Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 1 of 108 1 THE HONORABLE JAMES L. ROBART 2 3 4 5 6 7 UNITED STATES DISTRICT COURT WESTERN DISTRICT OF WASHINGTON AT SEATTLE 8 9 10 UNITED STATES OF AMERICA, Plaintiff, 11 v. 12 13 No. 2:12-cv-01282-JLR MEMORANDUM SUBMITTING NINTH SYSTEMIC ASSESSMENT REGARDING USE OF FORCE CITY OF SEATTLE, Defendant. 14 15 16 17 18 19 20 The Monitor presents to the Court the Ninth Systemic Assessment regarding the Seattle Police Department’s use of force. DATED this 6th day of April, 2017. s/Merrick J. Bobb Merrick J. Bobb, Monitor 21 22 23 24 MEMORANDUM SUBMITTING NINTH SYSTEMIC ASSESSMENT REGARDING USE OF FORCE - 1 12-CV-01282-JLR MERRICK J. BOBB, MONITOR Police Assessment Resource Center P.O. Box 27445 Los Angeles, CA 90027 (213) 623-5757 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 2 of 108 SEAT TL E P OLICE MO N ITO R Ninth Systemic Assessment: Use of Force April 2017 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 3 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Table of Contents Introduction & Executive Summary............................................................................................................................ 1 Part I. Overview of Use of Force & Purposes of the Assessment .......................................................................... 12 I. A. B. Background on Use of Force ...................................................................................................................... 12 Reforms to SPD’s Use of Force Policy ................................................................................................................... 13 Changes to SPD Use of Force Training ................................................................................................................. 18 A. B. Purpose of the Assessment ....................................................................................................................... 21 The Role of the Monitor’s Assessments Generally ............................................................................................... 21 The Purpose of the Use of Force Assessment ...................................................................................................... 23 II. Part 2. Quantitative Assessment of SPD Use of Force Data .................................................................................. 26 I. A. B. C. Methodology ............................................................................................................................................. 26 Sources of Data Analyzed in this Report ............................................................................................................... 26 A Note on the Data’s Limitations ........................................................................................................................... 28 A Note on the Availability of SPD Data .................................................................................................................. 29 A. B. Overview of Force Data & Trends .............................................................................................................. 29 Use of Force by Incident Type, Severity of Force, and Across Time.................................................................... 31 Use of Force by Location/Precinct ........................................................................................................................ 34 II. III. Force Instruments & Techniques Used by SPD Officers............................................................................. 35 IV. Use of Force by Officer ............................................................................................................................. 37 V. A. B. C. D. Use of Force by Officer Characteristics ..................................................................................................... 41 Use of Force by Perceived Subject Sex ................................................................................................................. 42 Use of Force by Race/Ethnicity .............................................................................................................................. 42 Subject Impairment ................................................................................................................................................ 45 Subject Injury .......................................................................................................................................................... 48 A. B. C. Incident Characteristics ............................................................................................................................. 49 How Force Incidents Were Initiated ...................................................................................................................... 49 Ultimate Incident Type ........................................................................................................................................... 50 Incident Disposition................................................................................................................................................ 51 V. VII. Data Limiations & Recommendations ........................................................................................................ 52 A. The Difficulty of Matching Databases ................................................................................................................... 53 B. The Lack of Reliable Arrest Data............................................................................................................................ 53 C. The Absence of Police Activity Data ..................................................................................................................... 54 Part III. Officer Safety ............................................................................................................................................ 55 I. Methodology ............................................................................................................................................. 55 II. Findings .................................................................................................................................................... 55 Part IV. Public Safety ............................................................................................................................................. 59 Part V. Qualitative Assessment of SPD Use of Force ............................................................................................ 64 i Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 4 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 I. A. B. C. Methodology ............................................................................................................................................. 64 What Force Incidents Were Reviewed ................................................................................................................. 65 How the Force Incidents Were Evaluated ............................................................................................................. 68 How the Findings Are Summarized ....................................................................................................................... 73 II. Findings .................................................................................................................................................... 74 Necessity ................................................................................................................................................................ 76 Proportionality ........................................................................................................................................................ 79 Reasonableness...................................................................................................................................................... 82 De-Escalation .......................................................................................................................................................... 85 1. General Findings ............................................................................................................................................... 85 2. The Issue of Solo, Affirmative Action ............................................................................................................... 89 E. Specific Force Prohibitions .................................................................................................................................... 95 1. Force Used to Punish or Retaliate .................................................................................................................... 96 2. Force Against Individuals Only Verbally Confronting Officers ........................................................................ 96 3. Force on Handcuffed or Otherwise Restrained Subjects ................................................................................ 97 4. Force to Prevent Subject from Swallowing a Substance or to Extract a Substance or Item from Inside the Body of a Suspect without a Warrant. ............................................................................................................. 98 F. Use of Deadly Force: Neck Maneuvers.................................................................................................................. 98 G. Force Instrument-Specific Guidelines ................................................................................................................... 99 A Note on “Blast Balls”.................................................................................................................................................... 