a much greater extent than one would anticipate based purely on the percentage of Black individuals living in Aurora. For instance, according to data that Aurora Police provided to the FBI through its Uniform Crime Reporting program, between the years of 2015 and 2019, Black individuals accounted for nearly 40% of all reported arrests in Aurora.173 Yet, according to the U.S. Census Bureau, as of July 1, 2019, Black individuals made up less than 17% of Aurora’s population.174 By contrast, white individuals (non-Hispanic and Hispanic)175 accounted for nearly 58% of reported arrests, which tracks their 60% population share. Population vs. Arrest Percentages (2015-2019) 80.0% 60.4% 60.0% 58.0% 40.0% 40.0% 20.0% 0.0% - 16.5% African Americans ■ Population White (Hispanic & Non-Hispanic) Arrests Sources: FBI Crime Data Explorer; U.S. Census Bureau ACS Demographic and Housing Estimates 172 We use the terms “Black” and “African American” interchangeably, though most of the data sources use “African American,” so many of the charts and graphics in this section use that term as well. 173 Federal Bureau of Investigation, Crime Data Explorer. 174 2019 ACS 5-Year Data Profile, Demographic and Housing Estimates, U.S. Census Bureau. 175 The U.S. Census and other government data often distinguish between white (non-Hispanic) and white (Hispanic), and we likewise recognize that distinction throughout our analysis. However, for readability, in this report we use the term “white” to refer to non-Hispanic white individuals, unless otherwise noted. White (Hispanic) individuals are considered non-white in our analysis. 37 This means that from 2015 to 2019, the relative proportion of Aurora Police arrests involving Black subjects was approximately 2.5 times higher than would be anticipated based on the relative percentage of Black individuals in Aurora’s population alone. By contrast, the relative proportion of Aurora Police arrests of white subjects (Hispanics and nonHispanics) was lower than would be anticipated based on that group’s relative percentage of population.176 Publicly reported use-of-force data exhibited even greater patterns of disproportionality. In its annual use-of-force reports, Aurora Police reported the following data177 for Black and white individuals for the period of October 2015 to December 2020: Percentage of Total Aurora Police Use-of-Force Incidents Year 2015-16 2017 2018 2019 2020178 White 30.7% 33.9% 31.2% 34.5% 30% African American 53.8% 44.2% 48% 47.3% 36% Comparing the average reported use-of-force rates by ethnicity against the census figures showed an even greater disproportionality than the one seen with arrests: Population vs. Use of Force Percentages (2015-2020) 50.0% 45.0% 40.0% 35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% 45.9% 44.4% 32.1% 16.5% White (Non-Hispanic) African-American ■ Population Use of Force Sources: Aurora Police Use-of-Force Reports; U.S. Census Bureau ACS Demographic and Housing Estimates 176 The FBI’s arrest statistics for all crimes do not break out “white” into “Hispanic” and “non-Hispanic.” It does give ethnicity information for certain violent crimes and property crimes. 177 APD, 2016 Use of Force Annual Report at 8 (Oct. 1, 2015 to Sept. 30, 2016); APD, 2017 Use of Force Annual Report at 7 (Jan. 1, 2017 to Dec. 31, 2017); APD, 2018 Use of Force Annual Report at 7 (Jan. 1, 2018 to Dec. 31, 2018); APD, 2019 Use of Force Annual Report at 8 (Jan. 1, 2019 to Dec. 31, 2019); APD, 2020 Use of Force Annual Report at 7 (Jan. 1, 2020 to Dec. 31, 2020). 178 We note that, according to this report, the use-of-force rate for Black individuals appears to have gone down in 2020. However, given the unique circumstances presented by the COVID-19 pandemic and other events in 2020, we do not draw any conclusions from this data alone. Rather, our own analysis of use-offorce incidents, described below, focuses on a longer time frame (2018-2021). 38 Here, Aurora Police’s use of force against Black subjects was almost three times higher than would be anticipated based solely on the relative percentage of Black individuals in Aurora’s population. By contrast, use-of-force incidents involving white subjects were about 25% lower than would be anticipated based on that group’s relative population share. These publicly reported statistics align with feedback expressed by the community. We heard from many stakeholders that officers engage in over-policing in certain neighborhoods. We also heard anecdotal reports from Black community members about unwarranted stops and other examples from personal and family member experiences. But the above data have their limitations. For example, the FBI arrest data only report arrests for white individuals as a whole, without breaking down the numbers by Hispanic and non-Hispanic ethnicity. They also do not take into account individuals who are arrested more than once; the data merely report all arrests in a jurisdiction during a given year. As discussed below, the data we obtained from Aurora Police provided far more information and granularity, allowing the investigation team to do a much more rigorous and reliable analysis. More importantly, however, arrest and use-of-force statistics are only one piece of the puzzle. Along with the data, we had access to tens of thousands of police reports, ridealongs with police and fire personnel, interviews with police and fire leadership, body-worn camera recordings of many uses of force, discussions with community members, as well as input from law enforcement experts. This information, along with the data that we gathered and analyzed, make clear that Aurora Police has engaged in a continuing pattern and practice of racially discriminatory policing.179 7.2. The Team’s Investigation of Aurora Police’s Data Given the publicly reported data and other anecdotal evidence of biased enforcement activities against people of color, we set out to perform a more robust statistical analysis of Aurora Police’s internal data. Our goal was to determine whether there was statistically significant evidence that Aurora Police had engaged in racially biased policing activities. 7.2.1. Assembling the Data Analytics Team To accomplish this task, we put together a data analytics team comprised of professional econometricians and statisticians to assemble and analyze available data from Aurora Police on its activities. The members180 of the team are: • Dr. David K. A. Mordecai, president of Risk Economics, an advisory firm that specializes in risk and liability management, as well as forensic analytics; Limitations on available data from before January 1, 2018, prevent us from issuing findings for earlier periods. This is not meant to suggest, however, that Aurora Police was not engaged in racially discriminatory policing before 2018. 180 A complete description of the qualifications of each data analytics team member (including academic credentials, professional experience, and research concentrations) is contained in the attached Technical Appendix. 179 39 • Ms. Samantha Kappagoda, chief economist at Risk Economics; • Mr. Michael Kwak, executive vice president at Compass Lexecon, a leading economic consulting firm that provides support and analysis to law firms, corporations, and government clients; • Mr. Mihir Gokhale, vice president at Compass Lexecon; • Mr. Noah Mathews, senior analyst at Compass Lexecon; and • Mr. Peter Horvath, senior analyst at Compass Lexecon. Each of the individuals listed above donated substantial time and resources to this effort, and we are grateful for their involvement. 7.2.2. Data Collection Assembling a reliable data set on which to perform a statistical analysis was challenging. The main repository of data capturing Aurora Police’s law enforcement activity resides in the department’s “records management system,” or RMS, a database that is locally maintained by the City of Aurora, but which employs third-party software provided by Versaterm, a Canadian company that develops software for fire, police, and other public agencies. The RMS database contains, among other things, the information commonly found in police reports, such as the details of an event, the names and ethnicities of parties involved, the officers’ narrative description, and photos of an incident. But there are also other important sources of data that track Aurora Police activities. Chief among these is Aurora Police’s Administrative Investigations Management (AIM) system, which stores data related to use-of-force incident reports and officer performance data. In addition, Aurora Police’s computer-aided dispatch (CAD) system contains information related to officer locations and response times, as well as call notes from dispatchers and officers that are entered as events unfold. The goal of the data analytics team was to assemble data from these three repositories into a single data set against which they could run queries. Discussions with relevant personnel at Aurora Police revealed that the portions of the CAD database relevant to our investigation also existed within the RMS environment. The AIM system, however, remained separate. The team therefore focused on obtaining access to the RMS and AIM databases. We discussed several alternatives with Aurora Police and the City of Aurora, including creating a modified copy of the RMS and AIM databases for our statisticians to use. Ultimately, the City of Aurora and Aurora Police agreed to provide our team with virtual access to their servers, where we had direct access to the RMS and AIM databases containing the raw data used and created by Aurora Police. Using a statistical programming language called R, the data analytics team queried the joined RMS and AIM databases to extract police incident data recorded by the system. The extracted data was joined as a single dataset, further anonymized to remove personal 40 identifying information (e.g., names of individual officers or witnesses), and subjected to standard data cleaning procedures. The result was a preliminary dataset of over three million records, comprised of anonymized fields of interest for about 20 years’ worth of Aurora Police incidents. Importantly, to our knowledge, neither Aurora Police nor the City of Aurora restricted the data analytics team’s collection efforts by time or subject matter, so the team had essentially unfettered access to the available incident data from Aurora Police. However, because our investigation focuses on current practices, and because record-keeping practices were not as thorough for older periods, the data analytics team’s queries typically focused on data from 2018 to 2021. Along with the Aurora Police dataset, the data analytics team also relied on data from the Census Bureau and U.S. Department of Housing and Urban Development supporting estimates of population level demographics (e.g., race, ethnicity, age, gender, and median income). The resulting data set enabled a more detailed empirical analysis of Aurora Police activity using established statistical methodologies, which is described in greater detail in the attached Technical Appendix to this report. The investigation team used this analysis, along with other evidence, to reach the conclusions set forth below. This analysis would not have been possible without extensive assistance and coordination from data professionals at the City of Aurora and Aurora Police. We are extremely grateful for their cooperation and professionalism. 7.3. Data Analysis Based on our review of the data, we conclude that there is statistically significant evidence that from at least January 2018 through February 2021, Aurora Police disproportionately interacted with, arrested, and used force against people of color, particularly Black people, as compared to white individuals. Specifically, the data show that Aurora Police’s interactions with, arrests of, and uses of force against people of color were disproportionately higher than would be anticipated based on a racial or ethnic group’s percentage of Aurora’s overall population. These disparities persist across income, gender, and geographic boundaries. They even persist when accounting for individuals who have multiple interactions with police, arrests, and uses of force. Furthermore, the disparities persist even when Aurora Police’s arrest and use-of-force activities are compared to the demographic makeup of the subpopulation who police interact with in the first place (as opposed to the city’s population as a whole). In short, no matter how one looks at the data, it is clear that observed law enforcement outcomes for people of color in Aurora differ significantly from those experienced by their white counterparts. These data—particularly for Black individuals—are deeply troubling. 7.3.1. Interactions Police cannot arrest or use force against individuals who they do not first interact with. There is statistically significant evidence that Aurora Police disproportionately interacted 41 with individuals of certain races and ethnicities, particularly Black people, as compared to other races and ethnicities, including white residents. For purposes of this analysis, an “interaction” was defined broadly to include any recorded interaction with an officer that was related to a call for service of any kind. Many individuals had multiple interactions with Aurora Police during the relevant time. The data analytics team looked at both (1) the overall number of interactions Aurora Police had with individuals, without accounting for individuals Aurora Police interacted with more than once, and (2) the number of specific individuals Aurora Police interacted with during the same period. The data show that Aurora Police disproportionately interacted with people of color, particularly Black people. Over the three-year period from January 2018 to February 2021, the data analytics team found that the number of interactions Aurora Police had with people of color versus white individuals was far higher than would be anticipated based on relative population percentages alone, particularly for Black individuals. Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Population 3,203 23,917 60,909 101,562 1,296 Interactions 259 2,793 39,873 31,629 272 (as % of Population) 8.1% 11.7% 65.5% 31.1% 21.0% 190,887 163,765 74,826 47,745 39.2% 29.2% Non-White (All) White/Non-Hispanic Source: Technical Appendix, Figure 1.D The data above show that, from 2018 to 2021, Aurora Police had 74,826 interactions with non-white individuals.181 During the same period, they had 47,745 interactions with white individuals.182 In other words, Aurora Police had nearly one-and-a-half-times the number of interactions with people of color than with white residents during this time frame, which is far higher than would be anticipated based on relative population alone. As the graph below shows, the number of non-white interactions amounted to 39.2% of Aurora’s non-white population, while interactions with white subjects equaled only 29.2% of Aurora’s white population. Technical Appendix, Fig. 1.D. As explained in Section V of the Technical Appendix, each interaction with the same individual is counted as a separate interaction, meaning the total number of interactions likely includes multiple interactions with the same people. 182 Id. 181 42 Aurora Police Interactions as a Percentage of Each Group's Population Qanuary 2018 - February 2021) 80.00% 65.5% 60.00% 39.2% 40.00% 29.2% 20.00% 0.00% White Non-White Black/ African American Source: Technical Appendix, Fig. 1.D The data for Black individuals are particularly revealing. From 2018 to 2021, Aurora Police had 39,873 interactions with Black residents, which was equivalent to 65.5% of Aurora’s Black population.183 That number contrasts with the 47,745 white interactions, which as noted above was equivalent to only 29.2% of Aurora’s white population.184 This does not mean that 65.5% of Black individuals or 29.2% of white individuals living in Aurora had interactions with Aurora Police during the relevant time frame. The number of interactions referenced above includes multiple interactions with some individuals, as well as interactions with people who do not live in Aurora.185 However, as explained in more detail in the Technical Appendix, comparing the number of interactions to the relative population size provides one useful metric for assessing whether Aurora Police disproportionately interacts with Black and other minority groups, as compared to the city’s white population. Some may argue that the disproportionate interaction rate for minority groups is not a useful metric if attributable to the fact that certain members of the community will have multiple interactions with the police over time. If those with multiple interactions are disproportionately members of a minority community (for race-neutral reasons), then interaction data would likewise be skewed towards minorities as well. There are two problems with that criticism. First, if police disproportionately focus policing activities on minority neighborhoods, it is more likely that individuals in those Id. Id. 185 Given that Aurora has a higher percentage of Black residents than the combined surrounding communities, there is no reason to believe that non-resident Black individuals visiting Aurora could somehow account for the disparity. If anything, it suggests the opposite. 183 184 43 neighborhoods will have multiple interactions with the police. Second, the data here do not support the criticism. To account for individuals with multiple police interactions with Aurora Police, the data analytics team also examined how many times Aurora Police interacted with unique individuals from 2018 to 2021.186 The same pattern emerges: Unique Interactions with Aurora Police as a Percentage of Each Group's Population CTanuary 2018 - February 2021) 60.00% 46.1% 40.00% 29.6% 23.9% 20.00% 0.00% White on-White Black/ African American Source: Technical Appendix, Fig. 10.B From 2018 to 2021, Aurora Police had interactions with 56,571 different non-white individuals, which equates to 29.6% of Aurora’s racial and ethnic minority population, compared to 39,131 white individuals, which equates to 23.9% of Aurora’s white population.187 The numbers are again particularly glaring for Black individuals. Aurora Police interacted with 28,084 unique Black residents, which equates to 46.1% of Aurora’s Black population.188 These results are robust from a statistical perspective. As explained in the Technical Appendix, the data analytics team performed a series of statistical analyses known as “chisquared” tests on the data.189 One of these tests, the “chi-squared test for homogeneity,” examines whether an observed deviation for a given cohort (e.g., Black individuals) relative to a control group (e.g., white individuals) represents a statistically significant disparity.190 Application of that test to this data shows statistically significant evidence that Aurora Police disproportionately interacted with individuals based on race, particularly for Black individuals. Technical Appendix, Fig. 10.B. Id. 188 Id. 189 Technical Appendix, Section I. 190 See id. 186 187 44 Specifically, the analytics team calculated chi-squared statistics with p-values of 0.000 for both (1) the number of interactions with people of color as a whole versus white individuals; and (2) the number of interactions with Black individuals specifically versus white individuals. 191 Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Non-White (All) White/Non-Hispanic Population 3,203 23,917 60,909 101,562 1,296 Interactions 259 2,793 39,873 31,629 272 (as % of Population) 8.1% 11.7% 65.5% 31.1% 21.0% 190,887 163,765 74,826 47,745 39.2% 29.2% Chi-Squared Statistic P-Value 485.0 0.000 2,367.2 0.000 15,008.2 0.000 82.8 0.000 29.5 0.000 2,573.2 N/A 0.000 N/A Notes: P-Values less than .05 represent statistical significance at the 95% confidence level. Source: Technical Appendix, Figure 1.D The analytics team found 0.000 p-values both for the number of overall interactions (shown above), as well as for the number of unique individuals interacted with, as compared to their group’s percentage of overall population.192 The lower the p-value, the less chance any disparity can be explained by random chance. A p-value of less than 0.05 is statistically significant, providing a 95% confidence level that the data do not result from random chance, rather than some other factor (such as race or ethnicity). The 0.000 p-values here mean there is less than a 1 in 1,000 chance that the disparity in number of interactions by race and ethnicity could be explained by random chance alone. These observations are critical because, again, interactions determine the population of individuals who may be subject to arrest, use of force, or other enforcement actions. 7.3.2. Arrests There is also strong statistical evidence that Aurora Police disproportionately arrested people of color between 2018 and 2021, particularly Black individuals. In fact, Aurora Police arrested more members of non-white racial and ethnic groups than would be expected not only based on each group’s relative overall population, but also based on the number of people of color with whom Aurora Police interacted. Below is a chart comparing frequency of arrests by race and ethnicity compared to Aurora’s racial and ethnic make-up generally for 2018 to 2021. 191 192 Technical Appendix, Fig. 1.D. Technical Appendix, Fig. 10.D. 45 Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Population 3,203 23,917 60,909 101,562 1,296 Non-White (All) White/Non-Hispanic 137 1,075 16,904 11,846 106 (as % of Population) 4.3% 4.5% 27.8% 11.7% 8.2% 30,068 18,334 15.8% 11.2% Arrests 190,887 163,765 Chi-Squared Statistic P-Value 135.9 0.000 906.0 0.000 7,760.4 0.000 12.1 0.001 10.5 0.001 1,340.9 N/A 0.000 N/A Notes: P-Values less than .05 represent statistical significance at the 95% confidence level. Source: Technical Appendix, Figure 1.E The data show that over a three-year period, Aurora Police arrested 30,068 racially and ethnically diverse individuals, which equates to 15.8% of Aurora’s non-white population, as compared to 18,334 white individuals, which equates to 11.2% of Aurora’s white population.193 In other words, based on relative population percentages, Aurora Police arrested people of color more than 1.4 times more than white individuals. That multiplier was even greater for Black individuals, who were arrested almost 2.5 times more (27.8%) than white individuals (11.2%) based on relative population percentages.194 Arrests as a Percentage of Each Group's Population Qanuary 2018 - February 2021) 30.00% 27.8% 20.00% 15.8% 11.2% 10.00% 0.00% White on-White Black/ African American Source: Technical Appendix, Fig. 1.E As with the preceding data showing disproportionate interactions by race and ethnicity, the arrest data shown here also have chi-squared statistics with p-values less than 0.050, and are thus statistically significant evidence of a racial disparity that is not the result of random chance. Furthermore, even when accounting for individuals who were arrested more than once (i.e., measuring the ratio of unique arrests to population counts), statistically 193 194 Technical Appendix, Fig. 1.E. Id. 46 significant disparities persist for both the non-white cohort as a whole and Black individuals.195 Unique Arrests as Percentage of Each Group's Population Ganuary 2018 - February 2021) 25.0% 20.6% 20.0% 15.0% 10.0% 12.5% 9.7% 5.0% 0.0% White Non-White Black/ African American Source: Technical Appendix, Fig. 10.D These data cannot be explained by the fact that Aurora Police disproportionately interacted with people of color more than whites. Even among the universe of people interacted with, Aurora Police arrested a higher percentage of people of color than white individuals with whom they interacted. To reach this conclusion, the analytics team looked at all the police interactions from January 2018 to February 2021 and isolated instances when an arrest was made. It then subdivided that data by race and ethnicity and calculated how often an interaction with a member of a given minority group led to an arrest. The disproportionate treatment of people of color, and Black individuals in particular, persisted. Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Non-White (All) White/Non-Hispanic Interactions 259 2,793 39,873 31,629 272 137 1,075 16,904 11,846 106 (as % of Interactions During The Period) 52.9% 38.5% 42.4% 37.5% 39.0% 30,068 18,334 40.2% 38.4% Arrests 74,826 47,745 Notes: P-Values less than .05 represent statistical significance at the 95% confidence level. Source: Technical Appendix, Figure 2.D 195 Technical Appendix, Fig. 10.D. 47 Chi-Squared Statistic P-Value 22.3 0.000 0.0 0.941 144.0 0.000 7.2 0.007 0.0 0.896 38.7 N/A 0.000 N/A Specifically, Aurora Police arrested 42.4% of Black subjects interacted with, but arrested only 38.4% of white subjects interacted with.196 Again, these differences are statistically significant and cannot be explained by chance, particularly for Black individuals. The data analytics team calculated a p-value of 0.000 for the disparity in the number of arrests per interaction for Black versus white subjects.197 7.3.3. Reported Uses of Force Finally, there is strong evidence Aurora Police used force disproportionately against people of color, particularly Black people, as compared to white residents.198 Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Non-White (All) White/Non-Hispanic Population 3,203 23,917 60,909 101,562 1,296 5 12 605 310 2 (as % of Population) 0.2% 0.1% 1.0% 0.3% 0.2% 934 374 0.5% 0.2% UoF Incidents 190,887 163,765 Chi-Squared Statistic 0.7 32.2 596.1 14.4 0.3 P-Value 0.395 0.000 0.000 0.000 0.578 162.7 N/A 0.000 N/A Notes: P-Values less than .05 represent statistical significance at the 95% confidence level. Source: Technical Appendix, Figure 1.F As shown in the data above, Aurora Police used force on members of non-white racial or ethnic groups 934 times, which equates to 0.5% of the non-white population in Aurora.199 It used force on white subjects only 374 times, which equates to 0.2% of Aurora’s white population.200 In other words, Aurora Police used force against members of non-white racial and ethnic groups 2.5 times more than white individuals based on relative percentage of the population. Again, the numbers were even more glaring for Black residents. From 2018 to 2021, Aurora Police used force 605 times against Black residents, which equates to 1.0% of Aurora’s Black population.201 This means that Aurora used force against Black individuals Technical Appendix, Fig. 2.D. Id. Interestingly, the data also show statistically significant evidence that Hispanics or Latinos were arrested less than would be anticipated based on their relative share of Aurora Police interactions. This may be because of differences in frequency of arrests between Hispanic/Latino males and females. When gender is accounted for, there is statistically significant evidence that Aurora Police disproportionately arrested Hispanic and Latino males interacted with at a higher rate as compared to white males interacted with (44.5% versus 42.8%). Technical Appendix, Fig. 7.C. 198 Police departments sometimes have differing definitions of what constitutes a use of force. For example, some departments may consider a soft empty-hand control technique to be a use of force, while others may not. In addition, use-of-force severity classifications often differ across police departments. To avoid confusion here, the data analytics team only considered a given encounter to be a use of force if it was so designated by Aurora Police. 