102 A. B. C. D. ii Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 5 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Introduction & Executive Summary The Department of Justice’s (“DOJ”) 2011 investigation of the Seattle Police Department (“SPD”) identified a pattern or practice of unconstitutional use of force by SPD, including “[d]eficiencies in SPD’s training, policies, and oversight with regard to the use of force,” which “contribute to . . . constitutional violations.”1 Thereafter, the City of Seattle (the “City”) and DOJ (collectively, the “Parties”) entered into an agreement to reform the SPD, which, when ordered by United States District Judge James Robart, became what is commonly referred to as the Consent Decree. The Court has twice approved revisions to the use of force policies and certified them as consistent with the Consent Decree.2 The Court has likewise approved many force-related training programs that have addressed those policies and provided SPD officers with instruction on strategies and tactics consistent with those policies.3 In terms of creating policy and training, the Monitoring Team has previously praised the Department’s efforts and compliance with requirements to create new policies and training. SPD’s use of force policies are clear, simple, balanced, and well-reasoned – perhaps, with their emphasis on de-escalation, among the best in the country. The Monitor has previously cited the training provided to officers on use of force as similarly excellent and exemplary. The purpose of this report is to evaluate whether the SPD has achieved initial compliance with the provisions of the Consent Decree between the United States and City of Seattle that address officer use of force. It focuses on whether the performance and conduct of officers in the field – over time and across incidents – can establish that SPD, after having “trained all relevant personnel as necessary to fulfill their responsibilities pursuant to the requirement[s]” on force and creating force policies, is “carrying out [that training and policy] in practice.”4 This report would not have been possible even just a few years ago. When the reform process began, “force often went unreported – leaving it subject to no departmental scrutiny.”5 When force was reported, it was documented “on paper stuffed, unreviewed, in file cabinets.”6 If reported force was investigated, those inquiries were typically incomplete or inadequate. Now that SPD is reporting, tracking, investigating, and reviewing its use of force as never before,7 this analysis of the Department’s use of force can entail both quantitative and qualitative Dkt. 1-1, Investigation of the Seattle Police Department,” United States Department of Justice - Civil Rights Division, United States Attorney’s Office – Western District of Washington” (Dec. 16, 2011) [hereinafter “2011 Findings Letter”] at 3. 2 Dkt. 115; Dkt. 225. Another revision of the force policies is pending. 3 Dkt. 144; 151; 152; 153; 165; 168; 199; 254; 277. 4 Dkt. 3-1 ¶ 184. 5 Dkt. 231, First Systemic Assessment at 1 n. 5. 6 Fourth Semiannual Report at 1. 7 See generally Dkt. 231 (First Systemic Assessment). 1 1 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 6 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 components, as envisioned by the Consent Decree.8 There is, however, no magic number or single type of evaluation that by itself can determine whether SPD is in compliance with the Consent Decree. Rather, the Monitoring Team and the Parties have analyzed and balanced important and sometimes competing factors in assessing whether SPD now is where it needs to be with respect to how, under its new policies and training, its officers are using force. For this report, the Monitoring Team reviewed both data relating to SPD use of force and at random, statistically significant samples of force cases across nearly two-and-a-half years (28 months). To be able to analyze trends and make comparisons, we divided this 28-month span into two time periods. The first, earlier time period – from July 2014 to August 2015 – covered a period soon after full implementation of the Consent-Decree-required use of force policies and related training. The second, later period – from September 2015 through October 2016 – covered more recent incidents that occurred well after officers were fully familiar with the expectations under the new force policies. The Monitor finds that overall use of force by the SPD is down – both across time under the Consent Decree and compared to the time period studied by the original DOJ investigation. Overall, use of force has gone down even as officer injuries have not gone up and crime, by most measures, has not increased. At the same time, the force that SPD officers do use is, by and large, reasonable, necessary, proportional, and consistent with the Department’s use of force policy. Because officers are using less force overall, without negatively impacting officer safety or public safety, and are using force consistent with law and SPD policy in those increasingly infrequent instances when force is deployed, the Monitor finds that SPD is in initial compliance with Paragraphs 69 to 90 of the Consent Decree. The significance and importance of this finding cannot be understated, as this report makes clear. It represents a singular and foundational milestone on SPD’s road to full and effective compliance – and represents Seattle crystallizing into a model of policing for the 21st century. A. SPD Officers Use Less, and Less Significant, Force Much of the quantitative data is promising and suggests trends consistent with SPD officers interacting with subjects and using force differently than they did just a few years ago. Of the many insights that the reporting and tracking of force allow, six facts are particularly notable. • Overall Use of Force Rates Are Down – Both Over the Past 28 Months And Compared to the DOJ Investigation Period. Use of force rates trended down over 8 Dkt. 3-1 ¶ 184. 2 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 7 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 this report’s July 2014 through October 2016 study period. Indeed, use of force decreased by 10 percent from the first half of the study period to the latter half. Thus, as the new use of force policies have been become more and more enmeshed in the fabric of the Department, force has appeared to go down. This downward trend is particularly notable when compared to determinations of the DOJ’s 2011 investigation, it appears that there has been a net decrease of 743 incidents – a 60 percent reduction – in the number of moderate- to higher-level uses of force (socalled Type II force, Type III force, and officer-involved shootings) in the 2014-2016 period analyzed.9 Of the 2,385 force incidents, only 39 – or 1.6 percent – involved Type III use of force, the most significant and serious type of force, including fifteen officerinvolved shootings. • Less-Lethal Instruments Are Used Infrequently – With Baton Use Dramatically Declining. SPD’s use of less-lethal weapons (which constitute a kind of Type II force) is relatively infrequent. With respect to one such less-lethal tool, the baton, the decline in use has been dramatic. In 2011, the DOJ investigation concluded that “SPD officers too quickly resort to the use of impact weapons, such as batons,”10 which included finding that a single officer had used his baton 12 times in a 14-month period. For the 28-month period studied for this report, all of Seattle’s officers combined used their batons just 23 times. This is a noteworthy finding. The frequency of Taser use also declined – from approximately 14 incidents per month from January 2001 through December 201011 to an average of 7 incidents per month between July 2014 and August 2015. There is a correlation between the use of the Taser and the presence of either (a) individuals experiencing a behavioral crisis (as defined by policy) or (b) exhibiting signs of drug or alcohol impairment, though not in crisis. Specifically, in nearly all of the incidents in which a Taser was deployed (67 of 73, including Type I incidents), the subject was assessed either to be experiencing a behavioral crisis event (42 incidents) or impaired by alcohol or drugs (25 incidents). As explored in the Monitor’s assessment about crisis intervention, use of force in true crisis events is rare (less than 1-2 percent), but it is worth noting that SPD officers seem to be using the Taser almost entirely when a subject has been affected by a behavioral crisis or substance abuse issue. Because SPD’s force reporting policies and systems were markedly different and substantially less rigorous in during the 2009 to 2011, the true number of pre-Consent Decree force incidents is likely higher, which would increase the magnitude of the decrease. 10 2011 Findings Letter at 4, 11-12. 11 “Taser Use Update,” Seattle Police Department, May 2011. 9 3 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 8 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 These declines in the use of less lethal weapons together directly alleviate one of the key patterns of unconstitutional behavior that the DOJ found in its 2011 investigation, and thus are supportive of initial compliance. • Low-Level, Type I Force Incidents Spiked Initially and Continue to Make Up a Large Portion of All Force Used. The number of low-level, Type I force incidents has, on the other hand, generally increased over time. The Monitoring Team found an average monthly increase of 4 percent per month in Type I force incidents from August 2014 to August 2015. The Team hypothesizes that this increase is at least partially due to changes and improvements in reporting this type of force, which was not reported or logged prior to the Consent Decree. For instance, more than half (55 percent) of Type I cases included the use of restraints (handcuffing), and 44 percent included no force other than handcuffing. A large proportion of Type I cases appear to encompass instances where a subject reported that the handcuffing caused pain. This initially-increasing number of Type I force incidents could reflect that officers are increasingly able to apply deescalation and tactical skills to reduce the number of incidents that might otherwise have involved a higher level of force incident – using more Type I force because they are using less higher-level force. Comparing the first part of the study period (July 2014 through August 2015) with the latter part (September 2015 through October 2016), the number of Type I incidents went down – though Type I incidents account for a greater percentage of incidents overall. What is clear from the Type I trends – both the early spike and the subsequent increase in proportion – is that, when force occurs, it happens increasingly at the lower end of the force spectrum. • The Monitoring Team also analyzed how frequently individual officers used force. SPD officers reported using force of some level in a total of 2,385 incidents between July 1, 2014 and October 31, 2016 or an average of about three (2.8) incidents per day. Most (80 percent) involved no higher than low-level, Type I force. Viewed in the context of the 759,383 unique incidents that officers reported, at least as logged in the Department’s Computer Aided Dispatch (“CAD”) database, very few of SPD “contacts” involved any degree of force at any time.12 Even with expanded or increased reporting, the use of force is an unusual event. The Typical SPD Officer Uses Force Very Infrequently. Although this is a notable statistic, the number of incidents in CAD is an imprecise number that is known to be potentially inflated in some respects – because it includes a host of incidents in which it would be unlikely that officers would encounter force – and yet under-counting the overall contacts that SPD officers made during the time period because they do not include voluntary or consensual encounters with the public. 