199 Technical Appendix, Fig. 1.F. 200 Id. 201 Id. 196 197 48 roughly 5 times more than they did against white individuals, based on the relative percentage of each group’s population: Uses of Force as a Percentage of Each Group's Population Qanuary 2018 - February 2021) 1.5% 1.0% 1.0% 0.5% 0.5% - 0.2% 0.0% White Non-White Black/ African American Source: Technical Appendix, Fig. 1.F Once again, the data analytics team found p-values of 0.000 for these results, representing a statistically significant disparity involving the use of force against people of color that cannot be explained by chance alone.202 And similarly, the racial disparities persist even when the accounting for instances where Aurora Police used force against the same person multiple times (i.e., measuring the ratio of unique uses of force to population counts).203 In fact, when measured as a percentage of population, for every unique use of force on a white subject, there were 2.5 times as many uses of force on a non-white subject and 5 times as many unique uses of force on a Black subject.204 Again, these disparities are statistically significant and are not the result of random chance.205 Furthermore, as with the arrest data discussed above, the disproportionate use-of-force data is not due to the fact that Aurora Police interacted with people of color more often than they did with white individuals. When looking at Aurora Police interactions from 2018 to 2021 during which a use of force occurred, statistically significant racial and ethnic disparities persisted: Id. Technical Appendix, Fig. 10.F. 204 Id. The data analytics team observed 893 unique uses of force against non-white subjects, which amounted to 0.5% of that group’s population, and 579 unique uses of force against Black individuals, which amounted to 1.0% of Aurora’s Black population. In contrast, the team observed only 359 unique uses of force against white subjects, which was 0.2% of Aurora’s white population. 205 Id. 202 203 49 Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Interactions 259 2,793 39,873 31,629 272 Non-White (All) White/Non-Hispanic 5 12 605 310 2 (as % of Interactions During The Period) 1.9% 0.4% 1.5% 1.0% 0.7% 934 374 1.2% 0.8% UoF Incidents 74,826 47,745 Chi-Squared Statistic 3.0 3.9 105.3 8.4 0.1 P-Value 0.084 0.048 0.000 0.004 0.798 59.2 N/A 0.000 N/A Notes: P-Values less than .05 represent statistical significance at the 95% confidence level. Source: Technical Appendix, Figure 2.F For instance, Aurora Police used force on roughly 1.2% of members of non-white racial and ethnic groups who had at least one interaction with police from 2018 to 2021 compared to 0.8% for white individuals.206 That disparity appears to have been driven in large part (though not exclusively) by Aurora Police’s more frequent use of force against Black subjects. Aurora Police used force against 1.5% of Black subjects who had at least one interaction with police from 2018 to 2021. That is nearly double the corresponding figure for white subjects.207 Racial differences in arrest rates also cannot explain the use-of-force disparity, at least as to Black individuals. Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Non-White (All) White/Non-Hispanic 137 1,075 16,904 11,846 106 5 12 605 310 2 (as % of Arrests During The Period) 3.6% 1.1% 3.6% 2.6% 1.9% 30,068 18,334 934 374 3.1% 2.0% Arrests UoF Incidents Chi-Squared Statistic 1.0 4.0 76.6 10.6 0.1 P-Value 0.307 0.046 0.000 0.001 0.815 48.9 N/A 0.000 N/A Notes: P-Values less than .05 represent statistical significance at the 95% confidence level. Source: Technical Appendix, Figure 2.E The above table examines arrests during which a use of force occurred between 2018 and 2021. The data show that Aurora Police used force roughly 3.6% of the time when a Black individual was arrested, as compared to 2.0% of the time for arrests involving white subjects.208 This difference is again statistically significant with a p-value of 0.000.209 Technical Appendix, Fig. 2.F. Id. 208 Technical Appendix, Fig. 2.E. 209 Id. The data analytics team calculated a p-value of 0.000 for the disparities in uses of force per arrest for Black versus white subjects, as well as for non-white subjects as a whole versus white subjects. There is also statistically significant evidence that Aurora Police used force incident to arrest against Hispanics and Latinos more often than against whites. 206 207 50 7.4. Explanations for the Data The above analysis makes clear that Aurora Police have disproportionately interacted with, arrested, and used force against people of color, particularly Black individuals. These disparities cannot be explained by random chance. The data analytics team consistently calculated chi-squared statistics with p-values of 0.000—much lower than the 0.050 necessary to suggest a statistically significant disparity—showing that the racial and ethnic disparities in interactions, arrests, and use of force were not coincidental. Having identified these disparities, the data analytics team next looked to see whether these findings were persistent when controlling for certain factors other than race or ethnicity, such as geography, gender, income, or severity of crime. The data reveal that the racial disparities persist in almost all circumstances, particularly for use of force and particularly for Black individuals. 7.4.1. Geography, Gender, and Age and Use of Force Geography cannot explain the racial and ethnic disparities observed in Aurora Police’s use of force. Aurora is divided into three police districts, each with different demographic and socio-economic characteristics. Yet the statistically significant relationships between race and ethnicity and use of force—both as a percentage of interactions and a percentage of arrests—remained consistent for non-white individuals as a whole and Black individuals specifically across all three Aurora Police districts.210 As to gender, there is statistically significant evidence that Aurora Police disproportionately used force against Black males and females as compared to white males and females—both as a percentage of interactions and as a percentage of arrests.211 As to age, there is statistically significant evidence that Aurora Police disproportionately used force against Black subjects as compared to white subjects for all age groups between 18 and 49, both as a percentage of interactions and as a percentage of arrests.212 The disparity goes away for individuals over 50, which is not surprising given the much lower frequency of uses of force against older individuals in general.213 Technical Appendix, Figs. 4.A & 4.B. The data analytics team calculated p-values of 0.000 for the disparities in uses of force per interaction and per arrest for Black individuals versus white individuals in each of the three districts. Statistically significant disparities for non-white individuals as a whole were found in all three districts, with 0.000 p-values in districts 1 and 3. 211 Technical Appendix, Figs. 7.A & 7.B. The data analytics team calculated p-values of either 0.000 or 0.001 for the disparities in uses of force per interaction and per arrest for both Black males and females compared to white males and females. As discussed above, gender did appear to play a role with regard to Hispanics and Latinos: there is statistically significant evidence that Aurora police disproportionately used force against Hispanic or Latino males but not against Hispanic or Latino females. 212 Technical Appendix, Figs. 8.A & 8.B. The data analytics team calculated p-values of 0.000 for the disparities in uses of force per interaction and per arrest for Black versus white subjects in the following three age groups: (1) 18-21, (2) 22-29, and (3) 30-49. Id. The data analytics teams looked only at age cohorts for Black and white subjects. 213 Id. 210 51 7.4.2. Type of Offense or Level of Force Used The statistically significant relationship between race and ethnicity and use of force also persists across different use-of-force tiers and misdemeanor versus felony arrests. When looking either at the population of individuals arrested or interacted with, the data show that Aurora Police disproportionately used force against non-white subjects in general, and Black individuals in particular, as compared to white individuals for uses of force categorized as Tier 1 (lowest) and categorized as Tier 2 or 3.214 In other words, the observed disproportionality in occurrences for non-white and Black subjects persists across incidents involving use of force with differing degrees of severity. Similarly, racial disparities exist regardless of the severity of the offense (i.e., misdemeanor vs. felony). Specifically, the data show that Aurora Police disproportionately used force against non-white subjects in general, and Black subjects in particular, as compared to white subjects for uses of force incident to both misdemeanor and felony arrests.215 7.4.3. Income It has been suggested that one of the reasons people of color may experience disproportionate levels of police interaction, arrests, and uses of force is that police often concentrate law enforcement activities in lower income areas, which tend to have higher crime rates, and where populations of non-white individuals are also higher. Under this theory, the observed differences in policing activities that members of non-white racial or ethnic groups experience are more a function of income than race. The data do not support this argument. For instance, if this argument were correct, one would expect the disproportionate treatment of people of color to disappear at higher income levels. But that is not at all what we found. The data analytics team looked at data for interactions, arrests, and uses of force between 2018 and 2021 and assigned each observed event to an income quartile based on the median household income of the zip code where the event occurred.216 For each of the four quartiles, the data analytics team calculated whether racial disparities existed for Technical Appendix, Figs. 3.A & 3.B. For Hispanics, there was not a statistically significant relationship between Tier 1 uses of force against Hispanics versus whites. However, there was for Tier 2 and 3 uses of force. 215 Id. Figs. 5.A & 5.B. With one exception, the data analytics team calculated p-values of 0.000 for the disparities in uses of force against people of color as a whole, and Black individuals specifically, versus white individuals for both felonies and misdemeanors—both as a percentage of interactions and as a percentage of arrests. For felonies, the p-value was 0.006—still statistically significant—for disparities in uses of force incident to felony arrests for people of color as a whole and for Black individuals specifically versus white subjects. For Hispanics, there was not statistically significant evidence about the disparity in uses of force incident to misdemeanors for Hispanics versus white subjects, but there was for uses of force incident to felonies. 216 The income level of each individual subject is not observed in the data. However, the median household income by region within each incident occurred can be assigned based on the zip code in which the incident is recorded to have occurred. In other words, while the data do not allow for the identification of high income and low income individuals, they do allow for the identification of high and low income areas in which incidents occurred. 214 52 minority groups based on their relative share of the population within the income quartile, versus white individuals in the same income quartile. For example, for interactions that occurred in zip codes in Income Quartile 1 (median household income between $45,431 and $58,992 annually), the data analytics team observed that Black subjects had 27,484 interactions with police, which equaled 69.6% of their population in zip codes corresponding to this income quartile.217 White subjects, by contrast, had 24,910 interactions, which accounted for 30.1% of their population share in zip codes corresponding to this income quartile:218 Panel A. Income Quartile 1 Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Non-White (All) White/Non-Hispanic Population 3,736 14,081 39,480 130,076 415 Interactions 196 1,620 27,485 22,712 209 (as % of Population) 5.2% 11.5% 69.6% 17.5% 50.3% 187,788 82,817 52,222 24,910 27.8% 30.1% Chi-Squared Statistic P-Value 759.8 0.000 1,516.3 0.000 9,756.0 0.000 3,601.4 0.000 56.2 0.000 103.8 N/A 0.000 N/A Notes: P-Values less than .05 represent statistical significance at the 95% confidence level. Source: Technical Appendix, Figure 13.A As shown above, Aurora Police interacted with Black subjects within Income Quartile 1 over two times more than they did with white subjects, when measured as a share of each group’s relative population in the quartile (i.e., 69.6% for Black subjects versus 30.1% for white subjects). This disproportionality was statistically significant, with a p-value of 0.000. This disparity did not disappear at higher income levels. In Income Quartile 4 (median household income between $105,658 and $136,144 annually), statistically significant disparities in police interaction rates persisted for both Black and Hispanic individuals: Panel D. Income Quartile 4 Race/Ethnicity American Indian/Alaska Native Asian Black/African American Hispanic or Latino Native Hawaiian/Pacific Islander Non-White (All) White/Non-Hispanic Population 1,365 24,165 15,201 32,467 466 5 333 2,009 1,584 17 (as % of Population) 0.4% 1.4% 13.2% 4.9% 3.6% 3,948 6,230 5.4% 2.5% Interactions 73,664 245,324 Notes: P-Values less than .05 represent statistical significance at the 95% confidence level. Source: Technical Appendix, Figure 13.A 217 218 Technical Appendix Fig. 13.A Id. 53 Chi-Squared Statistic P-Value 25.4 0.000 121.9 0.000 5,159.7 0.000 557.8 0.000 2.3 0.133 1,412.0 N/A 0.000 N/A As shown above, for zip codes within Income Quartile 4, Aurora Police had 2,009 interactions with Black individuals, which equaled 13.2% of their population in zip codes corresponding to this income quartile.219 This is over five times higher than the 2.5% interaction rate for white subjects in Income Quartile 4. Once again, the observed interaction disparity for Black individuals was statistically significant, as was a similar disparity observed for Hispanics in Income Quartile 4.220 The data analytics team next looked at arrests and uses of force and found similar disparities across nearly all income quartiles, particularly with respect to Black individuals.221 7.4.4. Officer Discretion The data analytics team further investigated whether the statistically significant disparities were indicated across cases where for which officers are granted greater discretion to act or arrest versus cases for which they are not. If the disparities are greater where officers have greater discretion, that is another strong indication that racial bias may be at play. The data analytics team evaluated three types of cases: suspicious occurrence, disturbance/noise complaint, and domestic dispute. Officers have the least discretion to stop and arrest in domestic dispute cases, where officers must arrest a suspect if there is probable cause to believe domestic violence has occurred.222 Suspicious activity cases fall towards the other end of the spectrum. Officers may decide to stop an individual based on a call or based on observing the individual directly, whether or not a complaint has been made. Whether the stop proceeds to an arrest is subject to far greater discretion, particularly if the stop reveals only potential misdemeanor activity. On our ride-alongs, we observed officers using significant discretion as to whether to stop or arrest subjects even where they observed a factual basis to do so. Disturbance/noise complaint cases fall somewhere in between. In those cases, there has been a report or observation of potential unlawful activity that impacts others, but officers still retain significant discretion over whether to make an arrest. The data analytics team identified statistically significant evidence that Aurora Police disproportionately used force against Black individuals, as compared to white individuals, in suspicious activity cases—both as a percentage of interactions (2.5 times more) and as a percentage of arrests (2 times more).223 By contrast, the analytics team did not find Id. Id. (showing p-values of 0.000 in Income Quartile 4 for non-white subjects as a whole as well as Black and Hispanic individuals in particular) 221 Technical Appendix, Figs. 13.B & 13.C. 222 § 18-6-803.6 C.R.S. 223 Technical Appendix, Figs. 6.A & 6.B. Aurora Police used force against Black subjects in 4.3% of suspicious activity interactions, while it used force against white subjects in only 1.7% of such cases (i.e., twoand-a-half times as much). Aurora Police used force against Black subjects in 7.8% of suspicious activity arrests, while it used force against white subjects in only 3.9% of such cases (i.e., twice as much). 219 220 54 statistically significant evidence that the uses of force in domestic violence224 or noise disturbance/complaint cases225 were disproportionate across race or ethnicity. The difference in use-of-force disparities between domestic violence cases and suspicious activity cases is striking. Granted, the analytics team only compared uses of force for three types of cases. But the fact that disparities in use of force all but disappear where officers have little to no discretion provides more evidence that bias plays a role in the disparities when officers have greater discretion. That biased-based policing would reveal itself where officers have greater discretion is unsurprising. When there is implicit or unconscious bias, it is more likely to make a difference, even if unintentional, in discretionary decision-making. If there is intentional bias, it is more likely to manifest when it is least likely to be detected—i.e., where the decision-making is discretionary.226 7.4.5. Use of Population-Level Demographics Some have suggested that comparing police activity to the relative demographics of a community’s population as a whole—as we have done above—is not an appropriate way to measure disproportionality. Instead, it has been suggested that police activity should be compared only to the demographics of the criminal offender population (as opposed to the population at large), under the theory that some non-white racial and ethnic groups may commit crime at different rates. We disagree. As an initial matter, the idea that individuals of certain races or ethnicities have a greater propensity to commit crimes because of their race or ethnicity is unsupported by any reliable evidence and contrary to law. While it is true that observed arrest and conviction rates can and do differ among racial and ethnic groups, we recognize that these differences could arise for many reasons, including income level disparities, differential policing efforts, community willingness to report crime, or other factors. To address this concern about the use of relative population percentages, we did four things. First, we looked at what comparisons Aurora Police itself uses when evaluating uses of force by its officers. In its Use of Force Annual Reports, Aurora Police compares the race of those it uses force against with relative population percentages.227 Id. Aurora Police used force against Black subjects in 1.7% of domestic violence related interactions, while it used force against white subjects in 1.4% of such cases. Aurora Police used force against Black subjects in 4.4% of domestic violence arrests, while it used force against white subjects in 4.1% of such cases. 225 Id. Aurora Police used force against Black subjects in 3.4% of disturbance/noise related interactions, while it used force against white subjects in 2.5% of such cases. Aurora Police used force against Black subjects in 8.7% of disturbance/noise complaint related arrests, while it used force against white subjects in 7.8% of such cases. 226 Whether officers are exercising discretion to stop, arrest, or use force more often against people of color or are simply patrolling neighborhoods where people of color live more often, both constitute bias-based policing. 227 See, e.g., APD, 2020 Use of Force Annual Report at 7. 224 55 Second, we examined the approaches used by other recent studies of policing. Population level data analysis features prominently in this analysis.228 Third, as described above, we looked to see whether Aurora Police’s disproportionate interactions with, arrests of, and uses of force against people of color could be explained by income level. They could not. Instead, a pattern of statistically significant racial disparities in interactions, arrests, and uses of force persisted across all income levels, even though people of color make up a higher percentage of the lower-income population in Aurora. Fourth, we did not limit our statistical analysis to only population-level comparisons. As noted below, our team also looked more narrowly at the subset of all Aurora Police interactions between 2018 and 2021 (which is smaller than the overall population of Aurora) and then identified instances during which an arrest or a use of force occurred. Even among this smaller subset population, the data show that once an interaction with Aurora Police had begun, people of color—particularly Black people—faced disproportionate levels of arrest and use of force as compared to their white counterparts. Other comparisons, such as to the reported crime rate by race or offender information provided by victims, create significant uncertainty as not all crimes are reported, victims do not always accurately describe offenders, and stereotypes often fill in gaps in uncertain data. Though the comparison of arrests and uses of force to population levels does not eliminate all error, we find it most reliable for our purposes, particularly when combined with our other analyses.229 7.4.6. Excluding Incidents Where Race Information is Unknown or Unavailable Race and ethnicity information was not always available for each of the interactions, arrests, and use-of-force instances that the data analytics team observed. Among possible explanations, this may be because the officer recording the incident could not determine the race or ethnicity of the subject, or because they simply failed to enter the information when filling out a police report.230 The data analytics team decided to exclude such instances from their analysis after confirming that doing so would not affect the validity of their results. Specifically, the data analytics ran two alternate versions of each of their calculations, one in which the “unknown race” instances were assigned to the “white” category and one in which the See, e.g., Andrew Gelman, Jeffrey Faga & Alex Kiss, An Analysis of the New York City Police Department’s “Stop-and-Frisk” Policy in the Context of Claims of Racial Bias, 102 J. OF THE AM. STATISTICAL ASS’N 813, 816 (2007) (discussing relationship between the race of those stopped and the racial breakdown of the city population); Bocar A. Ba, et al., The role of officer race and gender in police-civilian interactions in Chicago, 371 SCIENCE 696 (2021) (comparing officer behavior among districts with different racial demographics). 229 See also Technical Appendix at 6 n.5. 230 Indeed, as we discuss below in Section 9, Aurora Police has failed to properly document its interactions with the public as required by Colorado law, and part of that failure includes the lack of demographic information in many Aurora Police records. 228 56 instances were assigned to the “non-white” category. Across these test results, the statistical significance of observed disproportionalities with respect to race and ethnicity persisted.231 7.5. The Data Combined With Other Evidence Shows a Pattern and Practice of Race-Based Policing While the data indicate that Aurora Police has engaged in racially and ethnically biased policing, we do not rely on the data alone to conclude that Aurora Police engages in an unlawful pattern and practice of race-based policing. When the data and other evidence and information we gathered are combined, the unlawful pattern and practice becomes clear. All of the recent high-profile use-of-force incidents in Aurora involve non-white individuals. We heard from community members about their experiences with Aurora Police and what they perceived as Aurora Police’s focus on non-white residents. Defense counsel reported that they often saw non-white residents face highly discretionary charges such as for failure to obey a lawful order only to have those charges later dropped. We personally observed differences in how Aurora Police officers engaged with the community on our ride-alongs based on the race of the subject. We noted frequent escalations of force against non-white residents as we observed the Force Review Board meetings. These many observations, along with the data analysis, provide overwhelming support for our conclusion that Aurora Police engages in a pattern and practice of race-based policing. We do not believe that such a comprehensive data analysis is necessary to prove a pattern and practice violation. But given the availability of the data and the capabilities of the data analytics team, we performed such an analysis here. 7.6. Changes Needed To remedy and eliminate its practice of race-based policing, Aurora must make major changes across the organization to improve its culture, including improving its policies, training, recordkeeping, and hiring. These required changes all relate to contributing factors that cause Aurora Police to violate the law. 7.6.1. Improved and More Detailed Policies and Guidance to Prevent Racially Biased Policing Aurora Police Directive 8.32 is a four-page directive that prohibits “biased based policing,” defined as “enforcement action based on a trait common to a group.” 232 Such traits include race, ethnicity, gender, national origin, language, religion, sexual orientation, gender identity, age, and disability.233 Because “[b]iased based policing undermines legitimate law enforcement efforts, alienates a significant percentage of the population and fosters See, e.g., Technical Appendix, Fig. 1.D at n.6. Aurora Police Department Directive 08.32 (revised Oct. 7, 2020) By contrast, the directive addressing procedures for complaints and disciplining officers is a 32-page, highly detailed document. See Aurora Police Department Directive 10.02 (revised May 19, 2020). 