12 4 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 9 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Across the 4,272 individual applications of force by officers – a number that is higher than the overall number of force incidents because multiple officers applied force in the same incident – an average SPD officer used reportable force, of some level, in approximately 3.3 incidents over the course of the 28-month evaluation period. This average (3.3 incidents per officer) may actually be higher than the typical officer because the average that was computed considered all, rather than just patrol, officers. Nonetheless, of officers who did use force, about one-third were involved in just one incident and a vast majority of the remaining officers were involved in just one to three incidents. This suggests that the “typical” SPD officer uses force very infrequently. • A Small Group of Officers Use Force More Than Their Peers, But This Group Who Used Force More Do Not Use Different, or More Serious Force, Than SPD Officers Who Used Force Less. A relatively small number of SPD officers – 109 officers, or about 8 percent of all SPD officers and 13 percent of all SPD officers who used force at least once – were involved in eleven or more use of force incidents each. This included the officer who reported being involved in some 49 use of force incidents. In all, these 109 officers accounted for 40 percent of all force reported by SPD during the 28-month period.13 However, looking at the type of force used by officers, these 109 officers who use force more often are not distinguishable from those who use force less frequently. That is, in the aggregate, the relatively small group of SPD officers who used force the most did not use different, or more serious, force than SPD officers who used force less. • The Monitoring Team Found Some Racial Disparity in the Population of ForceIncident Subjects But Found No Statistically Significant Disparities With Respect to the Type or Severity of Force Used. The Monitoring Team identified some divergence between the racial makeup of the population and the racial makeup of force subjects. This potential disparity – analyzed in crude, Census population terms rather than more sophisticated analyses for the sake of simplicity – is notable if found to be true using more sophisticated statistical techniques and would need to be meaningfully unpacked, evaluated, and scrutinized going forward. However, the Monitoring Team concludes that, although there may be some disparate impact established by aggregate data with respect to use of force, there are no statistically significant disparities with respect to the type or severity of force used. Put differently, in terms of the overall Type (Type I/Type II/Type III) or severity of force used, there were no statistically significant differences noted in terms of what type of force was applied across subjects of different races.14 Thus, although non-white subjects may be overrepresented vis-à-vis the population, a subject’s race does not appear to predispose him or her to more or less serious force. It must be noted, however, that within the levels of force, it appears that Note that some unique incidents or events may be counted more than once, depending on the number of officers using force during the incident. 14 Chi-square = 19.9, df=15, p=.18 13 5 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 10 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 SPD officers are more likely to point firearms at historically-underrepresented than White subjects but are more likely to go hands-on with White subjects. Because nothing immediately obvious about the circumstances of the interactions reviewed in the Monitoring Team’s qualitative assessment suggested reasons why pointing a firearm at Black, Latino, and Asian subjects was more reasonable or necessary than for White subjects, the Monitor encourages more study by SPD, the Community Police Commission (“CPC”), and the anticipated Inspector General In sum, the overall statistical decrease in the frequency and type of force SPD officers use is suggestive of officers using force more strategically or being more likely to de-escalate force to the point of mitigating the need for force to be used in the first place. The Monitor’s previous assessment of SPD’s crisis intervention capacity and performance suggested as much, as well.15 Together, these aggregate data trends strongly suggest that the Department that the DOJ’s 2011 investigation found – where officers would escalate even minor offenses, particularly with persons in crisis – has changed in fundamental ways.16 There is no definitive way of identifying the total number of times overall where it was plausible for justified force to be used, but where officers de-escalated the situation to the point in which no reportable force in fact was employed. Nonetheless, there is also some highly encouraging quantitative and reliable anecdotal evidence of increased use of de-escalation, which this report also details. B. Force Has Gone Down Without Officer Injuries Going Up Officer injuries are flat to slightly down over the study period, although the decrease is not statistically significant, based on SPD injury and hospitalization data. Accordingly, officer force has gone down without any increases in officer injury. It appears to the Monitoring Team, then, that the decreased use of force has not placed officers at any higher risk or made officers less able or willing to use force to defend themselves from threats or harm. C. Force Has Gone Down Without Crime Going Up The Consent Decree seeks to ensure that constitutional policing occurs in a context where both officers and the public are kept safe. Consequently, we evaluated crime data – and the relationship between that data and use of force – to consider whether a decreasing incidence of force might be negatively affecting the ability of the Department to keep the Seattle community safe. Fifth Systemic Assessment, Dkt. 272, at 17 (finding that “SPD is almost always handling crises with a high level of skill and avoiding the unnecessary use of force in difficult situations”). 16 Dkt. 1-1 (2011 Findings Letter) at 11-13. 15 6 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 11 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 By most measures and accounting for seasonal trends, crime in Seattle appears relatively flat overall across the study period, with property crimes flat to slightly down and personal crimes showing a mild uptick. However, across all of overall crime and the separated personal and property crime categories, there is no obvious correlation between the use of force and crime incidence. In fact, not only does it not appear that decreased use of force has been associated with increased crime, but it is actually the opposite: officers have used the most force when crime has been the highest in Seattle. Therefore, an analysis of SPD crime data and use of force data lead to the conclusion that the decreases in force that have occurred over time have not been associated with increases in crime. Although the concept of public safety can be measured in many different ways, the failure of the data to establish a relationship of force going down while crime goes up – and, indeed, establishing the opposite relationship of more force occurring whenever more crime happens to be occurring in the City, all in a context where overall crime is stable – gives the Monitoring Team confidence that the reforms of the Consent Decree are not compromising community welfare and public safety. D. Officer Force Is Typically and Sufficiently Consistent with Law and SPD Policy The aggregate trends and statistical analyses of data on force, officer injury, and crime are critical factors for considering whether SPD is in compliance with the Consent Decree. However, the Consent Decree does not require that the number of use of force incidents or a particular type of force necessarily go down. In part, this is because it is at least possible that officers might use force across relatively fewer incidents but use such force in a manner contrary to the Department’s policies or the constitution when they do. The Consent Decree accordingly required that SPD implement force policies that take hold in practice – across Seattle’s communities, across incidents, and across time. Thus, the overall numbers do not, by themselves, establish whether the Decree’s force requirements have become effective in practice. Therefore, a qualitative review of a statistically significant sample of individual force events themselves is needed. To this end, the Monitoring Team conducted an in-depth, structured qualitative review of force investigation files for force incidents that occurred in the July 2014 through October 2016 study period.17 The Monitoring Team’s experts used a structured assessment instrument to guide their reviews of the sampled force cases. It must be emphasized that the Monitoring Team’s qualitative evaluations of force were not exercises in second-guessing or “Monday morning quarterbacking.” This Monitor, and this Monitoring Team, have unwavering respect for the men and women of the SPD who often face tense and dangerous situations and who are routinely called upon to interact with individuals who For both parts of the time period, the Department of Justice and City of Seattle elected to conduct a collaborative review of cases that the Monitoring Team randomly selected to review. 17 7 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 12 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 have been let down, left behind, or forgotten by the social service, mental health, educational, and criminal justice systems. When confronted with a threatening, silent, or resisting subject, police officers do not have the option that others in different spheres of public life might to shuffle subjects along, look the other way, or pass the buck. The Monitoring Team’s reviews of force incidents, as with its other inquiries and activities, actively sought to account for this reality. As the same time, the Team did not hesitate from identifying instances where SPD officers performed counter to SPD’s Consent-Decree-required and Court-approved use of force policies. Across the more recent half of the study period, the Monitoring Team found that officers used force that was consistent with SPD policy more than 99 percent (99.27 percent) of the time. This included using only force that was necessary under the circumstances more than 99 percent of the time (up from 85 percent in the earlier half of the study period). Force was likewise proportional and reasonable in the same more than 99 percent of force incidents (up from 88 percent in the earlier time period). And officers also complied with the duty to de-escalate in 99 percent of cases where that duty was applicable (up from 81 percent in the earlier time period).18 Focusing only on intermediate-level Type II and serious Type III force from the morerecent 14-month period, which is analogous to the moderate and serious uses of force identified and analyzed for numerical purposes in the 2011 DOJ investigation, nearly 96 percent (95.7 percent) of force incidents were consistent with SPD policy.19 The Monitoring Team concludes that many of the issues identified in the DOJ’s investigation with respect to the application of force have been eliminated or, otherwise, substantially eliminated. The Monitoring Team did not identify any force incidents that implicated the prohibitions against using force on individuals who had solely verbally confronted officers. In contrast, the DOJ found using force in violation of a person’s First Amendment Rights was one of the key patterns that formed the basis of their conclusion SPD had violated 42 U.S.C. 14141.20 Likewise, DOJ found that SPD officers routinely used excessive force against individuals who were already under physical control, restrained or rendered helpless.21 The Monitoring Team now finds that SPD officers are, appropriately and consistently with the new policy, generally not applying force to handcuffed or restrained subjects, though the Monitoring Team When analyzing force under policy, reasonableness, necessity, and proportionality are not separate inquiries; they are inextricably intertwined and dependent on each other. They are, in most instances, facets of the overall analysis: force that is not proportional is also, by definition, probably not reasonable or necessary; force that is not necessary is thus probably not reasonable or proportional; force that is not reasonable is probably not proportional or necessary. They are called out, and defined, separately in policy to ensure that each aspect of the inquiry is considered in the overall determination as to whether the force is within policy – but they are not necessarily independent inquiries. 19 The 95 percent confidence interval around this estimate is 0.3 percent to 8.2 percent. 20 2011 Findings Letter at 16-17. 21 Id. at 15-16. 18 8 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 13 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 identified a handful, but not statistically significant number, of instances that gave it concern, and those will be discussed below. Although a vast majority of cases involved force consistent with SPD policy, the Monitor nonetheless did find some instances where officer performance was not consistent with SPD policy. In other cases, Monitoring Team reviewers identified certain tactical issues that, even while not establishing the force as contrary to policy, were inappropriate or problematic. For instance, in some cases – which, encouragingly, tended to be clustered in the first half of the study period – SPD officers placed themselves at substantially higher risk by closing the space between, or by affirmatively initiating contact or pursuit of a subject on their own – without the benefit of backup, resources, or strategic cover and concealment, sometimes affirmatively escalating the situation. Although the Monitor appreciates that, in some circumstances, officers will need to take immediate action to intervene where a subject is presenting an immediate threat of harm to himself or others, this finding is consistent with the Team’s analysis of SPD aggregate data, which indicate that more serious uses of force (Type II and Type III) were more likely to be initiated through officer observations than by being specifically dispatched to an incident. Throughout Part V of this report, the Monitor and his Team highlight various issues like this, make recommendations, and provide technical assistance to further reduce the incidence of problematic or out-of-policy force. Although some incidents that the Monitor reviewed still involved force contrary to policy or the Consent Decree’s requirements, the compliance rates are high and, as noted, have been steadily increasing over time. Further, incidents involving problematic use of moderate to serious force are, in the larger context of SPD encounters, substantially infrequent. Again, of the 759,383 unique incidents to which officers responded or on-viewed during this review period, less than 0.00003% involved a greater than moderate (Type II) level of force. By extension, an exceedingly small fraction of SPD interactions involved force that was deemed inconsistent with policy. Even if a human organization could somehow attain perfection, the Consent Decree does not require a perfect police department. Instead, it requires a police department that, among other things, has rigorous policies governing the use of force; provides high-quality training and supervision to officers on those policies; uniformly reports, investigates, reviews, and critically analyzes all use of force incidents in a thorough, fair, and unbiased manner; and has robust, overlapping mechanisms for critical self-analysis that inspire constant innovation and internal improvement. The Monitoring Team is reassured that, in every case in which it determined officers had not complied with SPD’s use of force policy during the later half of the study period, the Force Review Board identified the force as out of policy. It therefore appears that when an officer performs in manner contrary to SPD’s use of force policy, the Department is able to catch and correct the error. Although this does not mitigate the exposure of a subject to unconstitutional or out-of-policy force, it goes a significant length toward ensuring that poor performance and human error can be addressed meaningfully such that the 9 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 14 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 performance is isolated to one incident – rather than existing as a part of a wider pattern or practice of deficient performance that is allowed to fester through inattention or indifference. *** The City of Seattle reaching initial compliance with the core provisions of the Consent Decree relating to use of force is a testament to the hard work, dedication, commitment, and performance of many over an extended period of time. Chief Kathleen O’Toole, and her command staff, have worked tirelessly since she became chief in June 2014 to implement comprehensively the forcerelated provisions of the Consent Decree. Her constant promotion of the new use of force policies as good for the men and women of the police department and the Seattle community has done much to cultivate buy-in and ongoing application by the rank and file. In the same vein, Mayor Ed Murray has been an unstinting champion of new approaches and robust training for officers on force – emphasizing the importance of de-escalation, the professional development of police officers, and the ongoing strengthening of ties between SPD and the diverse communities that the Department serves. However, the credit for this major milestone goes first and foremost to the men and women of the Seattle Police Department. This assessment is fundamentally an analysis of their performance over time. Their ability to meaningfully and effectively implement the use of force policies and apply the related use of force training on the streets of Seattle – while facing the unpredictable challenges that are part and parcel of law enforcement – is worthy of substantial praise. The Monitoring Team’s respect for the many committed professionals who serve, protect, and partner with the community to solve problems and make Seattle a place for all to live, work, and play is unwavering – but burnished all the more by the demonstration of strong trends and good performance. As this report elsewhere makes clear, police officers in Seattle are frequently tasked with addressing individuals and situations that the rest of the social service fabric has failed, left out, or left behind. Their ability to innovate, change approaches, and change the course of the Department while addressing these fundamental duties is commended. Although this report finds compliance on a core area of the Consent Decree, the Monitoring Team continues its work on its other major pillar: discriminatory policing. The Monitor’s assessment of SPD stop activity, involving substantial quantitative and qualitative analysis, is forthcoming. Meanwhile, the scope of this report is both fundamental and expansive. As such, there are limits to how much of the Monitor’s inquiry can be adequately summarized in this introductory summary. Consequently, the remainder of the report presents the Monitoring Team’s quantitative and qualitative findings relating to SPD’s use of force in detail. Part I provides important background on the development of the standards governing this review. Part II summarizes the results of the Monitoring Team’s quantitative analyses of aggregate data on SPD use of force. Part III analyzes data on officer injury and hospitalization. Part IV examines crime 10 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 15 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 trends. Finally, Part V discusses the findings of the Monitoring Team’s qualitative review of use of force incidents. 