233 Aurora Police Department Directive 08.32 (revised Oct. 7, 2020). 231 232 57 distrust of law enforcement by the community,” the directive states that officers “will not utilize biased based policing as a basis for a contact or the detention, seizure of persons or assets.”234 This directive does not define impermissible “biased-based policing.” It merely restates the constitutional requirements for a lawful Terry stop: “Reasonable suspicion must be supported by specific articulable facts that persons contacted regarding their identification, activity or location, has, is, or is about to commit a violation of the law or presently constitutes a threat to the safety of themselves or others.”235 This is a common problem throughout Aurora Police’s directives—they are too often cold recitations of statutes or legal standards pulled from court opinions with no practical examples of what is and is not permissible or any explanation of how to identify and avoid engaging in discriminatory behavior. Baltimore’s Policy 317, titled Fair and Impartial Policing, and associated policies provide a model for Aurora Police to consider.236 Policy 317 has detailed sections on “Required Actions,” “Prohibited Actions,” “Supervisory Requirements,” and “Training and Compliance.”237 Beyond setting forth what does and does not constitute discriminatory policing, Policy 317 requires officers who witness discriminatory policing to report it; makes clear that violations of the policy will “result in discipline, re-training, counseling or other remedial intervention”; provides for detailed collection and regular audits of data; mandates that officers provide mechanisms for individuals to file and pursue complaints; and makes clear that a history of discriminatory policing will be considered in performance evaluations, hiring, and transfers.238 Policy 317 does not stand alone. Aspects of the policy are woven into every major Baltimore policy.239 7.6.2. More Specific Standards and Expectations on When Officers Should Stop, Arrest, or Use Force The data, coupled with additional observations, showed differences among race are most pronounced when officers’ discretion over whether to stop and arrest is greatest. One way to address that problem is, as other major departments have done, to develop policies that provide more specific standards and expectations for stops and arrests and require officers to specifically articulate why they chose to stop or arrest. For instance, policies that permit and generally expect officers to stop an individual only when certain, clearly articulated conditions are met create more specific standards. Under such policies, officers can only arrest an individual if certain, clearly articulated conditions are met—and if those conditions are met, the officers should generally arrest that individual. Id. Id. 236 Baltimore Police Department Policy Number 317, Fair and Impartial Policing. This report cites several policies that were developed by cities under consent decrees. Because the policies were written in collaboration with federal policing experts, many are considered models in police reform. 237 Id. 238 Id. 239 Id. (identifying 17 associated policies). 234 235 58 Aurora Police must adopt policies and training that more clearly articulate when stops and arrests can and should happen.240 For instance, Aurora Police does not have a written policy dedicated to when officers may stop an individual. The only place it is addressed is Aurora Police Directive 8.32 on “biased based policing,” which recites the high-level constitutional requirements for a lawful investigative stop but does not incorporate the additional requirements necessary under Colorado law: Sworn members must have a legal basis for making a contact, whether consensual or non-consensual, for the purpose of enforcing the law or investigating possible violations of the law. Legal basis can be consent or reasonable suspicion or probable cause. Consent must always be truly voluntary. A consenting individual is always able to withdraw his/her consent at any time and is always free to stop talking and/or leave the encounter with the member whenever he/she desires. Reasonable suspicion must be supported by specific articulable facts that persons contacted regarding their identification, activity or location, has, is, or is about to commit a violation of the law or presently constitutes a threat to the safety of themselves or others. Probable cause refers to the reasonable belief that evidence exists establishing a basis to arrest, conduct a personal or property search or to obtain a warrant. Members will not utilize biased based policing as a basis for a contact or the detention, seizure of persons or assets.241 The policy provides no detail about what circumstances or activities satisfy the constitutional requirements, let alone the additional, more stringent state requirements. More importantly, there is nothing at all about when a stop should be made. Under the policy, once minimum constitutional requirements for a stop exist, the officer has complete discretion over whether to make the stop. Aurora Police must amend its policy to reflect Colorado law and provide additional support to officers in how to make these decisions. Aurora Police’s directives about arrests similarly do not provide specific guidance. Directive 6.01 (“Arrest Procedure”) states, “[o]fficers may arrest without a warrant only upon determination that probable cause exists to believe that a crime was committed and that the individual to be arrested committed the crime or for a crime committed in the officer’s presence.”242 This tells officers when an arrest may be made but does not make clear when an arrest should be made. It does not differentiate between crimes—e.g., it does not set different standards for non-violent misdemeanors or status offenses (except in some cases for juveniles) or violent crimes. Officers again have complete discretion over whether to arrest once it appears a crime has been committed. Aurora Police policies regarding the use of force are discussed below in Section 8. Aurora Police Department Directive 08.32 (revised Oct. 7, 2020) (emphasis in original). 242 Aurora Police Department Directive 06.01 (revised Nov. 4, 2020) (emphasis in original). 240 241 59 Under these policies, whether or not a particular individual is stopped or arrested depends not so much on the objective circumstances, but whether the officer on scene decides to make the stop or make the arrest. These decisions will differ depending on the officer, workload, directives from sergeants, and other factors leading to inconsistent policing and similar offenders receiving variable treatment. For example, during ride-alongs, we repeatedly observed that differences in policing and outcomes depended not on objective circumstances, but on the individual officer’s approach. At least one officer admitted that one of the things he liked most about the job was the discretion he had in the field—the discretion allowed him to be lenient or develop creative solutions short of arrest where he believed appropriate. While that sentiment is laudable and while some discretion can lead to positive policing and a reduction in harsh outcomes, too much discretion across an organization consisting of many different individuals with different training and backgrounds will inevitably lead to unjustifiably different policing outcomes. Baltimore again provides a good potential model for Aurora. For example, Policy 1112 is a 24-page document laying out, in detail, when officers may engage in different interactions with residents—e.g., voluntary interactions, investigative interviews, stops, weapons patdowns, and searches.243 For stops, the policy not only provides a more detailed explanation of what constitutes “reasonable suspicion” for a stop, but it also lays out specific procedures for how to conduct a stop and the scope of what may be done during a stop. It says that the “scope of the stop must be tied to the basis for it,” making clear that additional justification must exist for taking certain actions (such as demanding the individual’s ID, ordering a motorist out of a vehicle, applying handcuffs, conducting a patdown, or using any level of force).244 The policy prohibits officers from making “pretext stops” based on race or where the justification is loitering or misdemeanor trespass.245 It prohibits officers from stopping someone based solely on a person’s response to police presence or proximity to the scene of a suspected or reported crime.246 It prohibits officers from manufacturing a justification for a stop by provoking or trying to provoke a flight response.247 And it requires officers to document every stop, using specific and descriptive language detailing the reason for the stop rather than boilerplate or coded words.248 Policy 1106 contains similar detail and prescriptions for warrantless arrests.249 Because Colorado law prohibits using any force if it can reasonably be avoided, use-offorce policies similarly should provide more structure to support officer decision-making. Aurora Police should give its officers more guidance in how to avoid force and, as importantly, when force is an appropriate next step in an interaction with a resident. Baltimore Police Department Policy 1112, Field Interviews, Investigative Stops, Weapons Pat-Downs & Searches. 244 Id. at 1112.18-20. 245 Id. at 1112.25-26. 246 Id. at 1112.30, 1112.32. 247 Id. at 1112.31. 248 Id. at 1112.33-34. 249 Baltimore Police Department Policy 1106, Warrantless Arrest Procedures and Probable Cause Standard. 243 60 7.6.3. Track Outcomes for Those Arrested for Misdemeanors Baltimore’s focus on monitoring arrest outcomes for misdemeanor arrests provides useful guidance here. The Baltimore report suggests monitoring misdemeanor arrest outcomes to determine whether arrests are unconstitutional or discriminatory, because if booking officers or the district attorney does not advance charges, that arrest was likely not supported by probable cause. The report identified that central booking and district attorneys declined to advance a disproportionate percentage of misdemeanor charges against Black residents—an indication that many arrests of Black residents were unjustified and based in discrimination. “Tracking arrest outcomes is an important tool for imposing accountability as well as identifying officers who would benefit from additional training, guidance, or other early intervention.”250 Aurora has no articulated process in place that tracks outcomes of arrests or provides feedback to Aurora Police. As we discuss below, we found that charges are often dismissed for those arrested for failure to obey a lawful order in Aurora. Developing a process to track the disposition of misdemeanor arrests will reduce the ability of these charges to be used in a race-based manner. 7.6.4. Improved Academy and In-Service Training As noted above, the Aurora Police training academy is undergoing major changes and is shifting its focus away from militaristic techniques to an emphasis on empathy and interpersonal skills. This philosophical shift is a good start. And Aurora Police can improve its training (both at the academy and in-service level) in additional ways. We conclude that improvements to Aurora Police’s training are required in at least four areas: bias, deliberate decision-making, recordkeeping requirements, and specific articulation of the basis for an encounter. As to bias, while Aurora Police’s current in-service and academy training incorporate antibias components, such programming must more effectively incorporate scenario-based training so that officers are presented with real-world scenarios rather than abstract concepts. This is a concern that we heard repeatedly in interviews, and we agree with this assessment. While we note that there appear to be efforts underway to increase and improve scenario-based teaching (particularly for in-service training), specific, measurable goals would ensure that these changes are implemented. We also note that there has not, to our knowledge, been any widespread scenario-based training on the new requirements of Senate Bill 217. In addition, training needs to include feedback from the actual experiences of Aurora Police officers with critical incidents and the review of its officers’ uses of force. With respect to avoiding unnecessary escalation, a consistent theme that we heard during our interviews was that Aurora Police’s training needed to shift away from teaching officers what they can do, and instead focus on what they should do. The emphasis on what was permissible rather than proper was a chief contributor, according to several Aurora Police interviewees, to the August 2020 incident where officers held a Black family (including four 250 U.S. Department of Justice, Investigation of the Baltimore City Police Department at 46, Aug. 10, 2016. 61 children between the ages of 6 through 17) at gunpoint.251 We agree that Aurora Police’s training should focus more on what officers ought to do in a given situation, and not just on what can be justified under department policies or the U.S. Constitution and Colorado law. Colorado state law requires, starting in June 2020, police officers to track and report demographic information for every contact with civilians as well as requiring documentation of the reason for the contact.252 Aurora Police failed to implement this requirement as state law requires. This absence of required information prevented Aurora Police from knowing who its officers contacted and why, and limited our ability to analyze the full scope of Aurora Police’s interaction with the public. This failure to follow the law hindered our data analysis. Aurora Police must take immediate steps to comply with state law’s requirements for recording and tracking contacts. Finally, Aurora Police must improve its training for officers on how to properly articulate the basis for encounters. In our review of Use-of-Force and General Offense Reports, we often saw vague, conclusory articulations for the reasons why an encounter occurred. For example, an officer might vaguely report that a subject “resisted” without specifying what specific actions or behaviors the subject did to resist. We note that Aurora has developed a new in-service training on this issue, which must be a continued training focus going forward. Greater detail from officers is necessary to ensure proper evaluations and to develop necessary improvements to policies and training. One approach could include having reporting included in scenario-based training where the reviewers know the full facts of the scenario and can provide training covering both the officer’s conduct during the scenario and the reporting of the scenario. 7.6.5. Better Timely Recordkeeping One of the challenges we faced was in accessing comprehensive data reflecting the activities of Aurora Police. Broadly, that data were split across three repositories: (1) the Records Management System (RMS), which contained most of the police activity including General Offense reports; (2) the Computer-Aided Dispatch (CAD) system, which had callspecific dispatch information and was partially (though not completely) integrated into the RMS; and (3) Use-of-Force Reports stored in Aurora Police’s Administrative Investigations Management (AIM) database. While CAD data is partially available in the RMS environment, AIM is a separate system, and matching up AIM records to their corresponding RMS data proved to be challenging. Some of these systems have shortcomings that hamper Aurora Police’s ability to monitor and understand the activities of its officers. For example, the AIM system used to track use-of-force incidents was not designed for this purpose. Rather, it was designed as an early intervention system intended to track a broad range of performance issues (including absenteeism, complaints, and reprimands). Since its implementation with Aurora Police in 2011, the system has morphed into the primary tool by which Aurora Police tracks and See Family Sues Aurora Police Over Botched Stolen-Car Response That Left Children Handcuffed, Held at Gunpoint, The Denver Post, Jan. 25, 2021. 252 § 24-31-309(3.5), C.R.S. 251 62 measures its officers’ use-of-force incidents. Aurora Police personnel were candid with us about the shortcomings of AIM. For example, the system often requires manual entry of data, which can lead to errors and subjective bias. There was a broad consensus that AIM needed to be replaced, and we agree with this assessment. Based on feedback from the relevant stakeholders, a more comprehensive product would improve Aurora Police’s ability to track metrics that are both sought by the community and required under Senate Bill 217. More broadly, there is a critical need for Aurora Police to have better timely insight into the activities of its officers. Currently, Aurora Police releases an annual Use-of-Force Report, which is created from data extracted from AIM. That data requires substantial manual intervention and clean-up before it can be presented in a user-friendly way. And an annual review is not enough. Aurora Police’s use-of-force reporting should be done—at a minimum—on a monthly or quarterly basis and, wherever possible, shared with the community and oversight bodies. Aurora Police must also engage in similar real-time reporting for the statistically significant racial bias we have uncovered for police interactions and arrests. At a minimum, supervisors should be able to access a robust dashboard of relevant information about officers under their command to facilitate appropriate supervisory intervention.253 And at the leadership level, the ongoing use of more sophisticated analytics and dashboards is necessary to make sure that Aurora Police—and the public—can appropriately understand the effect of police activity at the community and organization level. Once an appropriate system is chosen, Aurora Police should establish a corresponding policy to ensure supervisors are familiar with the system and officers understand its role in both commendation and supportive intervention.254 7.6.6. Improved Hiring and Recruiting Procedures, Including Major Changes to the Civil Service Commission Standards Studies show that an officer corps that reflects the diversity of the community that it serves promotes professionalism and builds community trust, confidence, and legitimacy. In February 2021, researchers from four top universities published a detailed study examining the role officer race and gender played in policing outcomes in Chicago.255 The study examined records from 2.9 million officer shifts and 1.6 million enforcement events by nearly 7,000 officers covering 2012 to 2015.256 Researchers concluded that Black, Hispanic, and female officers made fewer stops and arrests and used force less often than white, male officers. The authors found that “these enforcement disparities are predominantly focused on relatively minor crimes, not violent offenses, suggesting little Karen L. Amendola & Robert C. Davis, Best Practices in Early Intervention System Implementation and Use in Law Enforcement Agencies, National Police Foundation. 254 New Orleans Police Department Operation Manual, Insight: Early Intervention System. 255 Bocar A. Ba, et al., The role of officer race and gender in police-civilian interactions in Chicago, 371 SCIENCE 696 (2021); see also Mark Hoekstra & CarlyWill Sloan, Does Race Matter for Police Use of Force? Evidence from 911 Calls, National Bureau of Economic Research Working Paper 26744 (Feb. 2020). 256 Ba, et al. at 1. 253 63 trade-off in terms of public safety.”257 The largest contribution to the gap for each group was attributable to interactions with Black civilians.258 For instance, Black officers made 12.55 fewer stops of Black civilians per 100 shifts than white officers, a reduction equal to 39% of typical white-officer volume. By contrast, Black officers made only 1.31 fewer stops of white civilians per 100 shifts than their white counterparts, a reduction equal to 17% of typical white-officer volume.259 Black officers also had fewer interactions where the underlying justification was discretionary—e.g., Black officers made 5.72 fewer stops per 100 shifts for “suspicious behavior” than white officers (a reduction equal to 31% of average white-officer volume).260 While we lacked enough information on officer race or gender to do a similar analysis in Aurora, Aurora’s use-of-force statistics from 2019 suggest at least the possibility that Black officers are less likely to use force than their white counterparts. Black officers were responsible for 2.5% of use-of-force incidents, while making up 3.9% of the sworn staff. That was the lowest use-of-force ratio for officer race/ethnicity tracked.261 Obviously these data do not prove a relationship in Aurora between officer race and policing outcomes, but they are consistent with findings from the Chicago study. Aurora must do more to attract and hire a more diverse group of qualified police officers. Aurora Police releases basic demographic information about sworn officers each year. Below is a graph with race and ethnic breakdown for sworn staff for 2016 to 2019262 compared to percentage of Aurora’s overall population:263 Aurora Police Sworn Staff Versus Population of Aurora 90% 80.8% 80% 70% 60% 50% 44.4% 40% I 30% 20% 10% 0% White 28.6% 1 4.1% 1 9.4% 16.5% Black - 1.4% 6.5% 0.9% 0.8% Hispanic/Latino Asian American Indian ■ % 2019 Population % Sworn Staff 0.4% 0.1% - 5.2% 3.4% Hawaiian/ Pacific Two or More Races Islander Sources: Aurora Police Annua1 Rcpons; U.S. Census Bureau ACS Demographic :and Housing Estim:ues Id. at 6. Id. at 3-6. 259 Id. at 4. Similar though less pronounced effects were observed for Hispanic and female officers vis-à-vis Black civilians. Id. at 4-5. 260 Id. at 4. 261 APD, 2020 Use of Force Annual Report at 7. White officers were involved in 78.8% of the use-of-force incidents while making up 80.1% of the sworn staff in 2019. Hispanic officers were involved in 10.1% of the use-of-force incidents while making up 9.7% of the sworn staff. Id. When it comes to gender, females were involved in only 8.4% of the use-of-force incidents while making up 12.4% of the sworn staff. Id. 262 We have averaged the reported percentages for 2016-2019 taken from: APD, 2016 Annual Report at 7; APD, 2017 Annual Report at 8; APD, 2018 Annual Report at 8; APD, 2019 Annual Report at 6. 263 2019 American Community Survey 5-Year Data Profile, Demographic and Housing Estimates, U.S. Census Bureau. 257 258 64 While Black residents make up over 16% of Aurora’s population and roughly 40% of arrestees,264 only about 4% of sworn officers are Black. Similarly, while Hispanics and Latinos make up over 28% of Aurora’s population, less than 10% of sworn officers are Hispanic or Latino. By contrast, over 80% of the sworn officers are white—nearly double the percentage of non-Hispanic white civilians in the general population. There are multiple steps Aurora and Aurora Police must take to attract and hire a more diverse group of qualified police officers. First, Aurora and Aurora Police must improve outreach and recruiting procedures to target diverse new and lateral hires. Second, Aurora must improve the Civil Service Commission’s hiring and promotion requirements and procedures for police officers and firefighters. If the current structure remains in place, a neutral, third-party expert should evaluate and make recommendations for revisions to any requirements or procedures that impair the recruitment, hiring, and promotion of female and minority candidates but are not otherwise necessary to ensure individuals are qualified for the job.265 Statistics presented by the Civil Service Commission in September 2020 suggest there is substantial room for improvement on this score.266 Below is a chart showing the number of individuals—by race, gender, and ethnicity—who (1) applied for positions with the Aurora Police, (2) then met the minimum qualifications, (3) then passed the ergometrics exam, (4) then underwent a full background check, and finally (5) were offered a job.267 Race/Ethnicity White Black Hispanic or Latino Asian 2+ or Other Total Applied 3,267 591 1,317 165 624 5,964 Met Minimum Qualifications 2,809 (86%) 454 (77%) 1,037 (79%) 139 (84%) 510 (82%) 4,949 (83%) Passed Ergometrics Exam 975 (30%) 122 (21%) 297 (23%) 48 (29%) 179 (29%) 1,621 (27%) Subjected to Full Background Check 413 (13%) 42 (7%) 112 (9%) 18 (11%) 80 (13%) 665 (11%) Made Formal Job Offers 119 (3.6%) 5 (0.8%) 32 (2.4%) 5 (3%) 22 (3.5%) 183 (3.1%) Federal Bureau of Investigation, Crime Data Explorer. Aurora’s hiring requirements and procedures are set forth in the Rules and Regulations of the Civil Service Commission. 266 Aurora Civil Service Commission 2020 Overview and Recent Entry-Level Hiring Summary (Sept. 9, 2020) at 56-60. 267 Id. 264 265 65 Only 0.8% of Black applicants (5 out of 591), as compared to 3.6% of white applicants (119 out of 3,267), were offered a job. Only 1.1% of Black applicants (5 out of 454) who met minimum qualifications were offered jobs, as compared to 4.2% of white applicants (119 out of 2,809) who met minimum qualifications. Indeed, at every stage of the process, a lower percentage of Black applicants than white applicants made it to the next step.268 This racial winnowing at every step of the process strongly suggests a problem with Aurora’s recruitment and hiring standards and procedures. Though the Civil Service Commission made some changes in response to the 2009 Department of Justice investigation, the way the current process works in practice shows that those changes were not substantial enough. We do not suggest that white officers cannot police in an unbiased way, or that the racial or ethnic make-up of a police force must precisely mirror that of the population policed. But law enforcement agencies—particularly those with a demonstrated history of racially biased policing—should strive to make their force more representative of the community they serve. Communities are more likely to trust police departments with cultural awareness, language skills, and similar backgrounds to the members of the community that they serve.269 And having such diverse officers help the entire law enforcement agency improve its approach to working with the public. Only 27% of Black applicants who met minimum requirements (122 out of 454) passed the ergonomics exam as compared to 35% of white applicants (975 out of 2,809). Only 34% of Black applicants who passed the ergonomics exam (42 out of 122) were subjected to a full background check as compared to 42% (413 out of 975) of white applicants. Only 12% of Black applicants subjected to a full background check (5 out of 42) were made formal job offers as compared to 26% of white applicants (109 out of 413). 269 How to Support Trust Building in Your Agency, U.S. Dept. of Justice & VERA Institute of Justice, 15 (2016). 268 66 8. Finding #2: Aurora Police Has a Pattern and Practice of Using Force Excessively We found that Aurora Police has a pattern and practice of using force excessively. Aurora Police generally approaches the use of force with a what-can-be-justified-underthe-outer-limits-of-the-law approach rather than a what-force-is-necessary approach. We observed officers using force to take people to the ground without first giving them adequate time to respond to officer commands; using more force when people subject to pain-control techniques respond with expected resistance; and generically reciting “stop resisting” when trying to control subjects when body-worn camera footage and other available information does not suggest such resistance. In our review of Force Review Board meetings, we saw Aurora Police continuing to analyze force only under the requirements of U.S. Constitution rather than also using the more limiting requirements of Colorado law that generally do not permit the use of force unless officers first try other possible alternatives.270 In addition, our investigation found that Aurora Police often views de-escalation as requiring officers to calm down after using force rather than avoiding unnecessary escalation in the first place. We observed that officers often approach scenes with a showof-force mentality, bringing many officers to the scene and using gunpoint and threatened force often disproportionate to the risk presented. Finally, many encounters we observed involved people undergoing mental health crises. The way that officers responded to these calls varied dramatically. Sometimes traditional high-show-of-force approaches were used, which tended to escalate the situation. Other times, much more low-key, problem-solving approaches were used that generally led to quicker and safer outcomes. Aurora Police must improve its method of approaching people undergoing mental health crises. 8.1. Law and Aurora Policies Aurora Police Department Directive 5.03, the current version effective August 6, 2021, summarizes Senate Bill 217’s use-of-force standard and states, in part: “when sworn members use force, they will de-escalate the amount of force used when that force is successful, and control is gained.”271 But then it says that “[p]hysical force may be used as allowed by State statutes” and then repeats the Senate Bill 217 language. This permissive language instructs officers that they may use force whenever it does not violate state statutes. § 18-1-707(1), C.R.S. Aurora Police Department Directive 05.08, (revised May 6, 2021) titled “Less Lethal Devices, Weapons and Techniques,” also contains a section on de-escalation techniques. 270 271 67 Other policies, such as those recently adopted by Boulder,272 Colorado Springs,273 and Denver274 provide a structure for critical decision-making and de-escalation that requires officers to try to determine why someone may not be complying with a command (mental condition or impairment, language barrier, drug or alcohol impairment, developmental issue) and, if able, attempt de-escalation alternatives. Rather than just set the outer bounds of permissible force, these policies require de-escalation throughout encounters and set forth guiding principles for the use of force. While other Colorado agencies275 include a subject’s mental state or diminished capacity as a factor officers should consider before applying physical force, Aurora’s policy does not address the need to consider de-escalation for people undergoing mental health crises. The only policy addressing mental health evaluations is Directive 6.13, which applies only when the subject is gravely ill or an imminent danger to themselves or others. But mental health issues exist on much more of a continuum, and assessment of the subject’s mental health should occur in almost all interactions where officers may use force. We spoke with several members of the Colorado law enforcement community who raised concerns that if a department’s use-of-force policy included restrictions beyond those mandated by state or constitutional law, officers and Aurora Police might be vulnerable to litigation. This concern of creating additional grounds for liability is not founded in our state or federal laws. Although an officer can be disciplined for violating policy, “[i]t is clear that a Fourth Amendment excessive force claim, whether arising out of deadly or nondeadly force, cannot be based upon violation of police department or city policy.”276 The Supreme Court held that police policies, which “vary from place to place and from time to time,” are too unreliable to use as a standard for constitutional policing.277 Colorado alone has over 250 different law enforcement agencies, most with their own unique use-of-force policy. True, some courts in other jurisdictions have included an officer’s policy violation as a factor in their analysis of whether the force used was reasonable. But in those cases, policy was just one factor of many, and that practice has not been adopted in Colorado’s jurisdictions. And Colorado law limits personal liability of officers only to those situations where the officer’s employer determines “that the officer did not act upon a good faith and reasonable belief that the action was lawful,” so whether or not an officer complies with a more restrictive policy than the law requires does not affect their potential personal liability.278 Boulder Police Department Policies & Procedures. Colorado Springs Police Department Policies & Procedures. 274 Denver Police Department Operations Manual. 275 See policies from Boulder County, Canon City, Durango, Englewood, Erie, Fort Collins, Jefferson County, Lafayette, Longmont, Northglenn, and Pueblo. 276 Martin A. Schwartz, Section 1983 Litig. Claims & Defenses, § 3.12 Excessive Force Claims: Fourth Amendment Protections (4th ed. Supp. 2020). 277 Whren v. United States, 517 U.S. 806, 815, (1996); See also Thompson v. City of Chicago, 472 F.3d 444, 455 (7th Cir. 2006) (“because police rules, practices and regulations vary from place to place and from time to time, they are an unreliable gauge by which to measure the objectivity and/or reasonableness of police conduct.”). 278 §13-21-131(4), C.R.S. 272 273 68 Rather, Colorado courts permit police agencies to create more restrictive use-of-force policies, and have held that departments are “free to adopt a policy that applies a more stringent standard than the baseline constitutional standard.”279 “[P]olice departments may—indeed, they should—impose higher internal standards on their officers than simply not violating state criminal law and avoiding federal damages liability.”280 Aurora’s policy, unlike other jurisdiction’s policies that emphasize limiting force, does not provide a structure for officer decision-making and leaves it to the officer to interpret the statute while in stressful and complex situations.281 This lack of structure could leave officers more open to civil liability. 8.2. Data and Information Reviewed 8.2.1. Documents Relating to the Use of Force We reviewed Aurora’s Use of Force Annual Reports from 2016 to the present, which summarize at a high level the use of force by Aurora Police each year.282 We also collected and reviewed over 2,800 Use-of-Force Reports from 2016 through 2020. For many of these reports, we requested and received the General Offense reports that provided more detail. 8.2.2. Force Review Board and Other Meetings We also attended nearly every Force Review Board meeting for the past ten months, where the Board discusses all significant uses of force. These weekly reviews typically examined General Offense reports for each incident, body-worn camera footage, and documentation of injuries. For the uses of force associated with Aurora’s support of Denver law enforcement’s response to the protests following the killing of George Floyd, the Force Investigation Unit prepared abbreviated presentations that generally focused only on bodyworn camera video. In addition, we attended several Internal Review Board meetings, where a mix of civilian members and police personnel assist the Chief of Police in imposing discipline, sometimes for use-of-force violations.283 8.2.3. Ride-Alongs We also spent over 190 hours in ride-alongs in all three districts, concentrated during high demand afternoon and evening times. During the ride-alongs, we followed a particular officer or sergeant for an entire shift, and observed how the officers interacted with the public and approached the use of force in a wide range of situations. We accompanied Johnson v. Civil Serv. Comm’n of City & Cty. of Denver, 2018 COA 43, ¶ 30. Turney v. Civil Serv. Comm’n, 222 P.3d 343, 350 (Colo. App. 2009). 281 Denver updated their use-of-force policies to include a decision-making model, explanations of the force continuum, de-escalation requirements, and a requirement that officers exhaust all non-force alternatives before using force. 282 APD, 2016 Use of Force Annual Report; APD, 2017 Use of Force Annual Report; APD, 2018 Use of Force Annual Report; APD, 2019 Use of Force Annual Report; APD, 2020 Use of Force Annual Report. 283 City of Aurora, Independent Review Board. 279 280 69 officers and witnessed first-hand as they, among other things, intervened in domestic violence incidents, encountered and took into custody someone undergoing a mental health crisis who threatened self-harm, used force to remove a resident from his home when he did not fully cooperate with officers investigating an assault by a relative, and used force to handcuff a witness who observed a subject in an escaping car that endangered an officer. We observed that the situations officers encounter vary dramatically and require quick judgment under uncertain information. Many officers were thoughtful, careful, and hesitant to escalate force. Others were quick to draw their weapon and escalate situations. 8.2.4. Community Concerns Based on Interviews Aurora stakeholders’ opinions of the police ranged from anger to distrust to support. The following themes arose during our interviews with community members. Accountability. Many in the community perceive that Aurora Police does not hold accountable officers who use excessive force or otherwise engage in racially biased actions. This perception underscores the tension created when disciplinary decisions made by Aurora Police are reversed or reduced by the Civil Service Commission. Many observe that Aurora Police leadership are not able to hold officers responsible for policy violations which undermines the credibility of the agency, leads to low morale, and erodes public trust. Community members observe the inequity in allowing officers who engage in unlawful conduct to remain employed with few, if any, consequences. Another related concern is Aurora Police’s internal investigation process. Community members believe that Aurora Police polices and investigates itself, resulting in inadequate investigations and presumptions of policy compliance. The perception is that investigations into excessive use of force are designed to exonerate the officers, regardless of the actual facts. Until 2021, Aurora Police had an agreement with the Denver Police Department to support each other in critical incident review. Many in the community criticized the way that Aurora and Denver implemented this agreement. Aurora Police has since ended this agreement with Denver and, in April 2021, joined the Critical Incident Review Teams with other law enforcement agencies in the 17th and 18th Judicial Districts. Since the change occurred recently, we have not had a chance to evaluate its operation. Transparency. Many community members expressed a desire to have more information about critical incidents promptly disclosed. Many felt that the way Aurora Police conducts internal affairs investigations and excessive use-of-force reviews is largely hidden from the public. Additionally, many stated that the Civil Service Commission’s influence on Aurora Police is opaque and undermines confidence in the disciplinary process. Over-Policing. Most of the Black community members we interviewed felt that their neighborhoods were over-policed. Increased surveillance and enforcement actions that involved confrontation and use of force were thought to be common. In contrast, mostly white or affluent neighborhoods had less police activity, and the interactions were perceived to be public-safety focused. One Black man stated that his children were detained by Aurora Police and transported home without justification, even though they 70 were just kids walking in their own neighborhood. Another Black woman described how she often saw youth smoking marijuana near two fast food restaurants in different Aurora neighborhoods. She said that police did not engage with the white youth although their conduct was obvious and illegal. However, she said police repeatedly confronted Black youth for the same behavior. Visibility. By contrast, many non-Black community members expressed the perception that Aurora Police is not reachable or visible in the community. They did not feel that Aurora Police has made an effort to engage with their communities—for example, an LGBTQ+ advocacy organization was unaware that Aurora Police has an LGBTQ+ liaison officer until finding the person’s name buried on Aurora Police’s website. A director of a shelter for individuals experiencing homelessness expressed the wish that officers would come by to chat and get to know those individuals, not just respond to calls for service. And other community members expressed the opinion that Aurora Police seems disengaged and likely to not take certain incidents—such as street racing or a dispute between two gay men—seriously, leading to community members avoiding calling the police. Response to Mental Health Calls. Aurora community members expressed concerns about officers responding to calls involving people experiencing a mental health crisis. At a fundamental level, members of the public question whether officers should respond to such calls. When people are having a mental health episode, it may be more suitable to respond with medical professionals who can provide treatment. One stakeholder talked about the scope of officers’ responsibilities creeping beyond appropriate boundaries. Other stakeholders worried about the lack of training officers receive and suggested that more education about mental health issues and response would be beneficial. Additionally, increasing community outreach should help familiarize officers with individuals who are likely to have increased contact with police due to their mental health status, which should lead to more treatment-based interactions. 8.3. Findings We found several types of encounters where Aurora Police used force in what appeared to be an excessive manner that did not satisfy federal or Colorado law. We set out the two specific categories of force where we find Aurora has a pattern and practice of violating the law and find that its general approach to the use of force leads to excessive force in other cases. 8.3.1. Force and Threatened Force Used Against Those in Mental Health Crises We observed a pattern of using force against those in mental health crisis without trying less confrontational ways to reduce the risk of force. This pattern of using force in this way violates applicable Colorado law because officers do not, as the law requires, “apply nonviolent means, when possible, before resorting to the use of physical force.”284 And our review of prior use-of-force incidents before Senate Bill 217’s effective date in 2020 shows 284 § 18-1-707(1), C.R.S. 71 that, based on information available to us, there was a prior pattern of using force against those in mental health crises that did not satisfy even the then-applicable, less-stringent requirement of being objectively reasonable. Inaccurate and confusing policies—both old and recently updated—likely contribute to this pattern. In May 2021, Aurora Police updated its use-of-force reporting policy to require that officers document when they used any force if the subject was not a criminal suspect. The old policy had many deficiencies. It did not require officers to file a report “when restraining persons solely for medical, emotional, or mental health purposes,” because “[t]hese situations do not involve an application of the criminal law.”285 Aurora Police properly no longer exempts these uses of force from reporting requirements. A separate policy on civil commitments authorizes officers to initiate a 72-hour medical hold if a person in a mental health crisis appears gravely ill or suicidal. Under the old useof-force policy, if officers determined that civil commitment was needed, they could use force to take the subject into custody, but only had to report the incident as a use of force if it resulted in injury. This policy wrongly created a separate category of residents who had force used against them by Aurora Police without the same level of review or scrutiny, leading to reduced accountability. Aurora Police updated this policy in May 2021 and again in August 2021, but the updated policy still does not comply with Colorado law and creates confusion by giving the wrong guidance to officers. It maintains the separate category of residents in crisis by specifically identifying persons who need restraining for “medical, emotional, or mental health purposes” who are “potentially causing or imminently threat[en]ing to cause Serious Bodily Injury or death [to] another individual.”286 First, Colorado law does not permit the use of force against anyone who just “is potentially causing” or “imminently threat[en]ing to cause Serious Bodily Injury.” Rather, officers can use force, as applicable here, only to prevent “an imminent threat of injury” and “shall apply nonviolent means, when possible” before using any force.287 Second, the policy does not define what “potentially causing” or “imminently threat[en]ing” mean. It does not provide guidance about whether verbal threats to harm someone, however remote, qualify or whether any possibility, however small, meets the “potentially causing” standard. Nor does it give any assistance to officers to understand what “medical, emotional, or mental health purposes” mean. Finally, this policy fails to address the authority officers have to act when someone is in a mental health crisis and there is probable cause to take that person into custody for purposes of a mental health hold. The use-of-force policy should be read in harmony with the civil commitment statute.288 Aurora Police Department Directive 05.04 (revised May 13, 2019). Aurora Police Department Directive 05.04 (revised Aug. 12, 2021). 287 § 18-1-707(1), C.R.S. 288 § 27-65-105, C.R.S.; Aurora Police Department Directive 09.06 (effective Sept. 7, 2021) on coordination with Aurora Fire, addressed below in Section 10, similarly fails to harmonize the use of force statute with the requirements of the civil commitment statute. 285 286 72 Our team could not locate another use-of-force policy that applied different reporting standards or use-of-force standards to different categories of residents.289 The new policy also strongly discourages officers from assisting paramedics should a patient become combative.290 But Senate Bill 217’s limitations on when an officer can use physical force reflects the limitations on when an officer should initiate an emergency mental health hold.291 And Colorado law does not permit paramedics to initiate an emergency mental health hold.292 This means that police and paramedics must work together to assist individuals who are experiencing a profound mental health crisis that places themselves or others in danger. Because Aurora Police implemented this new policy after almost all of our ride-alongs, our team could not observe how the new policy may have changed the way officers and paramedics worked together to assist a person in crisis, but this policy does not move Aurora Police forward in how to engage with community members who face mental health challenges. Under the prior policy in place in 2020, we saw officers engage with significant threatened force to those undergoing mental health crises. For example, when one of our team was on a ride-along, a Black man walked into a gym in a strip mall and claimed he had a knife and planned to kill himself. Several officers and members of the Crisis Response Team (unarmed, specially trained professionals not in uniform who focus on mental health) met in a nearby parking lot to plan the approach. In this meeting, and again on the radio, the sergeant reminded everyone to stay safe and use deadly force only as a last resort given the likely mental health profile of the subject. But in this encounter, officers quickly took cover behind their vehicles and adjacent vehicles and drew their weapons (mostly handguns with several tasers and at least one long gun). By aiming their weapons at the man, they were all pointing their weapons at a crowded gas station 25 yards behind the man. An officer started shouting at the man asking him to stop and put his hands up. He stopped advancing but did not comply with the “hands up” order. He started shouting that he didn’t want to live anymore, he had a knife, and no one knew how hard his life was. After a tense standoff, he finally began to We reviewed policies accessible to the public through law enforcement websites. Many Colorado law enforcement agencies, Aurora included, provide online access to their policies for transparency and to increase community trust. A few Colorado agencies include separate policies on emergency mental health holds, civil commitments, and subjects attempting suicide, but none created a separate use-of-force or reporting standard for civil custody encounters that deviated from their general use-of-force policy. 290 Aurora Police Department Directive 05.04 (revised Aug. 12, 2021). 291 An emergency mental health hold is permissible under § 27-65-105, C.R.S., only if one is gravely ill, or a danger to themselves or others, which are defined as conditions where the person poses an imminent threat of substantial bodily harm or serious bodily harm to themselves or others. The laws on when a civil commitment may be initiated are consistent with Senate Bill 217’s prohibition on physical force unless there is an imminent threat of serious bodily injury or death. 292 However paramedics may assist a patient who lacks capacity to give medical consent, which may include patients who qualify for an emergency mental health hold. See Aurora EMS Protocols, General Guidelines #0070. 289 73 kneel and eventually lay down. Several officers swarmed him, handcuffing him and placing him on a mental health hold. The officers did not, as the law requires, appear to “apply nonviolent means, when possible” before resorting to force. Rather, officers immediately escalated the confrontation by displaying an overwhelming number of cars and officers and immediately drawing their weapons, yelling at the subject to comply. The subject had all the hallmarks of seeking just such a confrontation, and the team made no effort to consider alternative, less confrontational approaches, particularly given the reports that he was armed only with a knife. A recent report by the Police Executive Research Forum recognized the importance of distinguishing between the threat posed by those likely armed with a knife or other weapon only capable of serious harm at close distance and those armed with a firearm.293 The next call on that same ride-along demonstrated a different approach. In the next call, a white man was in crisis because he was very drunk and exhibiting mental health issues. He was experiencing homelessness and had walked into a strip mall grocery store and asked the store to call 911 to take him to detox. When our team arrived, he was belligerent because Aurora Fire (who arrived before police) had told him he had to go to the hospital. In the middle of a tirade, he threw his cane at the officers and fire personnel in the middle of a tirade. Rather than drawing a weapon or yelling at the man, a junior officer who had been on the force for about a year then walked up to him, extended his hand, and said, “I’m [Joe]294, you look to be hurting. How can we help you?” The man replied, “It’s about *** time that someone treated me like a human being instead of yelling at me. I’m a former Navy SEAL and deserve some respect.” The officer then asked the man to sit down on the gurney next to him to talk about what was wrong. The man got agitated again a couple of times, but responded to the officer’s questions and within ten minutes the ambulance departed taking the man to the hospital. Though the first incident involved a man who claimed he had a knife and the second incident did not, overall differences in the two approaches still stand out. In the first incident, no effort for a less confrontational approach was considered or attempted as the law requires. And many in the community observed several white police officers yelling at a Black man held at gunpoint in the middle of a parking lot, potentially undermining faith in the police. In the second incident, one officer tried a less confrontational approach in spite of an agitated individual and, this time, it worked. In the Force Review Board meetings, we observed additional instances when the failure to recognize or address the mental health status of people led to unnecessary force. In one instance, a large Black man was in crisis at a party. He had cut his bare feet by walking through glass while in distress. Aurora Fire and Police responded but did not coordinate their response to the scene, so many Aurora Fire and Police members stood idly around Guiding Principles on Use of Force, Police Executive Research Forum, 2016. In reviewing other use-offorce reports, we often saw officers referring to the outdated “21 foot rule” that claimed an individual with a knife can cover 21 feet and harm an officer before an officer can respond, justifying the use of force against someone who was within 21 feet. The Police Executive Research Forum called this an “outdated concept” that conflicts with required concepts of proportionality and critical decision-making. 294 The officer’s name is changed consistent with our practice in this report. 