11 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 16 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Part I. Overview of Use of Force & Purposes of the Assessment I. Background on Use of Force A primary issue addressed by the Consent Decree between the United States and City of Seattle addressing the Seattle Police Department is the use of force. In December 2011, the United States Department of Justice found “that SPD engages in a pattern or practice of using unnecessary or excessive force, in violation of the Fourth Amendment to the United States Constitution.”22 Among other things, the investigation found that SPD officers: • • • • • Used force “in an unconstitutional manner” at an unacceptable rate; “[T]oo quickly resort to the use of impact weapons,” with the use of batons frequently unnecessary or excessive; Too frequently “escalate situations and use unnecessary or excessive force when arresting individuals for minor offenses,” especially individuals experiencing a behavioral crisis (including those experiencing the effects of mental illness or drug and alcohol intoxication); More often than not used excessive force against a single subject in conjunction with other SPD officers; and Too frequently used excessive force against individuals who “talk-back” but otherwise pose no physical danger.23 Among the systemic or structural “deficiencies [that] have led to the above-described pattern or practice of excessive use of force” were inadequate policies and training relating to the use of force, most especially with respect to “specific force weapons, such as the use of OC [pepper] spray, batons, or the ECW [Taser].”24 Likewise, the investigation found “deficiencies in training relating to verbal de-escalation techniques” such that officers were “trained how to win conflict, but not how to avoid it.”25 Although the City of Seattle did not agree with the findings of the 2011 investigation, the City entered into an agreement with the United States referred to as the Consent Decree (the “Decree” 2011 Findings Letter at 3. Id. at 4, 9-15. 24 Id. at 16–17. 25 Id. at 23-24. 22 23 12 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 17 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 or “Settlement Agreement”). The Decree outlined several “principles” that SPD needed to reflect in revised “use of force policies, procedures, and/or training,”26 including the following: • • • • • • “Officers should use de-escalation techniques, when appropriate and feasible, in order to reduce the need for force”; “. . . [O]fficers should de-escalate the use of force as resistance decreases, while staying in control and as safety permits”; “The number of officers on scene may increase the available force options and may increase the ability to reduce the overall force used”; “Officers should be trained that a hard strike to the head with any impact weapon, including a baton, could result in death” and such strikes “should be consistent with policy and training”; “Officers normally should not use reportable force against handcuffed or otherwise restrained subjects unless necessary or reasonable under the circumstances to stop an assault, escape, or as necessary to fulfill other legitimate law enforcement objectives”; and “Officers should not use force against individuals who only verbally confront them and do not impede a legitimate law enforcement function.”27 In addition to revised general policies on force, the Decree required that SPD revise its “weaponsspecific policies, procedures, and training” to “provide guidance for each [specific] weapon’s use” in a manner consistent with the Decree’s general force principles.28 Further, the Decree required that all officers and supervisors receive substantial and ongoing training on use of force policy, procedures, strategies, tactics, and expectations. A. Reforms to SPD’s Use of Force Policy Upon the Court’s approval of the Decree and the appointment of the Monitor, work began in earnest on revising SPD’s force policies to comply with the Consent Decree. The process of revising the Department’s force policies required “many months of extended drafts, redrafts, consideration of recommendations from community members and organizations, and significant negotiation.”29 This process “involved the Department of Justice, City of Seattle, SPD command staff and patrol officers, the two police unions, the Community Police Commission, and the public themselves during a period of public comment.”30 In ultimately evaluating the sufficiency of those policies, the Monitor “consulted police trainers . . . , law enforcement leaders, SPOG [the Seattle Police Officers’ Guild] in Seattle, and law enforcement rank-and-file . . . to make sure that the policies recommended by the Dkt. 3-1 ¶¶ 70-71. Id. ¶ 70. 28 Id. ¶¶ 74–75. 29 Third Semiannual Report at 9. 30 See Dkt. 107 at 1–2. 26 27 13 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 18 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 [P]arties did not compromise the safety of Seattle police officers and the public they serve.”31 The Court approved the new force-related policies on December 17, 2013.32 SPD’s core officer use of force policy “is the embodiment of the Consent Decree . . . [,] provid[ing] officers with clear guidance and expectations consistent with constitutional imperatives.”33 “The basic policy that governs what officers should and should not do in the field with respect to force . . . runs roughly 4 pages in length.”34 Other force-related policies “deal with many important but distinct areas” primarily relating to how force is reported, reviewed, and investigated, as well as special considerations for using various force instruments.35 Any policy “must balance concision with clarity and broad applicability – guiding officers across innumerable unforeseen circumstances yet being specific enough to allow the Department to effectively hold officers accountable for poor decision-making or substandard performance.”36 As the Monitor has previously summarized, four major concerns guide SPD’s use of force policy:37 • Reasonableness. Consistent with the Constitution and Supreme Court guidance,38 force employed by SPD officers must be objectively reasonable – in other words, appropriate and consistent with what a reasonable officer would do in light of all of the circumstances that the officer who used force encountered. • Necessity. SPD officers are authorized to use force only in the absence of “reasonably effective alternative[s].”39 • Proportionality. SPD officers may only use the force that roughly corresponds to, or reflects the totality of circumstances surrounding, the situation, including the presence of imminent danger to officers or others. Accordingly, the more immediate a threat or the more likely that a threat posed by a subject may result in death or serious physical injury, the greater the level of force that may be objectively reasonable and necessary. Officers “should assess and modulate the use-of-force as resistance decreases.”40 Id. at 5. Dkt. 115. 33 Fourth Semiannual Report at 17. 34 Id. 35 Id. 36 Fifth Semiannual Report at 13. 37 This discussion is adapted from the Monitor’s Fourth Semiannual Report at 17–19. 38 Graham v. Connor, 490 U.S. 386 (1989); Scott v. Harris, 550 U.S. 372 (2007); see also Tennessee v. Garner, 471 U.S. 1 (1985). 39 Seattle Police Manual 8.100(1), Dkt. No. 107-1 at 7; see also id. 8.100(4.1) (requiring officers to use only the force necessary to effectuate the lawful purpose of their actions). 40 Seattle Police Manual 8.100(3), Dkt. 107-1 at 8. 31 32 14 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 19 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • De-escalation. SPD’s force mandates that, “when safe and feasible under the totality of circumstances, officers shall attempt to slow down or stabilize the situation so that more time, options, and resources are available for incident resolution.”41 The policy provides a non-exhaustive list of approaches and tactics that officers can strategically employ to de-escalate situations, including “decreasing the exposure to [a] potential threat by using distance, cover, [or] concealment” and “using verbal techniques” to gain compliance.”42 As the Monitor noted in the Fourth Semiannual Report, the requirement that SPD officers de-escalate situations when safe and feasible to do so “represents a significant evolution – and one that was asked for by members of the Seattle community for years.”43 During the 28-month period addressed in this report, SPD’s force policy included de-escalation as “one requirement” of many in the core officer force policy.44 Updates to the force policy approved by the Court in July 2015 created a “standalone de-escalation policy” in the sequence of policies in the SPD Manual that relate to force45 in addition to continuing to reflect SPD’s ongoing, basic commitment to de-escalation in the primary policy addressing when officers may and may not use force in the field.46 Thus, regardless of whether the specific provisions governing de-escalation are part of SPD’s basic officer use of force policy, SPD Manual 8.200 (“Use of Force – Using Force”) or are a standalone policy, SPD Manual 8.100 (“Use of Force – De-Escalation”), the failure to deescalate when safe and feasible to do so constitutes a violation of SPD policy. In some particular instances, some force types or techniques are expressly prohibited – including force used “as a means of retaliation, against individuals who only verbally confront officers, and (except in certain exigent circumstances) against handcuffed or restrained subjects.”47 Other policy provisions “address the use of deadly force (to be used ‘where threat of death or serious physical injury to the officer or others is imminent’), use of force to prevent escape of a fleeing suspect (which reinforces the necessity requirement . . . ), and the provision of medical aid . . . . ”48 Id. Id. at 8-9. 