293 74 the scene. The man remained agitated after Aurora Fire treated his feet, and after becoming angry again he walked up and struck a Fire paramedic. The officers then tased him multiple times. Though the taser use was likely consistent with policy if examined from the time when the man struck the paramedic, the whole situation never had a coordinated approach focused on avoiding escalation and recognizing the man’s volatile state. Such a strategy would have avoided putting so many personnel near him and reduced the chance that officers would use force. We asked several members of Aurora Fire and Police on our ride-alongs about any joint training or efforts to coordinate control on scene and learned that Aurora Fire and Police do not train jointly on such on-scene coordination. Rather, personal relationships that members of Aurora Fire and Police have developed on past calls often prompt coordination. Aurora Police issued a new policy, dated Sept. 7, 2021, that addresses how officers should coordinate with emergency medical providers. It prioritizes patient care over policing, explains what information officers should give to assist paramedics, and formalizes a transfer of care process.295 While the new policy ensures that the ranking officers of both units make face-to-face contact to establish unified command for large scale events, small scale events do not carry the same expectation. The majority of interactions between Aurora Police and Aurora Fire are small scale events, like the one above, and would benefit from face-to-face coordination between agencies. We witnessed on ride-alongs, however, that it was often difficult even for officers to know who was responsible for scene control, especially when numerous officers responded. As a higher ranking officer appeared, control would theoretically shift to the new arrival, but that officer may not understand the complexities of the scene yet. Without direction and training on how to coordinate scene control during small scale events, the new policy will not address these concerns. In another incident discussed by the Force Review Board, a Black man was walking down the street after a snowstorm hitting parked cars with a snow shovel. Officers responded and at first did a good job of backing up and keeping distance between the man and themselves rather than using force. But when the man started to walk away from the officers against their command, one officer tackled him into the street and handcuffed him. The man was on foot, posed no serious threat, and was in an obvious mental health crisis. Rather than follow the man or see whether he would comply with the officers’ instructions given more time, the officers chose to tackle him, violating the legal requirement that they apply nonviolent means, when possible, before resorting to force. In a final example, officers responded on a welfare check to a family dispute. They spoke to the mother outside the home; she had called about her son, a Black man, who was inside the home alone and throwing things. The man came outside, spoke to officers, and voluntarily let them pat him down. But as the questioning continued, the man became increasingly agitated and tried to go back inside the house. Officers refused to let him leave and ultimately took him to the ground, punched him in the head twice, and handcuffed him. The reports written by the involved officers did not articulate any potential crime that they were investigating or a reason why they needed to detain him, but the prolonged 295 Aurora Police Department Directive 09.06 (effective Sept. 7, 2021). 75 detention led directly to the uses of force. And because the initial call was to conduct a welfare check, not to investigate criminal activity, the officers’ escalation of the situation which precipitated the use of force was particularly unnecessary. In all but one of these examples we saw officers respond to people having mental health crises with significant, and based on information available to us, excessive force rather than first trying less-confrontational methods as the law requires. The officers did not attempt to de-escalate the person in crisis, and in several situations, their tactics escalated the situation, creating the need to use force. We do not know if the officers could have resolved the situation with less force because they never tried. 8.3.2. Using Force When the Only Criminal Violation is Failure to Obey The justification for many of the arrest and use-of-force incidents we analyzed was an alleged violation of Section 5 of Aurora’s Disorderly Conduct Ordinance, which outlaws “[f]ail[ure] to obey a lawful order or command by a peace officer, firefighter, marshal, or detention officer acting under the color of official authority which causes or is likely to cause harm or a serious inconvenience.”296 A resident who has not otherwise committed any crime can be arrested, and subject to use of force, if he or she does not obey a police officer’s order and the officer decides such disobedience causes “serious inconvenience.” Without a policy limiting how Aurora Police may use this ordinance, it provides officers the ability to justify arrest, and force, when no basis would otherwise exist. Based on information available to us, we conclude that Aurora Police has a pattern and practice of using objectively unreasonable force to arrest those who police claim have violated this ordinance. We heard from community members that officers used this ordinance to escalate situations and arrest, and use force against, residents. We also observed discussions in the Force Review Board where members commented that officers should have given an order sooner in the encounter so that they would have a lawful basis to use force more quickly. And other departments have faced scrutiny for relying on similar ordinances that give broad discretion to justify arrests and force.297 Less than a year ago, in December 2020, the ACLU announced a $285,000 settlement with Aurora arising from Aurora Police’s 2016 use of force against a resident, who was arrested for failure to obey a lawful order, resisting arrest, and disturbing the peace.298 According to the lawsuit, Aurora Police officers responded to a noise complaint and ordered the man to come out of his garage. When he did not immediately exit his garage (he said he was City of Aurora Ordinance § 94-110(5) (emphasis added). See U.S. Department of Justice, Investigation of the Baltimore City Police Department at 36-39; 55-58, Aug. 10, 2016; U.S. Department of Justice, Investigation of the Ferguson Police Department at 19-22, March 4, 2015 298 See ACLU Settles Case With Aurora After Police Brutalize and Unlawfully Arrest Alberto Torres, ACLU, Dec. 10, 2020. For the original complaint filed in Colorado federal court, see here. 296 297 76 calling to his wife who spoke better English), officers brandished a firearm, grabbed him, threw him to the ground, and handcuffed him, causing serious injuries.299 As part of our review of documents, we saw several use-of-force reports where the only apparent underlying justification for escalation was the suspect’s failure to obey a lawful order. Several examples are below. In one incident an officer was responding to a report that a security guard saw a car back into another car while trying to park, drive away, and then return to pull into the same parking space front-first. The car had dark tinted windows, and the officer believed it was empty. But as he approached the car, a man opened the driver’s door, “which caught [the officer] by surprise.” The officer drew his weapon, pointed it at the man, and ordered him out of the car. The man obeyed and got out of the car. The officer then ordered the man to get on the ground. The man said that “he was not going to get on the ground” and instead, walked to the back of the car and placed his hands on top of the trunk. Although the man had substantially complied with the officer’s orders and was not resisting arrest, the officer “performed an arm-bar take down on him and took him to the ground.” The officer said he used force “[d]ue to the male not obeying my lawful order.” A supervisor’s review of the incident stated that the “dark, crime ridden area, and the party startling him” authorized the officer to use force against the man who “was not being cooperative” and to “prevent him from escalating his behavior.” But the officer did not report any concern about the man’s behavior or the possibility that he might resist arrest. And resistance was unlikely because the startled officer still had his gun pointed at the man, and another officer was on scene to assist if needed. When the man’s wife filed a complaint, another round of review found that the responding officers “were calm and professional during this incident,” and that the use of force was “reasonable and appropriate for the circumstances.” From our perspective, however, the use of force appears to have been retaliatory. The car had been in a fender-bender, but was not otherwise connected to criminal activity. Suddenly opening the car door, while surprising, was not obviously such dangerous conduct to justify taking down a resident who was substantially complying with officer requests. Once out of the car, the man signaled his willingness to comply by placing his hands on the trunk where the officer could see them so he could be safely searched. Because the officer could have searched the man without force, throwing the man onto the ground was an unnecessary use of force and appears retaliatory for “not obeying.” The following is a portion of the narrative from a Tier 1 use-of-force report for an incident involving a Black female on January 16, 2018. The report states that the use of force was “necessary to effect arrest” and “necessary to prevent a crime.”300 On January 16th, 2018 [Officer 1] was walking through the lobby of Headquarters on her way to Municipal Court when she observed a female ACLU Colorado keeps a list of the more severe alleged racially biased uses of force by Aurora Police on its website. See here. 300 Incident-Report Case Summary, Tier 1: 2018-TIER-0025. 299 77 that appeared to be sleeping in the entryway on one of the benches. This area is out of view from the front desk but in direct view of anyone entering the building. Based on the female’s dress and position [Officer 1] contact[ed] the party to see what business the person had in the building and if they needed assistance. The female indicated she was waiting for her brother but was not willing to answer any other questions. She had verbally identified herself as [Person 1] which [Officer 1] believed to be a fictitious name. The female was not able to provide any identification or an explanation as to what name she gave. The female attempted to leave and [Officer 1] detained the suspect. The suspect requested a supervisor and again attempted to leave at which point [Officer 1] requested a cover car and supervisor. [Officer 2] who was at the Municipal Court house arrived to assist. The female again attempted to leave and was advised to sit down. Because the female was not listening to lawful orders [Officer 2] grabbed her arm and assisted her in a seated position. The female then started pulling away from [Officer 2] and swinging her arms in the direction of [Officer 1]. The female was taken to the ground, where she continued to scream and kick her legs.301 The supervisor who first reviewed the incident concluded that “the use of force in this situation was reasonable and appropriate to detain a suspect while attempting to leave and giving a false name” and that “[Person 1’s] actions of pulling away and swinging her arms ‘wildly’ posed a danger of injury to the officers and placing her on the ground to arrest her [was] appropriate.”302 Later reviewers all agreed the use of force was “necessary” and “appropriate.” From our perspective, there was no criminal activity before Officer 1 interacted with Person 1. The report does not state why the officer believed the name was false. There was no basis for officers to detain Person 1, much less put hands on her. But the report stated that because she failed to obey lawful orders, officers believed they had justification to detain, arrest, and use force against her. Below are portions of another narrative from a Tier 1 use-of-force report for an incident on December 1, 2017, involving a white male. This was six weeks before the incident discussed above. The report states that the use of force was “necessary for subjects’ safety.”303 I spoke with [Officer A] who advised me that she contacted the male who was lying down on the grass and at first was not responsive. She said when Id. Id. 303 Incident Report Case Summary, Tier 1: 2017-TIER-0432. 301 302 78 he did respond she told him to get his hand out of his pocket and he complied and he also complied with handing her his identification. ... She advised the subject was cooperative until she began to clear him at which point he began to get nervous and began to walk away. Herself and [Officer B] told the male to stop and he did not comply with their orders to stop. She advised they grabbed the male by the arms to get him to stop. She then said the male went to the ground and that while on the ground she had ended up with his right wrist in a wristlock.304 The supervisor who first reviewed the incident concluded that both officers’ actions were “reasonable and appropriate,” noting that “[o]nce [Person A] found out he was being cleared for warrants he became more nervous and in a hurry to leave (even ready to leave his ID with officers).”305 Later reviewers agreed the use of force was “lawful, reasonable, and within APD policy.”306 From our perspective, there was no basis for detention or any use of force. Yet, simply because he failed to obey an order, officers believed they had a justification to detain and use force. But unlike the other incident described above, however, the suspect was not arrested and was allowed to leave, showing the considerable discretion officers have in such situations. As part of the investigation, the team spoke with Doug Wilson, the Chief Public Defender for the City of Aurora. He told us that many clients his office defends were charged only with failure to obey a lawful order—and that often the prosecutor’s office ultimately drops the case shortly before, or on the day of, trial. Data provided by the Court Administrator for the City of Aurora show just how often Aurora Police charges residents with violating § 94-110(5):307 Year Count 2015 604 2016 569 2017 493 2018 390 2019 406 2020 264 2021 (through August) 118 Id. This portion of the narrative was based on discussions with Officer B. Id. 306 Id. 307 Aug 23, 2021 Email from Court Administrator Shawn Day to Deputy Attorney General Janet Drake. 304 305 79 Aurora Police officers’ over-reliance on “failure to obey a lawful order” to justify arrest and uses of force must change. Aurora Police must adopt policies that carefully limit using force against an individual where the only alleged crime is failure to obey a lawful order unless that failure creates other risks that independently justify force. 8.4. Causes of Violations Based on our review of thousands of use-of-force reports, our interviews with the community and members of Aurora Police, our ride-alongs, and our observation of the Force Review Board, we find the following causes lead to Aurora Police’s use of excessive force. 8.4.1. A Culture at Aurora Police That Emphasizes Justification for Force, Rather Than Whether Force Was Lawful and Appropriate We observed that Aurora Police generally approached force with a focus on why force was justified, rather than what force is appropriate. Our review of thousands of use-of-force reports found that the reports often emphasized what conduct or perceptions justified the force, rather than describing why the officer thought that level of force was objectively reasonable and appropriate. The Force Review Board discussions routinely focused on finding reasons why the force could be justified, rather than whether the force met legal standards. Almost no discussion occurred about Colorado law’s requirement that officers shall apply nonviolent means when possible before using force. We concluded that officers often focus on control and submission when engaging with individuals. As described above, this focus led to arrests and force used for failure to obey. This emphasis on control also led to encounters escalating quickly, particularly with those who were not native English speakers, were impaired, or in a mental health crisis. While we observed some encounters handled with patience and empathy, other encounters began with officers behaving aggressively and quickly escalating to force. We saw a pattern of officers “taking subjects to the ground” quickly, often for minor reasons. We found this behavior arises out of a perceived need to immediately take control of situations and establish dominance in interactions with members of the community. Indeed, in the Force Review Board meetings, we heard several comments about how officers should have given orders earlier to individuals so that when they violated those orders they would have legal justification to arrest or take them to the ground. As the data discussion reveals, this type of force is used with more frequency against non-white members of the community. Because Aurora Police uses this tactic so often, we cannot determine whether engaging with the community with a less-confrontational approach would lead to fewer uses of force. We also saw a pattern of officers reciting “stop resisting” reflexively during encounters even when the body-worn camera videos did not show resistance. But officers did not use specific language to describe why they felt the subject was resisting and the body-worn cameras did not show any resistance. This purported resistance was often used to justify force, both in the moment and later in the Force Review Board discussion. Because 80 officers used this generic language, we often could not determine whether specific, articulable reasons supported the officers’ use of force. In addition, when using pain compliance techniques to control individuals, officers often treated the individuals’ expected pain response as active—not involuntary—resistance, to justify the use of even greater force. We also saw some body-worn camera video where officers used language that apparently was designed to support an excited delirium diagnosis and potential chemical sedative use by Aurora Fire. These officers used words like “superhuman strength” or “he’s jacked up” during the encounter. As we discuss below, the decision of whether to use chemical sedatives, even under old policies, lies with the medical professionals, not the officers. Attempts by officers to narrate their encounter using catchphrases that appear to be designed to influence a medical diagnosis, as opposed to describing the actual behavior of the subject, are improper. This focus on justification rather than appropriateness of force also leads to inconsistent levels of force used for similar circumstances. For example, we saw several instances of officers engaging with a sleeping person in a car reported stolen. Because the Force Review Board did not review instances when force was not used, we did not see as many instances when officers addressed the situation without using force. But where officers chose to use force, they sometimes broke glass on one side or the other; often used tasers and pointed weapons from different vantage points; and another time used pepper spray. But a consistent pattern emerged from these encounters—officers acted differently in each case with generally low levels of coordination, acted quickly when no need to do so was present, and escalated the situation often resulting in injury to themselves and the individuals. In one instance, the pepper spray surprised the person sleeping and caused him to drive away and, because of pepper-spray-impaired vision, he sideswiped several police cars and collided with another car in an intersection. By contrast, we saw a similar instance when Aurora Fire, rather than Aurora Police, took the lead on interacting with a sleeping person in a car. Despite circumstances which had resulted in Aurora Police’s use of force in other similar situations—the car was running and the subject did not put it into park immediately—no one used force and the subject, after reacting with surprise, quickly calmed down and discussed the situation with Aurora Fire and Police. Finally, because the Force Review Board generally focuses on a single question—Could the force be justified under the policy?—the Board does not serve as a mechanism for improving the operations of Aurora Police or addressing close calls. And because the consequence for a policy violation is seen as severe—a referral to Internal Affairs—the Board is generally reluctant to find policy violations. When the Board finds poor decisionmaking but no policy violation, no structured mechanism exists for Aurora Police to improve—either through policy revision, training enhancements, or feedback to particular officers. For example, we observed several incidents during the Force Review Board meetings and on ride-alongs where officers responded to domestic violence calls and failed to separate the victim and suspect. In a recent incident, officers let the victim lead the way into her home, where the suspect was. Officers believed that the suspect possibly had a gun and yet there was no effort made to keep the victim away from the suspect. Instead, the victim 81 remained present for much of the encounter, even as the suspect was backed into a corner appearing to conceal something behind his back. The Force Review Board identified this as a training issue—saying that the officers never should have let her into the house—but there was no further discussion about follow-up on how to address this issue. Of course, situations like these present safety concerns for the officers, victim, and subject, often heightening emotions, escalating the urgency of the situation, and requiring the application of force that may not have otherwise been necessary. And the Force Review Board identified several other recurring tactical issues from their use-of-force reviews—officers getting too close to subjects suspected to have guns or other weapons; one officer drawing a taser and gun at the same time; sergeants on scene failing to either coordinate the officer response or take control of the situation; and officers deploying a taser when pressing it against a subject, often called drive-stunning, when a subject was already on the ground, leading to increased resistance by the subjects and often increased force by the officers, among other issues. Force Review Board and Force Investigation Unit members reported that they would follow up with officers and supervisors on an ad hoc basis but no structure or criteria guided when such feedback should occur. Nor did we observe any follow-up to confirm that it had occurred. Nor did we observe any accountability mechanisms that ensured that others in Aurora Police heard of the issues identified in the Force Review Board. Training Academy personnel confirmed that they do not collaborate with the Force Review Board, even though recurring issues identified by the Force Review Board are necessarily those that officers need to be trained on. Without having a systematic way to ensure that issues identified in these types of reviews return to the organization in a structured manner— through formal training (both at the academy and in-service), policy review, and in shift briefings—Aurora Police will continue to repeat mistakes. The recent development of the Force Investigation Unit and the development of a review checklist for use-of-force incidents show movement in the right direction. In our limited observation, the review checklist provides more structure to the Force Review Board discussion and increases the chances that more thoughtful discussions occur. But we still observed situations when the Force Review Board discussion was perfunctory, failed to follow the checklist, and focused on condoning poor decision-making by officers. It is unclear that the Force Review Board intends to have members discuss each element of the checklist on a consistent basis going forward. And we continued to see an emphasis on justification rather than need, and no systematic effort to utilize use-of-force reviews to improve training or tactics. How the Force Review Board handled the recent high-profile case where two officers were charged with crimes from the pistol whipping of a resident by one officer and a failure to intervene by another officer illustrates the residual profound shortcomings of Aurora’s approach to reviewing force. In the Force Review Board meeting, many members stated that pistol whipping a subject was not necessarily outside of policy, applied the wrong legal standard to evaluate the use of force, and spent significant time trying to justify the force used. The Force Investigation Unit’s presentation minimized the seriousness of the incident, describing the repeated pistol whipping as the officer “jab[bing] [the resident] in the face a couple of times with the barrel of his weapon” and characterizing the interaction 82 as a “struggle” and “brawl,” despite there being no evidence of such events in the video evidence reviewed. And none of the Force Review Board members mentioned the officer choking the resident. Instead, members discussed their reticence to have a “knee-jerk reaction” and determine whether the officer’s uses of force were compliant with policy without knowing his perspective. 8.4.2. Use-of-Force Policy and Implementation Focuses on Maximum Force Permitted Under Law Aurora’s policy does not contain more specific guidance to structure officer decisionmaking when using force like policies from other Colorado police departments. Officers encounter many challenging situations. Equipping them with a more detailed framework on when to use force and what force to use helps them do their jobs well and reduces the chance that their split-second decisions will result in excessive force. 