43 Fourth Semiannual Report at 18. 44 Third Semiannual Report at 13. 45 Id. 46 Seattle Police Manual 8.000(2), Dkt. 107-1 at 2 and http://www.seattle.gov/police-manual/title-8---useof-force/8000---use-of-force-core-principles (last visited Aug. 14, 2016) (“When safe under the totality of the circumstances and time and circumstances permit, officers shall use de-escalation tactics in order to reduce the need for force.”); see also Dkt. 115 (order approving consensus use of force policies). 47 Fourth Semiannual Report at 18 (citing Seattle Police Manual 8.100(2), Dkt. No. 107-1 at 7). 48 Id. (citing Seattle Police Manual 8.100(5)-(8), Dkt. No. 107-1 at 8-9). 41 42 15 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 20 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 SPD policies, “consistent with the Decree, set forth ‘different levels of departmental reporting and review that become more rigorous depending on the type of force used,’”49 which are known within the Department as Type I, Type II, and Type III force. Those “categories account for the nature of the force used, the outcome of the actual force used, and – regardless of the actual outcome – the risks that can reasonably be expected to be associated with” a given type or application of force50: • Type I force is relatively minor, minimal, or “low-level physical force.” 51 • Type II force is intermediate-level force that is “reasonably expected to cause physical injury” more significant than minimal or fleeting pain but less significant than something that might reasonably run the risk of causing “great or substantial bodily harm.” 52 • Type III force is the most serious or severe force that actually causes, or is of a type that could be reasonably expected to cause, great or substantial bodily harm, the loss of consciousness, or death.53 The Monitor has previously described how the type of force used dictates the reporting, investigative, and review requirements associated with a given incident – which are the primary ways in which the category or type of force used is important. This assessment at times refers to force types when analyzing data, in order to capture aggregate trends among various applications of force of similar seriousness or severity, and in discussing some specific force incidents, in order to situate a real-world incident in terms of the overall level of seriousness or severity that SPD policy attributes to the incident. Although the Department’s force “reporting requirements . . . are not merely bureaucratic,”54 nothing in SPD’s policies relating to the reporting, investigation, or review of force “guide or First Systemic Assessment at 10 (quoting Dkt. 3-1 ¶ 93). Id. 51 Type I: “Force that causes transitory pain, the complaint of transitory pain, disorientation, or intentionally pointing a firearm or bean bag shotgun at a person.” Complaints of pain associated with handcuffs, regardless of whether due to the application itself or the discomfort of handcuffs even when properly applied, is an example of a recurrent Type I force incident. 52 Type II: “Force that causes or is reasonably expected to cause physical injury greater than transitory pain but less than great or substantial bodily harm, and/or the use of any of the following weapons or instruments: CEW52, OC spray, impact weapon, bean bag shotgun, deployment of K-9 with injury or complaint of injury causing less than Type III injury, vehicle, hobble restraint.” (The acronym CEW stands for Conducted Electrical Weapon, for instance a Taser, the most common brand of this device.) 53 Type III: “Force that causes or is reasonably expected to cause, great bodily harm, substantial bodily harm, loss of consciousness, or death, and/or the use of neck and carotid holds, stop sticks for motorcycles, impact weapon strikes to the head.” (Although discharges of a firearm at a person are currently included in this category, they are tracked separately by the IA Pro database and in this report as officer-involved shootings.) SPD policy also places into this category any use of force that involves serious misconduct or criminal conduct by officers. 49 50 16 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 21 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 constrain an officer’s determination as to whether to use force in the field.”55 Whether an officer uses his or her firearm (high-level force) or points it at a person (low-level force), an officer’s use of force must be reasonable, necessary, proportional, consistent with the duty to de-escalate, and otherwise conform to SPD policy. Thus, even as this assessment provides information about trends among the different types of force, officers are applying a single policy and the same rules regardless of the nature or severity of the force. The use of force policies began to be implemented on January 1, 2014, with officers beginning to face discipline for failing to adhere to new requirements upon completion of an interim training held in March and April 2014. However, the Consent Decree calls for periodic review and reappraisal of the force policies, which provides an opportunity for the force policy to be adjusted in light of actual performance trends and real-world lessons learned.56 The first update of the force policy was approved by the Parties, the Monitor, and the Court in July 2015.57 As with the original use of force policies, stakeholders from across the Seattle community – including members of the CPC and SPD – provided feedback and input in the process. Noteworthy updates included: • • • • Setting forth de-escalation as a standalone policy requirement and section – regardless of whether force does, in fact, end up being used in a given situation (as noted above); Clarification of requirements relating to the application of force to a handcuffed, non-compliant subject in the back of a police car when “reasonable attempts to gain voluntary compliance have failed” and removal of a subject from the car has received supervisor approval; Clarification of some prohibitions on the use of less-lethal instruments; and Revision of SPD’s Manual preface to provide more specific guidance and clearer expectations on what reasonableness entails for an SPD officer policing within the Seattle community.58 SPD is currently engaging in this year’s review of Consent Decree policies, including the officer use of force policy.59 The Monitor and DOJ recently have received from the City proposed revisions, Fourth Semiannual Report at 28 Fifth Semiannual Report at 13. 56 Dkt. 3-1 ¶ 180 (“SPD will review each policy, procedure, training curricula and training manual required by the Settlement Agreement 180 days after it is implemented, and annually thereafter . . . to ensure that the policy or procedure continues to provide effective direction to SPD personnel and remains consistent with the purpose and requirements of the Settlement Agreement and current law.”). 57 Dkt. 225. 58 Adapted from a discussion in the Fifth Semiannual Report at 13-14. 59 See Dkt. 294. 54 55 17 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 22 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 which include input from the CPC. The Monitor and DOJ hope that those proposals are reviewed and perhaps re-submitted in light of the findings herein. B. Changes to SPD Use of Force Training The Consent Decree recognizes that any police department’s policies can only be beneficial if officers have a clear, pragmatic understanding of how to apply those rules on a day-to-day and shiftto-shift basis. The Decree required that SPD provide “use of force training for all patrol and other relevant officers” that addresses, among other topics: • • • • • SPD’s use of force policy . . . ; [P]roper use of force decision-making; [T]he Fourth Amendment and related law; [R]ole-playing scenarios and interactive exercises that illustrate proper use of force decision-making; and [T]he appropriate use of de-escalation techniques.60 As the Monitoring Team has elsewhere discussed, officers also have needed to receive ongoing training on the closely-related issue of interacting with individuals experiencing behavioral crisis.61 After the core policy addressing when officers are authorized to use force, and other force-related policies, were approved by the Court in December 2013, the primary attention of the Consent Decree necessarily shifted to the creation and provision of “high-quality, interactive training” initiatives for officers on the revised force policy that “translate the clear expectations of the use of force policy into everyday officer performance.”62 Seattle Mayor Ed Murray correctly observed in February 2014 that 2014 was “the year of training” necessarily focused on “getting training right” by “[d]eveloping training manuals and programs to translate the policies” created in the Consent Decree into practice across Seattle’s communities.63 Indeed, immediately after approval of the revised force policy, SPD, “with significant input from the Monitor and DOJ,” created an interim use of force training program.64 This program consisted of several components: Dkt. 3-1 ¶ 128. See generally Fifth Systemic Assessment. 62 Fourth Semiannual Report at 19. 63 Liz Jones, “Seattle Police Reforms Shift to Officer Training,” KUOW.org (Feb. 5, 2014), http://kuow.org/post/seattle-police-reforms-shift-officer-training. 64 Third Semiannual Report at 19. 