8.4.3. De-Escalation is Misunderstood and Not Consistently Practiced Aurora Police often views de-escalation as what happens after force is used, rather than an alternative to using force in the first place. Repeatedly in the Force Review Board meetings, members applauded de-escalation that occurred after officers had tased or tackled someone, rather than focusing on whether the taser or tackle was necessary in the first instance. Deescalation, properly understood, focuses on tactics that reduce the need for force in the first place, rather than decreasing the amount of force used after the fact. In addition, at scenes with multiple officers, we often witnessed a lack of coordination, with the most aggressive officer often setting the tone for use of force. We saw, for example, officers • use pepper spray against a sleeping individual in a stolen car, • raise long guns for close encounters where the need for force had not been established, • tackle someone who posed no immediate threat, • encourage a police dog to continue to bite well after an individual had been taken to the ground, • fire two less-lethal shotgun rounds from close range at an intoxicated individual who was not complying with commands because he did not understand English, • quickly draw a gun on an individual producing paperwork during a traffic stop which ultimately led to a multi-officer takedown and taser deployment, • engage in a foot chase and tasing of a person who stole $30 worth of goods from a store, and 83 • immediately draw weapons when dealing with an individual in a mental health crisis. Many of these incidents occurred despite a lack of urgency and even though officers had time to coordinate and attempt engagement without using force. Aurora Police officers’ frequent practice of having all officers on scene participate in engagement with the subject leads to this lack of coordination reducing the ability to deescalate the situation. For example, we observed all available officers respond to a domestic violence call where the subject was outside the home, yelling and disruptive but not physically aggressive. At first, there were a handful of officers on scene. The man was verbally aggressive, but one officer approached and spoke with him, and he began to calm down. The man appeared ready to leave the scene, which would have ended the encounter. But as more and more officers arrived, the man became agitated and started yelling again. This was not an isolated incident. Rather than have one or two officers first engage with an individual, Aurora Police often has all officers on scene engage, including drawing their weapons, instead of first assessing whether less aggressive shows of force may lead to a calmer situation without the need for force. As other examples, we saw several incidents in the Force Review Board where foot pursuits led to the use of tasers. Many times interactions quickly escalated from a routine engagement and led to the use of force when alternatives to using force were not given a chance. For example, we saw officers begin a foot pursuit that led to the use of a taser after an individual had shoplifted an energy drink from a big box store and poor decisionmaking by store personnel allowed the individual to escape. Other instances of foot pursuits leading to taser use arose from routine encounters that escalated when individuals did not want to interact with police and fled. There was often little connection between the basis for the stop and the pursuit leading to the use of a taser. In general, we observed that Aurora Police officers were quick to use tasers to address challenging situations. For example, we saw an incident where an officer tried to tase a fleeing suspect who had climbed over a fence. The officer missed, but even if the taser had been effective the suspect would have been on the other side of the fence, too far away for the officer to handcuff him while he was subdued by the taser. While Aurora Police policy describes tasers as less-lethal force, they still pose significant risk to individuals and, as we saw, often do not work effectively in certain situations, including in the winter when people wear heavy coats. The use of a taser escalates the situation dramatically, often foreclosing other engagement requiring less force. In addition, tasers can cause serious injury or death even when used appropriately.308 Although Aurora Police Department Directive 5.08 does provide more guidance on the use of tasers, what we saw in practice was a pattern of quickly resorting to threatened and actual taser use instead of, as the policy requires, “exhaust[ing] all available non-physical means.” We also observed that officers were often quick to escalate and use dangerous and unnecessary tactics that led to unnecessary force. Aurora needs to change its approach to Lethal Force? Tasers Are Meant to Save Lives, Yet Hundreds Die After Their Use By Police, USA Today, Apr. 23, 2021. 308 84 civilian encounters by using force as a last resort, not as the immediate option, as state law requires.309 While the Force Review Board would often identify poor tactics in their meetings, that information was not passed along in a standardized manner to the officers, their supervisors, or the trainers. In many cases, that information was not passed along at all. When the Academy trained recruits on de-escalation tactics, such as returning to their patrol car and waiting for backup if a gun was visible during a traffic stop, that training was not passed along to the Field Training Officers, who then misunderstood the new tactic and did not reinforce it. It is not enough to identify a concerning tactic as a training need; officers must then be given the training to understand why the tactic should be avoided and what to replace it with. Aurora Police lacks a systematic method to flag bad tactics that lead to unnecessary force, and then ensure that leadership highlights tactics to be avoided in performance reviews, shift briefings, and in-service trainings. Aurora Police also lacks a systematic method to meaningfully engage and train experienced officers on new deescalation tactics and their value. Without such clear direction, officers will continue to use tactics that lead to unnecessary force. 8.4.4. Mental Health Issues Are Not Adequately Addressed Aurora Police inappropriately responds to those in mental health crisis. Officers are not adequately trained and often do not consider the mental health status of people they engage with. Some officers have Crisis Intervention Training, a week-long course on interacting with people facing mental health challenges, but other officers do not empower those trained officers to take the lead in situations involving mental health crisis. On our ride-alongs, officers told us that Crisis Intervention Training has a bad reputation within Aurora Police, and some officers do not think it is effective. The use-of-force policy does not emphasize de-escalation or explain that officers should consider a person’s mental capacity in their decision-making. If officers are unconvinced that crisis intervention techniques play an important role in de-escalating a person in crisis, they are unlikely to use them. And if policy does not emphasize the need to attempt de-escalation, officers are unlikely to allow CIT or similarly trained officers to use other tactics that often reduce the likelihood that force will be used. It is unsurprising, then, that based on our observations, the Crisis Response Teams seldom take the lead in encounters involving mental health issues. 8.4.5. Force is Used Disproportionately Against the Black Community Our observation of the Force Review Board noted that the targets of force by Aurora Police were often individuals of color. Of course, because the Force Review Board only reviewed incidents when force was used, our observations do not substitute for the robust data analysis above. But the pattern was still striking. As we discuss above, Aurora uses force disproportionately against Black residents. § 18-1-707(1), C.R.S. (requiring officers to “apply nonviolent means, when possible, before resorting to the use of physical force”). 309 85 8.4.6. Aurora Police Inadequately Documents the Use of Force Aurora Police does not prepare or keep records that allow for review to understand why force was used in a given circumstance. We saw a pattern of using conclusory, boilerplate language to describe why force was used, rather than the required specific articulation of the reason the officer actually decided to use force and why other approaches to engagement were not used as required by Colorado law. Without this detail, it is difficult for supervisors or others to evaluate the appropriateness of the use of force. 8.4.7. Aurora Fire and Aurora Police Do Not Adequately Coordinate In both our Force Review Board observations and our ride-alongs, we observed how Aurora Fire and Aurora Police interacted with each other. Generally, we saw poor scene coordination with Aurora Fire and observed several encounters where there was too little communication between Fire and Police personnel. At times, this lack of communication and coordination led to unnecessary force. And we learned that Aurora Fire and Police do not conduct joint training exercises that would allow them to practice interagency communications. Some officers and fire personnel talk to each other well, effectively communicating important information in high-stress environments. Other times, the second agency to arrive will waste time gathering information from civilians on scene rather than learning from the first agency. We observed no standardization of this communication and no opportunity for teams to debrief or discuss how to improve communication or coordination in the future. The new policy adopted by Aurora Police does not address these concerns. Finally, we saw a pattern of no specific, identifiable point when someone transitioned from a subject of police investigation to a patient of Aurora Fire. While we generally observed cooperation and professional engagement, we also often observed confusion on scene as to how to handle unexpected developments that arose. Because of this confusing transition, we observed, at times, delay in the provision of medical care to individuals. The new coordination policy prioritizes medical care over custody concerns and requires officers to facilitate prompt patient care. This is an improvement. But fire personnel (including those from Falck) need to understand how Aurora Police are trained to try to de-escalate a situation and must not hinder those efforts. Aurora Police, in turn, need to train with Aurora Fire to support providing prompt medical attention to those in need. Also, we observed one instance of an exceptional after-action debriefing by Aurora Fire with members of an emergency room department after a single-car fatal accident. There, paramedics from Aurora Fire engaged in extensive discussion with members of the Emergency Department at the Medical Center of Aurora over how care was provided on scene and how care could be improved to increase the chance of survival. This discussion stood out because in our many ride-alongs, where we observed many scenes where Aurora Fire and Police both responded, we saw no similar after-action discussion about how to learn from recent experiences. Many of these situations involved high-stakes, joint responses, such as responding to three different individuals in mental health crisis threatening a household member with a knife, dealing with a drunk individual who was reluctant to receive needed medical care, and responding to domestic violence incidents. 86 Aurora Fire and Police are partners in supporting the community, and they should train and better support each other in doing so. These shortcomings in how Aurora Fire and Police train and work together are particularly surprising given their shared use of a joint training center that was developed, in part, to improve how the two organizations work together. That facility went into service in early 2016, but does not appear to have supported a culture of joint training or coordination. During our visit to the training center, police officers said that they have very little interaction with fire personnel despite the shared space. The Training Academy plans to incorporate joint Aurora Police and Fire training into academies in the future, but there currently is no plan to implement coordinated in-service training for existing officers. 8.5. Changes Needed Based on our investigation, Aurora Police must make the following changes to ensure its use of force follows the law. 8.5.1. Improve Use-of-Force Policies to Give More Specific Guidance to Support Officers Aurora Police must change its policy to provide more specific guidance to officers as to when they may use force and what they must do before using all types of force. The policies must require efforts to use alternatives to force whenever feasible. The policies must focus on the entire encounter with the officer and information provided by dispatch or bystanders, not just the immediate decision to use force. The policies must require officers to have specific, articulable reasons for using force, and document those reasons in the required reports. Examples of appropriate de-escalation and use-of-force policies include the Baltimore310 policy, as well as policies from Boulder,311 Colorado Springs,312 and Denver.313 8.5.2. Enhance Training on New Policies Aurora Police must comprehensively train all of its officers under this new policy, both at the academy and with in-service training. This training must include scenario-based training that addresses a mental health crisis, critical decision-making, de-escalation, and the likely use-of-force consequences of decisions like foot pursuits. In addition, Aurora Police must substantially expand training that addresses bias. Baltimore Police Department Policies. Boulder Police Department Policies & Procedures. 312 Colorado Springs Police Department Policies & Procedures. 313 Denver Police Department Operations Manual. 310 311 87 8.5.3. Change the Focus and Purpose of the Force Review Board and the Force Investigation Unit The Force Review Board and Force Investigation Unit must review whether force was necessary and legal, not just whether force could be justified in hindsight. The Board and Unit must change its approach to reflect the values and requirements of the new policies. 8.5.4. Develop Measurable Goals to Improve How Officers Engage With Those Experiencing Mental Health Crises Aurora Police must set and meet specific, measurable goals on improving how officers engage with those experiencing mental health crises beyond just making policy changes. Options include significantly increasing officer participation in Crisis Intervention Training; training officers to routinely request support from colleagues who have received Crisis Intervention Training on calls with a mental health crisis; creating specific practices to empower Crisis Response Teams, as opposed to officers, to lead on calls with a mental health crisis; and similar efforts. 8.5.5. Develop a Joint Policy for Aurora Police and Fire to Coordinate on Scenes and Conduct Joint Trainings Aurora Police and Fire must develop and implement a joint policy for scene coordination and conduct joint trainings so that, in particular, Aurora Fire personnel know and understand Aurora Police’s approach to de-escalation and Aurora Police officers know and understand Aurora Fire’s approach to patient care. Aurora Police issued a new policy on patient care and transfer of custody that addresses some of the concerns raised in this report. However, both agencies must adopt a more comprehensive policy and appropriate training. 88 9. Finding #3: Aurora Police Has a Pattern and Practice of Failing to Document Stops as Required by Law We found that under policies that have been in effect since 2020, members of Aurora Police conduct resident stops without documenting those stops as Colorado law requires. This failure prevents Aurora Police from adequately supervising its personnel and contributes to violations of the U.S. Constitution and Colorado law as discussed throughout this report. 9.1. Legal Background Senate Bill 217 requires that officers have a legal basis for making any contact with an individual, consensual or not, when they are conducting law enforcement or investigatory activities. Whenever such contact is made, Colorado law requires law enforcement to keep detailed records of the interaction, including: (a) the perceived demographic information of the person contacted; (b) whether the contact was a traffic stop; (c) the time, date, and location of the contact; (d) the duration of the contact; (e) the reason for the contact; (f) the suspected crime; (g) the result of the contact; and (h) the actions the officer took during the contact. Subsections (g) and (h) require more information, such as whether the subject was released with no action taken, whether the officer sought and was provided consent to search, or whether a weapon was unholstered during the encounter.314 These documentation requirements deter unnecessary and improper stops by requiring officers to articulate their motives and describe their actions for every investigatory or enforcement encounter with a member of the public. They allow for review by supervisors and the state. And applying these requirements not just to formal arrests but also to investigative-type Terry stops and other less formal interactions provides critical oversight for the types of police activity that have been historically under-scrutinized and subject to abuse. 9.2. Aurora Police Fails to Document All Stops as Required by Senate Bill 217 Senate Bill 217’s expanded recordkeeping requirements took effect when the law was enacted in June 2020, yet we found that Aurora Police failed to follow this law and record this information about all stops. Multiple members of our team witnessed incidents during our ride-alongs where non-white residents were detained, and even handcuffed, before being released with no charges filed. During those incidents, we saw no effort made to properly record the information required by Colorado law. And we confirmed with supervisors and senior leadership that such data was not recorded and that there was no way to determine how many other instances, beyond those that we personally observed, occurred where people of color were detained, handcuffed, and then ultimately released. 314 § 24-31-309(3.5), C.R.S. 89 This failure causes an entire category of Aurora Police activity (i.e., interactions that do not lead to a formal charge being filed, such as Terry stops where the subject is ultimately released) going unrecorded contrary to state law. We therefore conclude that Aurora Police has a pattern and practice of violating the recordkeeping requirements of Colorado law, which are contained at Colo. Rev. Stat. § 2431-309(3.5). Aurora Police must implement new procedures to ensure it is capturing all the information required by Senate Bill 217 and must likewise begin training to inform its officers of these requirements. Understanding who Aurora Police interacts with and why provides the basic information necessary to determine whether Aurora is following the law. This data collection allows prompt analysis to discover patterns of police misconduct and proactively address breakdowns in community policing. When departments do not record data, or keep it siloed and disconnected, leadership cannot identify trends in unlawful practices, individual officers or units who bear a disproportionate share of responsibility for illegal practices go unnoticed, Aurora Police cannot identify real training needs, and illegal conduct and practices remain harder to identify. 9.3. Aurora Police Reports Do Not Include Required Detail Describing the Reasons Supporting the Contact or Stop Even in the instances when a written report is filed, Aurora Police officers frequently fail to adequately describe the reasons supporting the contact or stop. Aurora Police Department Directive 5.4 addresses the reporting requirements in cases involving a use of force. Although sections 5.4.4 through 5.4.6 discuss the reporting responsibilities of officers and their supervisors, there is no reference to what officers should record if a resident stop involves or requires a use of force. Although section 5.4.4 directs an officer to “file a written report detailing the justification for the use of such force, type of force used, resulting effect of the force used and subsequent actions taken by the member,” the policy fails to address the information that should be recorded where force is used after or in connection with a stop—such as the specific articulable facts (reasonable suspicion) that justified the resident contact in the first place. We saw several examples of General Offense reports completed by an officer that failed to describe a sufficient basis for the initial contact or continued detention—for example, an officer wrote in one report that his reason for contacting and detaining a Black woman was that she “appeared to be homeless” and refused to answer the officer’s questions. Other examples of deficient reports include vague boilerplate language explaining the use of force, such as “I observed [the suspect] clench up and become more aggressive”, “I could feel [the suspect’s] body tense. [Suspect’s] body was stiff as a board and I needed to place [suspect] into a rear escort wrist lock and placed [suspect] in handcuffs for my safety and the safety of fellow officers”, and “[d]uring the contact, [suspect] began to resist and fail to follow instructions.” We also saw General Offense reports with limited detail that describe the report as a “summary” and direct the reader to the body-worn camera footage “for more information.” 90 The identified reporting deficiencies reflect a lack of officer accountability or failure by supervisors to ensure officer reports are complete and include sufficient detail. Aurora Police Directive 5.4.6 addresses the responsibilities of supervising officers to ensure that reports “contain a description of events leading up to, during, and after the use of force, which are relative to the use of force.” Despite the language in this directive, reports are often completed without sufficient detail, which reflects a lack of critical review by supervising officers. This lack of critical review undermines Aurora Police’s ability to ensure officer accountability to conduct stops and arrests that comply with applicable constitutional and statutory requirements. 9.4. Changes Needed 9.4.1. Develop Policies and Create Systems to Comply With Colorado Law on Documenting Stops Aurora Police must immediately develop policies and systems that comply with Colorado law on documenting encounters with the public. These systems should form part of a robust and functional data dashboard that allows for real-time analysis of this information. 9.4.2. Amend Aurora Police Policies to Include a Separate Directive Addressing Stops That Complies With both Colorado and Federal Law The directive should include at least the following provisions: • A clear and detailed statement of the key requirements of the Terry decision and Colorado analogues and what those decisions mean, including the legal differences between an “investigatory stop,” “consensual encounter,” and arrest. • Guidance on what constitutes “reasonable suspicion” supporting a stop—with usable examples of how officers should articulate the specific factual basis supporting their belief that a crime has been or will be committed. This guidance must emphasize the importance of including detailed factual statements in reports supporting “reasonable suspicion” and prohibiting the use of vague, canned, or boilerplate language. In addition, this guidance must include a list of factors that may not be used in making stops. • A requirement for supervisors to review stop reports to ensure reports avoid vague, canned, or boilerplate language and to document deficiencies in a timely manner with documentation of that review. 91 10. Finding #4: Aurora Fire Rescue Has a Pattern and Practice of Administering Ketamine Illegally We found that Aurora Fire has a pattern and practice of administering ketamine illegally. Patients were routinely given ketamine in doses that exceeded the maximum allowed by protocol, and paramedics often failed to estimate a patient’s weight. Once sedated, some patients were not properly monitored, placing them at risk for life-threatening complications. In some instances, paramedics administered ketamine at the request of officers without completely assessing the patient’s condition and over the patient’s stated objections. Patient care reports were detailed when symptoms supported a diagnosis suggestive of ketamine, but were scant on details when combativeness was the primary complaint. Aurora Police often failed to document that paramedics sedated a person in their custody, so supervisors were unable to review whether sedation was requested and should be reviewed as a use of force. Finally, medical supervisors did not intervene when paramedics failed to follow agency protocols to prevent future violations. Despite reviewing every instance of ketamine administration, the medical directors did not flag a single case of excess dosing or failure to properly monitor a patient as triggering the need for additional training or remediation. Aurora Fire must improve its training, protocols, and case-review policies to prevent the unnecessary use of chemical sedatives. 10.1. Ketamine and Its Use by Aurora Paramedics 10.1.1. Ketamine Use by Paramedics Ketamine is a short-acting “dissociative anesthetic that has some hallucinogenic effects.”315 The Colorado Department of Public Health and Environment regulates the use of ketamine by paramedics outside the hospital setting, providing waivers to medical directors who deliver the appropriate training, oversight, and review for the emergency medical providers under their supervision.316 It announced a review of the ketamine waiver program for administration in patients with symptoms of excited delirium in August 2020 that is still ongoing. A suspension of all waivers was initiated in July of this year while it works to develop a program that complies with House Bill 1251.317 The waiver program had permitted paramedics to use ketamine to sedate patients with “a presumptive diagnosis of excited delirium” or “extreme or profound agitation.”318 Ketamine Drug Fact Sheet, Drug Enforcement Administration (2020). Press Release, State health department announces plans to review the ketamine waiver program, Colorado Department of Public Health and Environment (CDPHE) (Aug. 22, 2020, updated Apr. 15, 2021). 317 Colorado Suspends Ketamine Use on Agitated People as New Law Takes Effect, Fox 31 News (July 7, 2021). 318 Ketamine Waiver Guidance, Emergency Medical Practice Advisory Council, CDPHE; Non-medical personnel use “excited delirium” to describe a variety of agitated or violent behaviors, but the medical community defines it as “delirium with agitation (fear, panic, shouting, violence and hyperactivity), sudden 315 316 92 According to the waiver guidance, “[e]xcited delirium is [a rare] medical emergency in which a person develops extreme agitation, aggressiveness, overheating, and exceptional strength that cannot be managed by routine physical or medical techniques.”319 Other health organizations, such as the American Medical Association and American Psychiatric Association, do not recognize “excited delirium” as a diagnosis.320 Across the country, doctors had recognized a pattern of people who showed signs of excited delirium who then died suddenly in police custody.321 The individuals were excessively combative and resisted arrest with abnormal strength. A few individuals had pre-existing mental illness, but most were under the influence of stimulant drugs, especially cocaine and methamphetamine.322 Everyone died while restrained and lying down; many had both their hands and feet restrained behind their backs.323 By 2010, about 250 people died each year in custody while exhibiting signs of excited delirium.324 The spike in custodial deaths alarmed the medical and emergency responder community, which issued recommendations for “rapid sedation” to prevent the bizarre “multisystem failure.”325 The recommended drug was ketamine because it was fast-acting when “victims may not have minutes to spare as they continue to struggle against law enforcement or physical restraints in a state of hyperthermia and metabolic acidosis.”326 By 2017, a nationwide study revealed that nearly all paramedics had received training on ketamine and one-third of EMS agencies “allowed prehospital use of ketamine” or had protocols permitting its administration.327 Chemical restraint and sedation accounted for more than half of all ketamine use.328 The appearance of ketamine administration in Minneapolis police reports filed between 2010 and 2018 demonstrated its increased use from 0 to 2 incidents per year early in the decade to 62 incidents in 2017.329 cessation of struggle, respiratory arrest and death,” Asia Takeuchi, M.D., et al., Excited Delirium, 12 W. J. EMERG. MED. 1, 77 (2011). 319 Ketamine Waiver Guidance, Emergency Medical Practice Advisory Council, CDPHE. 320 Position Statement on Concerns About Use of the Term "Excited Delirium" and Appropriate Medical Management in Out-of-Hospital Contexts, American Psychiatric Association, Dec. 2020; New AMA Policy Opposes “Excited Delirium” Diagnosis, American Medical Association, June 14, 2021. 321 Michael S. Pollanen, Ph.D., et al, Unexpected death related to restraint for excited delirium: a retrospective study of deaths in police custody and in the community, 12 CANADIAN MED. ASSN J. 158, 1603-07 (June 1998). 322 Asia Takeuchi, M.D., et al., Excited Delirium, 12 W. J. EMERG. MED. 1, 77-78 (2011). 323 Pollanen, et al., at 1604. 324 Lt. Col. Edward L. Hughes, Special Panel Review of Excited Delirium, Institute for Non-Lethal Defense Technologies (Dec. 2011). 325 Excited Delirium, at 80. 326 Id. at 81. 327 Daniel M. Buckland, M.D., Ph.D., et al., Ketamine in the Prehospital Environment: A National Survey of Paramedics in the United States, 33 PREHOSPITAL AND DISASTER MEDICINE 23, 25 (Feb. 2018). 328 Id. at 26. The other uses for ketamine were primarily for therapeutic pain management. 329 MPD Involvement in Pre-Hospital Sedation (July 26, 2018), MINNEAPOLIS POLICE DEPARTMENT, at 10. 93 Falling methamphetamine and cocaine use in the past decade should have led to fewer cases of excited delirium and less need for ketamine.330 But the rise in ketamine use during the last half of the decade, especially in connection with police-involved patients, raised concerns that officers were asking for ketamine to help subdue suspects.331 10.1.2. Changes to Colorado Law on the Use of Ketamine In June 2021, Colorado lawmakers passed House Bill 21-1251, a bill preventing paramedics from giving ketamine to police-involved patients unless a justifiable medical emergency required its use.332 The law specifies that excited delirium is “not a justifiable medical emergency.”333 The new law, which became effective on July 6, 2021, prohibits police officers from directing paramedics to give ketamine to an individual.334 Should an officer make such a request, the bill includes mandatory reporting provisions for both police and emergency medical personnel. Police are also mandated to intervene and prevent officers from directing paramedics to administer ketamine to effect an arrest or “to facilitate ease and convenience in law enforcement encounters.”335 Officers are permitted, however, to provide paramedics and emergency medicine service providers with “information about the individual or the scene of the emergency that may assist … assessment of the need to administer ketamine.”336 But paramedics may not base their “medical decision or diagnosis exclusively on information provided by a peace officer.”337 In addition, House Bill 1251 outlines statewide regulations on ketamine administration in the presence of law enforcement. Before giving ketamine to a patient, paramedics must weigh the patient to ensure accurate dosage. If a patient cannot be weighed, paramedics must have three trained personnel estimate the patient’s weight and try to receive verbal permission from a medical director. Paramedics must be trained in ketamine administration and advanced airway support techniques, provide urgent transport to a hospital, and record any complications arising from the ketamine. The drug cannot be given unless there is proper monitoring and respiratory management equipment immediately available. Methamphetamine Research Report: What is the scope of methamphetamine misuse in the United States?, National Institute on Drug Abuse (Oct. 2019); Cocaine Research Report: What is the scope of cocaine use in the United States?, National Institute on Drug Abuse (May 2016). 331 Ketamine That’s Injected During Arrests Draws New Scrutiny, Associated Press, Aug. 22, 2020. 332 House Bill 21-1251 was passed on June 22, 2021. 333 § 25-3.5-103(8.6), C.R.S. 334 § 18-1-707(1.5), C.R.S. 335 § 18-8-805(3), C.R.S. 336 § 18-8-805(2)(c), C.R.S. 337 § 18-8-805(2)(b), C.R.S. 330 94 10.1.3. Aurora’s Suspension of Ketamine Use in 2020 The Aurora City Council suspended the use of ketamine for patients exhibiting excited delirium on September 14, 2020, pending the outcome of an independent investigation into the death of Elijah McClain.338 The Council’s suspension expired at the end of March, thirty days after the investigation report was released.339 Chief Gray shared his intention to keep the ban in place indefinitely. And the scope of practice waiver authorizing Aurora Fire to use ketamine expired in June 2021. Because ketamine was not authorized for use during our period of observation, we could not see first-hand how ketamine is used in the field. 10.1.4. Structure of Aurora Emergency Medical Services Falck Rocky Mountain contracts with the City of Aurora to provide ambulance services in Aurora. Both agencies might respond to an emergency, but all transport is conducted by Falck. Aurora Fire, however, is responsible for all emergency medical services in Aurora and drafts the emergency medical protocols used by both agencies. Aurora Fire and Falck share the same medical director, who submitted a scope of practice waiver in 2018 permitting both agencies to use ketamine. 10.2. Findings on Aurora’s Practices of Ketamine Use To evaluate the agencies’ practices, we reviewed Aurora Fire Rescue and Falck Rocky Mountain ketamine administration records which are provided to the Colorado Department of Public Health and Environment as required by the waiver program. We reviewed redacted patient care reports for these incidents and the few Aurora Police reports that identify or correspond with ketamine administration. We also spoke with the medical director. 10.2.1. Records of Past Ketamine Use Show Disparate Dosing of Ketamine Not Based on Weight and a Failure to Follow Monitoring Protocols The ketamine administration records we reviewed covered 21 months, from January 2019 through September 2020. During that period, Aurora Fire reported administering ketamine for excited delirium 22 times.340 These records show that in more than half the incidents, paramedics failed to follow ketamine monitoring protocols or administered ketamine at doses above the maximum allowable dose. Proper ketamine dosing is based on a patient’s weight, and medical experts connect high doses of ketamine with increased risk of medical complications.341 The records show that Aurora City Council Minutes of Sept. 14, 2020 Council Meeting 4-5. Id. 340 Colorado Health Facilities and Emergency Medical Services Division, Denver Metro Waivered Administration of Ketamine for Excited Delirium and/or Extreme or Profound Agitation. 341 Ketamine Use in Prehospital and Hospital Treatment of the Acute Trauma Patient: A Joint Position Statement (Aug. 27, 2020). The statement represents the collective consensus of The American College of Surgeons Committee on Trauma (ACS-COT), the American College of Emergency Physicians (ACEP), the National Association 338 339 95 nine patients experienced adverse complications following ketamine administration, including sinus tachycardia (fast heart rate), hypoxia (low oxygen supply), and bradypnea (slow breathing rate).342 One patient was given assisted ventilation while in transit, and placed on a ventilator at the hospital. All but one patient who experienced an adverse complication was either given the maximum dose of ketamine, 500 mg, or given ketamine following the administration of another sedative. The risk of adverse complications borne out in the patient records raises concerns about the agencies’ frequently high dosing of ketamine. Aurora Fire records show that at least three patients from the table below were given too much ketamine for their weight, 50 mg above the maximum dose. A review of redacted patient care reports, however, showed that paramedics failed to document a patient’s weight five different times. Failure to document this information makes it impossible to determine whether proper ketamine doses were administered. Moreover, each time a report failed to include the patient’s weight, paramedics gave patients 500 mg of ketamine, the maximum dose permitted for any patient. Perhaps more concerning is that even though these weights were not recorded in the patient care reports, Aurora Fire reported those patient’s weights in the records they provided to the Department of Public Health. For these patients, the reported weight was not always enough to justify the maximum dose given. And in five other instances, the weights listed in Aurora Fire’s care reports did not match the records they gave to the Department of Public Health. Literature from Aurora Fire’s excited delirium training emphasizes the need to accurately estimate a patient’s weight to administer the correct dosage.343 Hospitals where patients were transferred, however, do not submit data confirming patient weights, so there is no way to verify whether paramedics are overestimating patient’s weight, or not estimating it at all, as the information we reviewed suggests. In July 2020, Aurora Fire included guidance on how to estimate a patient’s weight to their protocol manual.344 The guidance recommends paramedics “use your own size and weight in comparison to your patient,” and suggests several crew members independently assess the patient’s weight and compare to establish an accurate estimate.345 The protocol was adopted after most of the administrations we reviewed and shortly before the City Council suspended all ketamine use, so we cannot know if it would have prevented excess dosing. of State EMS Officials (NASEMSO), the National Association of EMS Physicians (NAEMSP) and the National Association of EMTs (NAEMT). 342 Colorado Health Facilities and Emergency Medical Services Division, Denver Metro Waivered Administration of Ketamine for Excited Delirium and/or Extreme or Profound Agitation. 343 Nick Anderson, Ketamine and Versed 22, on file with authors. 344 Aurora EMS Protocols, Summer 2020, 9000 General Guidelines: Medication Administration. 345 According to the protocol, “If there is discrepancy in weight estimation among the crew members, the lead medic should discuss the weight estimations with the crew and choose the lower weight estimate if the discrepancy persists.” Id. 96 House Bill 1251 now requires paramedics to weigh or triple-check a patient’s estimated weight through a process similar to Aurora Fire’s protocol outlined above. The law does not address the pattern of overdosing patients even when paramedics properly estimate and record a patient’s weight. Additionally, in eleven incidents below Aurora Fire reported that medical observation protocols were not completely followed.346 Generally, protocols require paramedics to begin capnography (carbon dioxide levels), pulse oximetry (oxygen saturation), and cardiac monitoring after ketamine is administered.347 Aurora Fire’s records show that five patients did not receive capnography monitoring, eleven patients did not receive cardiac monitoring, four patients did not receive oximetry monitoring, and two of these patients received no medical monitoring. The patient care reports document at least one instance when patient behavior prevented the paramedics from conducting proper medical monitoring. But most care reports do not explain the failure to follow protocols. And, of course, when Aurora Fire and Falck reported that they followed the monitoring protocol, that does not mean that the monitoring protocol was, in fact, followed. In at least two instances, Aurora Fire’s care reports do not show proper monitoring, but the reports to the Department of Public Heath list that monitoring was followed. Because we observed only the records for these incidents, we cannot independently verify in any circumstance whether Aurora Fire and Falck did, or did not, follow the monitoring protocols. Ketamine Administrations as Reported by Aurora Fire Rescue from January 2019 through September 2020.348 Date Recorded Patient Weight / Conversion Dose Given Protocol Dose ExcessDosage Monitoring Protocol Followed? Jan. 2019 weight not recorded 500 mg unknown unknown No Feb. 2019 150 lbs / 68 kg 300 mg 350 mg – No Feb. 2019 weight not recorded 500 mg unknown unknown Yes Feb. 2019 weight not recorded 500 mg unknown unknown Yes Feb. 2019 weight not recorded 500 mg unknown unknown No Feb. 2019 90.72 kg / 200 lbs 500 mg 450 mg + 50 mg No Mar. 2019 90.72 kg / 200 lbs 450 mg 450 mg – Yes Mar. 2019 170 lbs / 77.1 kg 200 mg 400 mg – No Mar. 2019 81.66 kg / 180 lbs 450 mg 400 mg + 50 mg No Apr. 2019 weight not recorded 500 mg unknown unknown No Id. Aurora Fire Rescue and Falck Rocky Mountain, Aurora EMS Protocols: Summer 2019, 226. 348 The data from this table was compiled from Aurora Fire documents created to comply with state law. As mentioned earlier, this information excludes the reported information for Mr. McClain’s ketamine administration. 346 347 97 Date Recorded Patient Weight / Conversion Dose Given Protocol Dose ExcessDosage Monitoring Protocol Followed? Apr. 2019 280 lbs / 127 kg 500 mg 500 mg – Yes June 2019 250 lbs / 113.4 kg 500 mg 500 mg – Yes July 2019 250 lbs / 113.3 kg 500 mg 500 mg – Yes July 2019 190 lbs / 86.2 kg 500 mg 450 mg + 50 mg Yes July 2019 220 lbs / 99.8 kg 500 mg 500 mg – No Aug. 2019 170 lbs / 77.1 kg 400 mg 400 mg – No Sept. 2019 220 lbs / 99.8 kg 500 mg 500 mg – No Dec. 2019 170 lbs / 77.1 kg 400 mg 400 mg – Yes Jan. 2020 81.65 kg / 180 lbs 400 mg 400 mg – Yes May 2020 113.4 kg / 250 lbs 500 mg 500 mg – Yes Aug 2020 160 lbs / 72.6 kg 350 mg 350 mg – Yes Sept. 2020 160 lbs / 72.6 kg 200 mg 350 mg – No The Department of Public Health ketamine waiver program requires the medical director to review every ketamine administration incidence within seven days.349 Cases with adverse reactions must be reviewed within 24 hours. According to the medical director overseeing ketamine administration for Aurora Fire and Falck, those reviews were conducted either by himself or one of three assistant medical directors, who also reviewed other types of EMS cases. The ketamine review included ensuring that the protocol was indicated by the patient’s symptoms, proper dosing was administered, and appropriate medical monitoring was initiated. Each month, the four physicians met to discuss the cases and identify any concerning trends or training needs. But the monthly meetings were not limited to ketamine reviews, and did not include a longitudinal analysis beyond what the medical directors recalled from the previous month. Absent a systematic process to review all ketamine administrations over a three or six month period, trends in excess dosing or failure to follow monitoring protocols may have been difficult to identify from memory alone. Even so, the medical director stated in his waiver application that paramedics whose cases fall outside the protocol would have to repeat the initial ketamine training.350 Despite evidence showing paramedics often failed to follow the agencies’ ketamine protocol, the director said that no ketamine calls triggered a further case review and no remediation or training needs were identified. Colorado Department of Public Health and Environment, Ketamine and RSI Waiver Reporting Guidance, 2021. 350 Dr. Eric Hill, Application for Excited Delirium Ketamine Administration Waiver, on file with authors. 349 98 The only change to emerge from the reviews was to include in the general medication guidance protocol that paramedics independently assess a patient’s weight and then compare their estimates. Aurora Fire training reports confirm that only one online training on excited delirium was offered in 2020, and it was prepared by a paramedic, not a medical director or experienced trainer.351 The protocols for ketamine remained unchanged during the period of administration. Our review of the patient care reports also showed that not all paramedics included detailed summaries of the patient’s presentation. To be sure, some of the patient care reports cited symptoms consistent with descriptions of excited delirium, including patients removing their clothing while actively fighting with police—a possible indication of hyperthermia—fast pulse, incoherence, and increased strength. But several care reports lacked important details to confirm that patients met the protocol for ketamine. The same was true in the corresponding Aurora Police records for these incidents. Both departments thoroughly documented when a subject demonstrated behaviors were consistent with a diagnosis of excited delirium. But when the subject’s behaviors mainly consisted of aggressive resistance, the records are more spare, and it is hard to confirm that patients were thoroughly assessed before paramedics administered ketamine. The training provided to Aurora paramedics emphasizes that the most likely cause of excited delirium is stimulant drug misuse, although psychiatric disorders coupled with drug use may also place a patient at risk.352 Paramedics identified “overdose” as the suspected cause of excited delirium in only seven patients. For the other patients, paramedics listed “psychiatric/behavioral” as the suspected cause.353 These data points neither align with the training provided to Aurora paramedics nor with the medical literature on the etiology of excited delirium. When the reported symptom is psychiatric or behavioral, the patient care reports emphasize the patient’s combativeness, suggesting that patients who are resisting treatment or arrest while experiencing psychosis may be administered ketamine instead of slower-acting, but perhaps safer, chemical sedatives. It also raises the question of whether ketamine has been used to transport a patient who refused treatment, in violation of the constitutional right to reject medical care. 10.2.2. Officers Sometimes Requested Chemical Restraints and Failed to Document That Paramedics Chemically Restrained the Person We found that officers often failed to document paramedics’ use of chemical sedatives— ketamine or otherwise—in their general offense and use-of-force reports. We found five incidents of ketamine administration that corresponded with a use-of-force review. The investigators did not include chemical restraints in their reports, despite the constitutional implications of giving a person medication without their consent or, in some cases, over Aurora Fire Rescue, Active Training Events for period 01/01/2018 Through 12/31/2020, on file with authors. 352 Travis Chambers, Pre-Hospital Ketamine, on file with authors. 353 Colorado Health Facilities and Emergency Medical Services Division, Denver Metro Waivered Administration of Ketamine for Excited Delirium and/or Extreme or Profound Agitation. 351 99 their objections. Officers did not report their own requests for paramedics to sedate suspects. And investigators who reviewed the body-worn camera footage documenting officers’ requests failed to include that information in their use-of-force reports. A review of use-of-force reports for dates corresponding to ketamine administrations shows that Aurora Police conducted only five use-of-force reviews out of the 24 times paramedics used ketamine. Aurora Fire’s patient care reports, however, show that Aurora Police were involved in most ketamine administrations. Because the use-of-force policy in place during the period of ketamine administration did not require officers to file a report “when restraining persons solely for medical, emotional, or mental health purposes,” this failure to document chemical sedation was policy-driven.354 The lack of documentation by Aurora Police has prevented oversight on officers’ interaction with paramedics, which may have led to blurred boundaries between medically necessary interventions and unnecessary chemical restraints. The Colorado EMS Chiefs, Managers and Directors and the Emergency Medical Services Association of Colorado issued a statement that “paramedics must make the critical determination if [a] person is experiencing a medical emergency.”355 The American Society of Anesthesiologists issued a statement that the organization “firmly opposes the use of ketamine or any other sedative/hypnotic agent to chemically incapacitate someone for a law enforcement purpose and not for a legitimate medical reason.”356 In response to the precipitous increase in ketamine administrations in Minneapolis, the police department issued a policy prohibiting officers from making “any suggestions or requests regarding medical courses of action to be taken by any medical personnel.”357 But Aurora Police have requested sedation for combative suspects, and Aurora paramedics have complied. Our review of patient care reports shows that police-involved patients sometimes received scant documentation that would confirm the medical need for sedation. During our ride-alongs with Aurora Fire, paramedics said that police officers used to feel comfortable telling paramedics to restrain patients. The paramedics were quick to add that this was no longer common practice. We matched the date-stamp from body-worn camera footage aired on a television news report to ketamine administration records, the patient care report, the general offense report, and the use-of-force report for that incident. The body-worn camera footage shows Aurora Police Department Directive 05.04, May 13, 2019. This policy has been updated to require officers to document when force is used when a person is in medical and not police custody. The new policy does not require officers to document when paramedics sedate a person who had been in police custody. 355 Statement on Sedation of Prehospital Patients, July 18, 2020. This statement was adopted by most Colorado EMS physician associations, including the Denver Metro EMS Physicians, but it has since been removed from the Emergency Medical Services Association of Colorado’s website. The Association replaced the July 2020 statement with a similar statement on November 5, 2020 with a substantially similar recommendation that “Medical emergencies must be handled by emergency medical professionals.” The July 2020 position statement is still available here. The November 2020 position statement is available here. 356 ASA Statement on the Use of Ketamine for a Non-medical Purpose, July 15, 2020. 357 See MPD Involvement in Pre-Hospital Sedation, (July 26, 2018). 354 100 an officer repeatedly asking paramedics to sedate a suspect.358 The officer asks “can we get this guy like some Versed or something and get him to calm down?” A few seconds later he asks, “can we get like Versed and calm this dude down or Haldol or something? Let’s give him some juice to go to sleep … and then we can deal with him.” And again, “If we give, if we give him some meds, he’ll probably, yeah we’ll give him some calm-down juice.” According to the body-worn camera footage, the paramedic on scene does not appear to evaluate the man, but asks the officers if “he’s still fighting you guys?” The paramedic then returns with an injection, and the suspect protests: “No. You’re not hitting me with nothing. I said no.” Records show that Aurora Fire paramedics injected the man with 450 mg of ketamine. The patient care report states the man had a one-inch laceration on his left forehead when they arrived. It says the man was showing “signs of excited delirium --hyper-aggression / increased strength.” The report notes the patient had been tased, handcuffed, and hobbled, but was “still combative and a danger to himself and others.” Combativeness, however, is not by itself an emergency medical condition. The use-of-force investigation reported that the man already had a cut on his face when officers first approached him. The report documented that the man was “being uncooperative” as officers held him on the ground. An officer on scene reported that the man was “actively trying to fight his way out of soft restraints, and needed to be transported to the hospital for his injuries.” But the only documented injury, other than the taser barbs the paramedics removed on scene, was a one-inch laceration that was not lifethreatening. The officer reported that “Aurora Fire issued a chemical restraint,” and identified the restraint as ketamine. But no one documented that police had requested paramedics sedate the man so “we can deal with him.” None of the officers on scene reported their interactions with paramedics in the general offense report, either. The news report included a statement from Aurora Police that “officers have already been reminded they are not to suggest or attempt to direct medical treatment.” But Aurora Police did not have a written policy confirming this verbal admonition at the time. In the statement, the Chief of Police also said that while the officer’s choice of words were “not ideal,” he was “trying to get the man in crisis help.” This clarification appears to undermine the verbal policy, because officers cannot both refrain from suggesting medical treatment and try to get the man medical help. Aurora Fire also released a statement that “Aurora Police Department does not influence our decision-making for treating patients,” and their paramedics “maintain medical control.” That statement does not appear borne out in this incident or in other anecdotal reports. House Bill 1251 addresses some of these concerning interactions by preventing police from directing paramedics to administer ketamine and preventing paramedics from relying Aurora Body Camera Video Shows Police Asking Medics to Give Powerful Drugs to Suspect to Calm Him, The Denver Channel, Dec. 3, 2020. 