60 61 18 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 23 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 (i) (ii) (iii) A message from the then-Chief of Police introducing the updated force policy; Five electronic learning modules addressing the basic use of force policy; and A one-day, live classroom training covering the new force policy and related reporting requirements.65 In March and April 2014, nearly all sworn and other relevant officers completed the interim training.66 Subsequently, the Consent Decree process worked to develop a multi-year, multi-part training initiative to build on the basic, interim training. This comprehensive use of force training consisted of a number of discrete modules: • Use of Force Core Principles. A two-hour module covering SPD’s revised policies, with particular emphasis on de-escalation and other approaches to force reduction that do not compromise officer or public safety. • Less-Lethal Tools. A four-hour course combining legal and policy principles with hands-on skills training to allow officers to comply with the force policy’s requirement that all officers carry at least one less-lethal tool. • De-Escalation and Contact/Cover Techniques. A four-hour course on using deescalation techniques to defuse volatile situations and prevent situations from getting to the point where force would need to be contemplated. • Threat Assessment and Subject Control. A four-hour course addressing how officers may safely and effectively respond to potential risks posed by subjects in particular positions. • Firearm Skills. A four-hour skills course reinforcing basic firearms skills, including one- and two-handed shooting techniques, flashlight/firearm techniques, and positioning a firearm safely out of the holster. • First Aid. A two-hour skills course providing officers instruction on basic trauma response, consistent with the policy’s requirements to provide medical aid at the scene of force incidents. 65 66 Id. Id. 19 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 24 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • Team Tactics. A four-hour training course addressing rapid intervention techniques, team searches for suspects in buildings, and team response to an “officer down” call. Students needed to demonstrate proficiency in role-playing exercises that included “shoot/don’t shoot” scenarios.67 The Consent Decree requires that SPD continue to incorporate these principles in its yearly training.68 Consequently, SPD officers have continued, since the completion of the initial “comprehensive” use of force training initiative in 2014, to receive significant, integrated, scenariobased use of force training seeking to reinforce and strengthen the basic force training. In 2015 and 2016,69 officers received training related to the use of force that included, but is not necessarily limited to: • Tactical De-Escalation. A four-hour integrated, scenario-based course addressing various proactive de-escalation skills and strategies.70 • Individual Firearms Skills. A four-hour integrated, scenario-based program addressing safe and effective firearms practices, including de-escalation. • Individual Defensive Tactics Skills. A four-hour course emphasizing individual defensive strategies and tactics, including de-escalation. • Integrated Team Tactics. A four-hour course emphasizing team defensive strategies and tactics, including de-escalation.71 • Rapid Intervention Tactics. An eight-hour scenario-based program focused on addressing rapidly evolving, active shooter situations and other similar threats. • Taser X2. An eight-hour scenario-based program, training the new ECW device. Officers have also received ongoing training in the area of crisis intervention, which featured significant instruction on interacting with and de-escalating, where necessary, situations involving individuals experiencing a behavioral crisis.72 Fourth Semiannual Report at 20-22. Dkt. 3-1 ¶ 127. 69 See Dkt. 191 (2015 Training Plan), 277 (2016 Training Plan). 70 See Dkt. 198. 71 See Dkt. 254-1. 72 See Dkt. 145-1-4, 254-5-8. 67 68 20 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 25 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Finally, SPD’s supervisors have received significant training, which includes principles and methods on overseeing officer use of force.73 II. Purpose of the Assessment A. The Role of the Monitor’s Assessments Generally One of the enumerated duties of the Monitor under the Consent Decree is, through compliance reviews, to “verify that all of the substantive reform measures in the . . . Agreement are implemented.”74 For the City and SPD to be considered as having “full[y] and effective[ly]” complied with a given provision of the Consent Decree, they must be certified to have: (a) incorporated the requirement into policy; (b) trained all relevant personnel as necessary to fulfill their responsibilities pursuant to the requirement; and (c) ensured that the requirement is being carried out in practice.75 Since September 2015, the Monitor has been engaged in a focused program of formalized assessments of SPD’s performance across all areas of the Consent Decree.76 The First Systemic Assessment outlined the purpose of the Monitor’s series of formalized assessments: For the Parties, Monitor, and Seattle community to have confidence that the requirements of the Consent Decree are being carried out in practice – not merely on paper, the performance of the Department and its officers must be assessed across a material span of time and a number of incidents. The assessments will more formally gauge whether SPD is where it needs to be in complying with the Decree.77 To date, the Monitoring Team has updated the Court and public on the Department’s progress in the areas of: • • • • • Force reporting; Force investigation; Chain of command reviews and analysis of force incidents; The Force Review Board (“FRB”), which analyzes many force incidents; Community confidence in, and partnership with, SPD, including through a scientific survey; See, e.g., Dkt. 254-2. Dkt 3-1 ¶ 173(a). 75 Id. ¶ 184. 76 Dkt. 317 at 7–9 (identifying all paragraphs of the Consent Decree subject to assessments or not). 77 First Systemic Assessment at 5. 73 74 21 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 26 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • • • • • Investigation of Officer Misconduct by the Office of Professional Accountability (“OPA”); Crisis Intervention; Supervision; Type II Force Investigations; and Early intervention (EIS).78 The formalized assessments are “systemic” because they focus on “whether the Department has the systems, policies, structures, and culture in place” that the Consent Decree contemplates across time and incidents.79 Accordingly, the assessments do not demand that “SPD is uniformly perfect at all times,” as the Consent Decree recognizes that “[n]oncompliance with mere technicalities, or temporary or isolated failure[s] to comply during a period of otherwise sustained compliance, will not constitute [a] failure to maintain full and effective compliance” with a given area of the Decree.80 However, the Consent Decree also requires, and the Monitor will accordingly continue to require, that “there is clear evidence that the various requirements of the Consent Decree are in fact being carried out in practice and are sufficiently manifest throughout the Department.”81 Thus, the Monitoring Team’s “examination of individual cases or data will be done with an eye toward determining what those individual instances say, in the aggregate and overall, about the performance of SPD and its officers.”82 The present assessment, like previous assessments, presents a host of numbers and statistics. No single number, whether purely quantitative or an aggregate summary of qualitative determinations made by the Monitoring Team’s experts, is singularly dispositive, or capable of telling the whole story about SPD’s progress to date. As the Monitor has previously indicated: All readers must be cautioned to note that the general numbers themselves may not tell the entire story. There is no single threshold number that SPD must reach across each and every area that represents initial compliance. For example, an assessment might judge 70 percent of a given type of [incident as consistent with SPD policy] – because the quality of the remaining 30 percent of cases leaving room for improvement but being relatively close to where it should be. On the other hand, even if 90 perfect of [incidents were consistent with SPD policy], there still might not be initial compliance with the relevant provision[s] of the Consent Decree because the remaining 10 percent of [incidents] were wholly inadequate . . . . 83 See Seattle Police Monitor, http://www.seattlepolicemonitor.com (last visited Apr. 1, 2017). Fifth Semiannual Report at 9. 80 Dkt. 3-1 ¶ 184. 81 Fifth Systemic Assessment at 5. 82 Id. 83 Id. at 6. 78 79 22 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 27 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 For this reason, the Monitoring Team and the Parties have analyzed and balanced important and sometimes competing factors in determining whether SPD now is where it needs to be in how, under its new policies, its officers are using force. B. The Purpose of the Use of Force Assessment This report explores the extent to which SPD’s policies, procedures, and training on officer use of force can or cannot be considered to be effective not merely on paper but in practice. That is, the assessment considers – over time and across numerous incidents – whether SPD officers are adhering to the requirements of the Consent Decree, SPD policy, and their training. The Monitor’s First Systemic Assessment, filed with the Court in September 2015, focused on the reporting, investigation, and administrative review of force incidents – and whether the Department had meaningfully implemented updated policies on its internal response to those instances where its officers deploy force. Because it focused on post-incident administrative investigation and review, that assessment necessarily did not evaluate the performance of SPD’s officers across force incidents to consider whether SPD’s officers are, in fact, affirmatively complying with the requirements of SPD policy, the Consent Decree, and the laws of the United States and State of Washington. Undoubtedly, ensuring that SPD has systems and processes in place that rigorously review force incidents and hold officers accountable for deviations from policy and training is an important aim of the Consent Decree. Indeed, the Monitoring Team was heartened to see that the investigations of force, from Type I to Type III, continue to be well done. Although it was not the focus of this assessment, the Monitoring Team found that the investigations were sufficiently well done to permit the Team to evaluate (a) all uses of force, (b) compliance with the duty to de-escalate, (c) all tactical decision-making, and (d) the legal bases for all seizures. The Department, however, must be able to demonstrate, not only that it has systems for identifying and dealing with problematic performance after the fact, but also that the performance of officers across Seattle consistently conforms to the Department’s policies and requirements. Indeed, the Monitoring Team has previously noted that “[f]or the policy changes addressing the use of force to become ingrained like muscle memory, they must be fully implemented in practice by officers – in the field and on a day-to-day basis.”84 Consequently, in the context of this assessment, the Monitoring Team is looking to see whether there is yet sufficient evidence that SPD has not only revised its force policies to conform to the Consent Decree or changed its system of internal review and analysis of force incidents but that, most fundamentally, its officers are affirmatively complying with the rules of the road articulated in SPD’s force policy. 