358 101 solely on officers’ reports in assessing a patient. And paramedics reported that new verbal policy requires the senior paramedic to announce to dispatch when they take over custody of a patient from officers. Our observations of paramedics and officers, however, revealed that when policy changes emphasized what should not be done, both paramedics and officers were confused about how and what they should do instead. For example, when Aurora Police changed their policy to require officers to report when they use any force on a person who needs “medical, emotional, or mental health” assistance, we were told that officers quit helping paramedics with these patients. Paramedics reported that they were being trained to search patients for weapons, but had not been told what to do if they found a weapon or who should take custody of a weapon. Both officers and paramedics expressed frustration in the gaps in their training, which left them confused about how to interact on community care calls. Aurora Fire’s restraint protocol recommends coordination with officers, but it does not address how paramedics should interact with officers or how to maintain medical objectivity when a person is resisting arrest. Aurora Police released a new policy dated September 7, 2021 that attempts to address how officers should coordinate with paramedics and EMS providers on scene.359 The new policy prioritizes the medical needs of a person, regardless of their custody status, and describes what information officers should provide to emergency medical personnel. It includes an official transfer of care procedure for when paramedics take custody of a police suspect. It outlines the limits on an officers’ ability to assist emergency medical personnel when a patient resists medical care, has a weapon, needs restraint, and it suggests non-physical assistance such as de-escalation or enlisting crisis response teams, mental health counselors, or family and friends of the patient. This recent policy is a substantial step in the right direction. But the new coordination policy fails to address the authority officers have to act when someone is in a mental health crisis and there is probable cause to take that person into custody for purposes of a mental health hold. The complete prohibition on officers using force to assist emergency medical providers is inappropriate when officers may need to prevent a patient from escaping transport to a mental health facility. The new Aurora Police coordination policy is designed to prevent the illegal use of sedatives on combative suspects. Policy must be reinforced with training, practice, and better documentation of patient restraints. Because police officers have routinely failed to document when paramedics used chemical restraints, it is impossible to determine how often paramedics illegally sedated patients in police custody. The new policy, however, does not require officers to document when a police-involved patient was restrained or sedated after they were transferred to Aurora Fire’s custody. Without documentation and accountability measures, it will remain difficult to assess how effective the new policy will be. Additionally, Aurora Fire does not have a policy on how paramedics ought to coordinate with officers. Aurora Fire’s restraint protocol recommends coordination with officers, but it does not address how paramedics should interact with officers or how to maintain medical 359 Aurora Police Department Directive 09.06 (effective Sept. 7, 2021) 102 objectivity when a person is resisting arrest. The Coordination with Aurora Police policy covers radio communication and initial requests for police assistance, but does not address on-scene coordination or care and custody transfers.360 Similarly, the policies on “Response to APD request for Emotional/Psych Patient Assistance” and “Citizen Assistance Notification” addresses how to report responses to dispatch and whether both Aurora Fire and transport or transport only will respond.361 The above incident is one example of paramedics quickly administering ketamine to a person who does not appear to present the protocol’s description of excited delirium symptoms and who is capable of articulating their own symptoms and needs. Because officers and paramedics will continue to work together on difficult calls, Aurora Fire must also establish effective accountability measures to prevent paramedics from sedating patients illegally, regardless of the type of sedative used. Aurora Fire and Aurora Police must develop joint training to ensure patient care is prioritized and first responders adhere to the expectations of the law and policies of both agencies. 10.3. Findings on Aurora’s Ketamine Protocols 10.3.1. Protocols Did Not Emphasize the Need for Paramedics to Conduct an Independent Examination of a Patient The passage of House Bill 1251 in the summer of 2021 precludes the use of ketamine for excited delirium “when a peace officer is present” and permits its use only in a justifiable medical emergency, which is defined as “a medical, traumatic, or psychiatric condition posing an immediate safety risk to the individual, emergency medical service provider, or the public.”362 Because excited delirium is a diagnosis based on a set of symptoms without a specified medical cause, the law also bans the use of ketamine for “any subsequent term for excited delirium, or any acute psychiatric diagnosis not recognized in the most recent edition of the Diagnostic and Statistical Manual of Mental Disorders.” The law does not prohibit ketamine administration for excited delirium when paramedics are not in the presence of law enforcement. Aurora Fire added a protocol for “control of excited delirium” in the summer of 2018,363 which it updated again in the summer of 2019.364 These protocols established how paramedics were to administer ketamine to patients who were showing symptoms of excited delirium. In September 2021, Aurora Fire released a new set of protocols, which adds several measures to protect patients from unnecessary sedation. It does not include ketamine as a drug paramedics may administer to patients. The medical director who helped draft the new protocol said that Aurora Fire included a pre-sedation checklist based Aurora Fire Rescue Manual of Procedures, Policy 6.14. Aurora Fire Rescue Manual of Procedures, Policy 6.19: Unconventional Request for Service. 362 House Bill 21-1251 363 Memorandum from Fernando M. Gray, Sr., Fire Chief, regarding Prehospital Protocol Updates (Oct 2, 2018). 364 Aurora Fire Rescue and Falck Rocky Mountain, Aurora EMS Protocols: Summer 2020. 360 361 103 on suggestions from the report issued by an independent review panel investigating the death of Elijah McClain.365 The new protocols are another step in the right direction. Because the new sedation protocol was released after our period of investigation, it is impossible to know whether it would have prevented the violations we found. House Bill 1251 maintains the option for agencies to use ketamine. Aurora Fire could apply for a new ketamine waiver and add it back into their protocols. It is important, therefore, to outline how past policies impacted ketamine administrations to establish what steps Aurora Fire would need to take before applying for a new ketamine waiver, should they choose to do so. The Aurora EMS manual includes excited delirium in the behavioral protocols,366 which is hyperlinked to the Agitated/Combative Patient protocol and the restraint protocol.367 In past manuals, these were linked to a protocol for ketamine.368 Both the old and the new protocols describe excited delirium in essentially the same language: “These patients are truly out of control and have a life-threatening medical emergency. They will have a severe metabolic acidosis which is life threatening. They will have some or all of the following symptoms: paranoia, disorientation, hyper-aggression, hallucination, tachycardia, increased strength, and hyperthermia.”369 The protocols instruct responders to “assume the patient has a medical cause of agitation and treat reversible causes” before moving to a potential diagnosis.370 They do not provide paramedics a method to determine which symptoms in which combination are more likely to signify a medical emergency.371 The restraint protocol, which includes both physical and chemical restraints and was unchanged in the 2021 manual, permits paramedics to restrain a person “only when necessary to prevent a patient from seriously injuring him/herself or others (including the EMS providers), and only if safe transportation and treatment of the patient cannot be accomplished without restraints.”372 It requires personnel to use “the minimum amount of force necessary to control the patient and prevent harm,” and also precludes using restraints “for the convenience of the crew.”373 It encourages alternatives to chemical City of Aurora, Investigation Report and Recommendations at 123, Feb. 22, 2021. Aurora Fire Rescue and Falck Rocky Mountain, Aurora EMS Protocols: Summer 2019 at 53. 367 Id. at 99. 368 Aurora Fire Rescue and Falck Rocky Mountain, Aurora EMS Protocols: Summer 2019 at 143. Aurora Fire’s current manual now links the Agitated/Combative Patient and Excited Delirium Syndrome protocols to the medication Benzodiazepines (Midazolam), which is the generic form of Versed. It does not include a ketamine protocol. 369 Id. at 48; Aurora Fire Rescue and Falck Rocky Mountain, Aurora EMS Protocols: July 2021 at 105. 370 Aurora Fire Rescue and Falck Rocky Mountain, Aurora EMS Protocols: Summer 2019 at 143 371 See White Paper Report on Excited Delirium, American College of Emergency Physicians, Sept. 10, 2009. 372 Id. at 53. 373 Id. 365 366 104 sedation, with verbal de-escalation the only listed technique.374 Coordination with law enforcement is also encouraged to assist with patients who exhibit violent, combative, or uncooperative behaviors and who do not respond to verbal de-escalation. Paramedics are instructed to evaluate the patient “to determine his or her medical condition, mental status and decision-making capacity.”375 Patients are to be treated with respect, and restraints used only if safe transportation and treatment “cannot be accomplished without restraints.”376 Documentation on the justification for restraints and efforts used to de-escalate before chemical sedation is required.377 The Agitated/Combative protocol in place when ketamine was used was a flow chart which included as a general guideline the need for “safety, appropriate use of restraints and aggressive treatment of the patient’s agitation.” Ruling out excited delirium syndrome was the third step in the chart, which came before verbal de-escalation.378 It recommended administering ketamine to induce “rapid tranquilization” and “minimize time struggling” if a patient shows any of the listed symptoms.379 The rule-out process listed the common symptoms of excited delirium and referred paramedics to the ketamine protocol.380 The protocols in place when Aurora Fire used ketamine called for a dose of 5 mg/kg IM (intramuscular) with a maximum initial dose of 500 mg. Paramedics were required to contact base before administering a second dose, which was set at one-half the initial dose. No alternative medications were referenced in the ketamine or the excited delirium protocols.381 In accordance with the waiver program, both the Agitated/Combative Patient protocol and the ketamine protocol required paramedics to start medical monitoring of cardiac function, carbon dioxide levels, and oxygen saturation. The Agitated/Combative Patient protocol required paramedics to begin high flow oxygen and IV saline. The Ketamine protocol required physical restraints once a patient was sedated and during transport. Both protocols emphasized that patients were experiencing a medical emergency and required rapid transport to a hospital. All three protocols included mention of law enforcement involvement, but none discussed how to coordinate with officers at the scene, when and whether to intervene on a patient’s behalf, or that medical decisions are the exclusive province of the assessing paramedics. Id. Id. 376 Id. 377 Id. 378 Id. at 99. 379 Id. at 99. 380 Id. at 143. 381 The current protocol instructs paramedics to use Midazolam (Versed). 374 375 105 10.3.2. National Ketamine Protocol Trends and Recommendations From Medical Associations Conflict With the High Doses Permitted by Aurora Fire’s Protocol The National Emergency Physician and EMS Consensus Protocol considers administration of ketamine at 5 mg/kg IM a “high dose” which is “associated with an increased intubation rate.”382 In evaluating Aurora Fire’s Agitated/Combative Patient and ketamine protocols, we examined ketamine protocols from rural, urban, and statewide regions across the United States. We found many consistencies, including tiered intervention responses to agitated patients and requiring documentation when paramedics administer ketamine. Most protocols recommend requesting law enforcement assistance for combative patients.383 Many include medical monitoring protocols following ketamine.384 A few policies suggest paramedics determine whether patients have been tased, which increases the likelihood of cardiac arrest.385 Many protocols included a nuanced rating scale, such as the Richmond Agitation Sedation Scale or the Altered Mental Status Scale to help paramedics assess whether chemical interventions are indicated, and if so, which ones.386 The protocol survey showed that dosing recommendations for ketamine, however, are inconsistent across the United States. For example, • Wentzville, Missouri (a suburb of St. Louis) and Minneapolis both permit initial 5 mg/kg ketamine doses. • The National Emergency Physician and EMS Consensus Protocol recommends a dose of 3-5 mg/kg IM (intramuscular).387 Ketamine Use in Prehospital and Hospital Treatment of the Acute Trauma Patient: A Joint Position Statement (Aug 27, 2020). The statement represents the collective consensus of The American College of Surgeons Committee on Trauma (ACS-COT), the American College of Emergency Physicians (ACEP), the National Association of State EMS Officials (NASEMSO), the National Association of EMS Physicians (NAEMSP) and the National Association of EMTs (NAEMT). 383 Policies from Connecticut; Houston, Texas; Kansas City, Missouri; Maryland; New York; and Wenatchee, Washington recommend involving police when restraints are required. Minneapolis, Minnesota began a review of police involvement in pre-hospital sedation and instituted several policies on officer interactions with paramedics. 384 See protocols from Kansas City, Missouri; Pennsylvania; Portland, Oregon; New York; and Wentzville, Missouri. 385 See protocols from Maryland. 386 Aurora’s newest protocol now includes the Richmond Agitation Sedation Scale. Exemplary scales and explanations on how to assess a patient’s agitation are included in the protocols from Portland, Oregon; Maine; and Pennsylvania. See also Ketamine Use in Prehospital and Hospital Treatment of the Acute Trauma Patient: A Joint Position Statement (Aug 27, 2020). The statement represents the collective consensus of The American College of Surgeons Committee on Trauma (ACS-COT), the American College of Emergency Physicians (ACEP), the National Association of State EMS Officials (NASEMSO), the National Association of EMS Physicians (NAEMSP) and the National Association of EMTs (NAEMT). 387 Ketamine Use in Prehospital and Hospital Treatment of the Acute Trauma Patient. 382 106 • The National Model EMS Clinical Guidelines and protocols in Albuquerque, New Mexico; Connecticut; Flathead County, Montana; Maine; Maryland; Pennsylvania; Vermont; and Central Washington limit ketamine doses to 4 mg/kg. • Miami and New York use a flat 250 mg initial dose. • Kansas City, Missouri; Portland, Oregon; and Houston, Texas do not permit ketamine use for agitated patients. The maximum dose permitted ranged from 300 mg to 500 mg. The survey revealed that a 5mg/kg dose, such as Aurora Fire’s protocol permits is at the highest level found in the nation. 10.4. Required Changes Several stakeholders in Aurora have stated they do not plan to use ketamine in the future. Before Aurora Fire uses ketamine or similar sedatives in the field, it must make these changes. 10.4.1. Review Ketamine Dose Recommendations The protocol for Aurora Fire Rescue and Falck Rocky Mountain instructs paramedics to administer ketamine at 5 mg/kg. This dose falls at the highest end of national recommendations. Medical directors should consider lowering the recommended dosage to align with nationwide trends which limit ketamine dosage to 4 mg/kg or less. 10.4.2. Maintain a Uniform Method to Assess Patient Agitation House Bill 1251 requires paramedics to record complications arising from ketamine use, but protocols should also require detailed records of the symptoms supporting a paramedic’s assessment that ketamine is necessary. The new protocols now include the Richmond Agitation Sedation Scale, which is a uniform method to assess the severity of a patient’s agitation, and requires paramedics to document each patient’s scaled agitation level prior to using sedation. This is a substantial step towards addressing this concern. Paramedics should be thoroughly trained in using the scale to quickly assess whether ketamine is indicated. House Bill 1251 requires medical directors to develop training for paramedics to ensure their compliance with the new law. Frequent citing of psychiatric/behavioral causes for excited delirium reflect that training is needed to help paramedics distinguish agitated states caused by a choice to resist arrest, mental illness, intellectual disability, or other medically less-dangerous situations from “a justifiable medical emergency.” Paramedics should be given clear guidance on what constitutes an emergency requiring ketamine. Current Aurora Fire protocols on coordinating with law enforcement are limited to transporting handcuffed patients. As addressed above, Aurora Fire and Aurora Police should develop a joint policy and training on effective coordination that includes the 107 transfer of care and custody from officers to paramedics. House Bill 1251 makes clear that paramedics alone should make medical assessments and decisions, such as if a person becomes medically distressed and requires intervention or chemical sedation. But emergency responders often turn to bystanders for clues on what might be causing a patient’s medical emergency, and House Bill 1251 permits officers to provide pertinent medical information that could help a paramedic assess whether the patient is experiencing a justifiable medical emergency.388 The new Aurora Police policy on coordinating with Aurora Fire lists the types of information officers should provide to paramedics.389 Joint training is needed to ensure both officers and paramedics are comfortable requesting and providing necessary medical information that complies with the new law. 10.4.3. More Stringent Review to Ensure Protocol Compliance Because the use of ketamine can create life-threatening situations, Aurora Fire must engage in rapid and thorough intervention when protocols are not followed. More stringent review procedures are necessary to ensure that paramedics follow protocol dosing and conduct complete medical monitoring during transport as required by House Bill 1251. To ensure patient safety, Aurora Fire should conduct a Post-Incident Analysis for each ketamine administration, create systems for addressing any violations of protocol or law, and develop a formal method for evaluating chemical sedation over a period of months to catch alarming trends.390 We observed several apparent policy violations in how Aurora administered ketamine. First, records given by Aurora Fire to the Department of Public Health showed that 40% of all ketamine administrations exceeded the maximum dose permitted by protocols. Because Aurora Fire did not record all patient’s weights, we cannot verify how many patients received an excessive dose. . But when the patient’s weight was recorded, the excessive doses did not result from mis-identifying a patient’s weight. Second, records also show that paramedics did not start required medical monitoring in 50% of administrations. Although some monitoring was reported in all but two instances, incomplete application of the necessary monitoring protocols may place patients at risk for complications. Coupled with the excessive dosing instances, the records reveal that more than 60% of all ketamine administrations failed to follow existing protocols completely. Third, Aurora Fire did not use a uniform method to assess the severity of a patient’s agitation. The old ketamine protocol was indicated for “Adult patients with signs of excited delirium where the safety of the patient and/or providers is of substantial concern.”391 House Bill 1251 bans the use of ketamine for excited delirium, but describes “a justifiable medical emergency” authorizing ketamine as a “medical, traumatic, or psychiatric condition posing an immediate safety risk to the individual, emergency medical service provider, or National Registry of Emergency Medical Technician Patient Assessment. Aurora Police Department Directive 09.06 (effective Sept. 7, 2021) 390 Aurora Fire Rescue Manual of Procedures, Policy 3.6 and 6.28. 391 Id. at 143. 388 389 108 the public.”392 The law introduces an immediacy requirement for the safety risk which must be part of any future ketamine protocol. The old protocols did not include a scale for paramedics to rate the severity of a patient’s agitation or combativeness. Without a uniform scale, determining the severity of a patient’s agitation or the immediacy of any safety risk may differ based on who is conducting the assessment. Improper ketamine administrations are more likely when paramedics must rely only on individual judgment rather than judgment based on uniform assessment practices. The new protocol addresses this concern, and the required documentation of a patient’s agitation scale will help ensure patients meet criteria for sedation. Fourth, failure to follow ketamine protocols, such as exceeding proper dosages or omitting medical monitoring, did not trigger supervisory intervention and the medical directors had no formal process to evaluate longitudinal trends which might have identified such failures. The medical director’s waiver application included a plan to intervene with training if paramedics did not follow ketamine protocols. But no interventions occurred during the eighteen-month period despite a pattern of over-administering ketamine and under-using medical monitoring. Fifth, lack of documentation by Aurora Police of paramedics administering ketamine or other chemical sedatives prevents adequate oversight into police and paramedic coordination. Aurora Police use-of-force reviews do not consistently include paramedics’ use of chemical restraints on people in police custody. Lack of documentation or incomplete documentation of these interactions makes it difficult for supervisors to address the full spectrum of restraint on a subject. The lack of documentation also prevents review of police and paramedic on-scene coordination. Because both Senate Bill 217 and House Bill 1251 require detailed reporting on use-offorce incidents, and because House Bill 1251 identifies chemical restraint as a type of force, Aurora Police must improve documentation of incidents that result in a paramedic’s decision to sedate a person, even if no officer used any level of force during the encounter. To ensure accurate reporting, Aurora Fire personnel should communicate to dispatch and directly with the officer-in-charge on scene that chemical restraints were needed for a patient. Aurora Police reports for such incidents must be shared with Aurora Fire. If the chemical restraint occurred following police restraint or use of force, a joint review with Aurora Police and Aurora Fire should be conducted to evaluate the agencies’ coordination on scene and the care and custody transfer. Even if officers did not use restraint or force, including the use of chemical restraints in general offense reports will enable leadership from both agencies to evaluate the interactions, ensure accountability, and help identify if training is needed to improve coordination and patient outcomes. 392 § 25-3.5-103(8.6), C.R.S; § 25-3.5-209(3), C.R.S. 109 11. Facilitating Meaningful Reform To ensure that the changes required above lead to real, meaningful improvements, we will also require that Aurora pay for an independent monitor, chosen with input from Aurora Police, Fire, and City Council, who will report to a court and provide periodic public updates about Aurora’s progress in implementing these changes. In addition, to ensure that these changes can be implemented to the greatest extent possible, we will require the Aurora Civil Service Commission to make its work publicly transparent and available for review to the fullest extent that current law permits for all work addressing Aurora Fire and Police. This requirement includes providing transcripts or public access for discipline hearings and deliberations on hiring and promotional considerations whenever possible. When participants choose to exercise the option that City Charter gives them for confidential procedures, the Civil Service Commission must make the fact of the proceedings public and note that the individual chose to keep proceedings confidential. The Civil Service Commission must publicly publish, on at least a quarterly basis, a specific list of all proceedings where a participant denied public participation. Importantly, we will require Aurora, as a condition of any negotiated resolution, to have a public assessment of the current structure of the Civil Service Commission and whether changes to the city charter should be made to ensure that hiring, promotion, and discipline review accomplish the overall goals of the city and ensure that Aurora Fire and Aurora Police can quickly improve to meet the needs of the City. 110 12. Next Steps Senate Bill 217 requires a negotiation period with Aurora so that the City and the Attorney General can try to make these required changes part of a voluntary consent decree.393 We want Aurora to succeed in these improvements and strongly believe that an agreement provides the best way to do so. Initial conversations with Aurora Police, Aurora Fire, and city management tell us they share that goal. We will work over the coming weeks with Aurora and other stakeholders to create a consent decree that makes sure these requirements are implemented promptly. 393 § 24-31-113, C.R.S. 111 13. Acknowledgments Our team included many professionals outside of our office who donated hundreds of hours of their time. We thank them for their commitment to supporting this important work: • Sundeep K. (Rob) Addy and Sean Grimsley of Bartlit Beck LLP • Dr. Theron Bowman of The Bowman Group • Mihir Gokhale, Peter Horvath, Michael Kwak, and Noah Mathews of Compass Lexecon • Samantha Kappagoda and Dr. David K. A. Mordecai of Risk Economics • Wendy Olson of Stoel Rives LLP 112