84 Fourth Semiannual Report at 19. 23 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 28 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Likewise, for the Consent Decree’s required training to be considered effective “in practice,” the content of that training must be evidenced in real-world police interactions. To that end, the Monitor’s Fourth Semiannual Report provided four types of evaluation across which the Department’s training initiatives would need to be considered. The Monitoring Team and DOJ have previously evaluated all SPD training curricula since late 2013, with final curricula submitted to the Court for approval. Further, members of the Monitoring Team and DOJ have observed attended inclass training and provided real-time feedback and comments about SPD’s implementation of various training programs. Through these modes of oversight, assistance, and monitoring, the Monitoring Team has previously identified that “there has demonstrably been a sea change in the depth, rigor, and sophistication of SPD’s approach to training officers” – with curriculum more comprehensive and educational techniques substantially more sophisticated.85 Although improvements in SPD’s capacity to develop and provide training according to curriculum that are consistent with best practice were necessary, they are not, by themselves, sufficient. The ultimate test of SPD’s training on use of force is whether it assists officers in complying with SPD’s force policy. Indeed, the Monitoring Team previously noted that one of the central modes of evaluation of adult education programs involves assessment in terms of “behavioral criteria,” or “measures of actual on-the-job performance” that “can be used to identify the effects of training on actual work performance.”86 Put differently, training can be considered effective only to the extent that officers adhere to it and base their performance on it in the real world – across time and across incidents. This report weaves together what the Parties and Monitoring Team originally conceived as three independent, standalone assessments: • Officer Activity. An evaluation of the overall workloads and aggregate types of endeavors in which SPD officers engage in the field over time. • Use of Force Data. A quantitative evaluation of SPD’s aggregate data on the use of force by SPD and its officers. • Use of Force Incidents. A qualitative assessment of incidents in which SPD officers have used force.87 As the Monitoring Team worked on assessing each of these significant areas, the Monitor and Parties agreed that the interrelated nature of each of these areas – what officers are doing overall on Fifth Semiannual Report at 16. Winfred Arthur, et al, “Effectiveness of Training in Organizations: A Meta-Analysis of Design and Evaluation Features,” 88 Journal of Applied Psychology 234, 235 (2003). 87 Dkt. 221-1 at 15–18, 35–36. 85 86 24 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 29 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 the streets of Seattle, aggregate trends with respect to when force is used during officer activity, and in-depth analysis of those incidents in which officers in fact deploy force – make reporting on all three subjects concurrently more comprehensive. Consequently, this report considers the nature and circumstances in which SPD officers use force overall, with discussions of use of force data, and the Monitor’s qualitative evaluations of use of force incidents woven together. 25 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 30 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Part 2. Quantitative Assessment of SPD Use of Force Data I. Methodology A. Sources of Data Analyzed in this Report This report relies on the direct access to three sources of data provided to the Monitoring Team by SPD. The three sources correspond to three core law enforcement database systems that SPD uses: (1) IAPro; (2) CAD; and (3) RMS. IAPro is the department’s database used to track information about officer use of force, citizen complaints and Office of Professional Accountability (“OPA”) investigations, and other risk-related incidents. The information contained in the IAPro is useful for understanding which officers used force, in what circumstances, which types of force were applied, and the effectiveness of the use of force. For instance, IAPro contains information on the conditions of the encounter, including whether the subject was impaired. IAPro also contains demographic information about both officers and subjects, as well as administrative information about the process and progress of the investigation. Tables from the IAPro database were downloaded on February 1, 2017, to include records for the time period from July 1, 2014 to October 31, 2016. Individual tables were downloaded and collated using appropriate joining fields, as described in the document “IA Pro Database Diagram” provided to the Monitoring Team by SPD. The beginning of this period was chosen to capture the beginning of SPD’s application of policies and training implemented in accordance with the Consent Decree. It is worth emphasizing that the time period covered by this report begins with a relatively early period in the Consent Decree process. The use of force policies governing much of the behavior described in this report became effective January 1, 2014. Training to those policies began in the spring of that year on the new reporting obligations.88 In-person training on the substance of the force policy began later that year and continued into 2015.89 The end of this period was chosen in order to increase the likelihood that investigations and reviews of use of force incidents had been completed by the date the tables were downloaded. The central table of the IAPro database is INCIDENTS, and within this table is a field INCIDENT_CODE. Use 88 89 See Dkt. 194 at 23. Id. at 24; Dkt. 212 at 17. 26 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 31 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 of force entries are flagged in the field INCIDENT_CODE, and only records flagged “UOF” were analyzed for this report. IAPro data are collected in multiple levels. At the highest level is the individual incident, designated by a GO Number or CASENUM (the two are interchangeable). There are 2,385 unique incidents that involved one or more uses of force by one or more officers in the time period under study.90 Within this level are records for individual officers on each incident, in which officers report their uses of force separately. There are 4,272 separate officer reports nested within the 2,385 incident reports. Finally, each officer reports each type of force separately (i.e., handcuffs, Taser). Applications of different types are reported in 6,171 separate records, nested within officer-level reports. In this report these nested records are aggregated to reflect activity at both the officer and the incident level. CAD is the Department’s database for recording communications, including the date, time, and location of each incident, how the case was received (9-1-1, on-view (or an officer viewing some activity, action, or behavior and initiating a contact with an individual), etc.), the initial case type (i.e., traffic stop, assault), the final case type (i.e., DUI, domestic violence), and the disposition (i.e., arrest, report written). The information contained in the CAD database is useful for understanding department activity, including calls for service, how many calls were dispatched, how many contacts were made, the type of calls, and to some degree, the disposition. During the time period selected, there were 424,735 unique 9-1-1 calls to which officers were dispatched and 334,648 incidents that were called in as on-views by officers in the field.91 It should be noted that these include all calls and incidents and cannot reasonably be considered as the number of times that an officer encountered a situation in which force might have been used. The CAD database is not a reliable source of information about the case disposition, however, as it only allows for one clearance code per incident. Therefore, there is only one record per incident in the CAD database. These can be matched with the IAPro database using an ID number assigned to each record that is also used as the GO_NUMBER. RMS is the Department’s database that contains detailed information about individual offenses from arrests and reports, as well as information about subjects who are booked into King County Jail. The information contained in the RMS database is useful for understanding the offenses to which officers respond, in each case, as well as whether the case included an arrest and booking in the King Reported uses of force that were determined to be “de minimis” have been removed from the analysis. There were 6 cases and 16 individual officer reports that were categorized as de minimis. 91 This number excludes calls handled by Parking Enforcement Officers only, 11,460 incidents to which a detox van was the only unit called to respond, and on-view events created only to document the location of officers on secondary employment in the field. 90 27 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 32 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 County Jail. Where a person is booked, the Department uses NIBRS codes92 to record the offense type. For the period July 1, 2014 through October 31, 2016, there are 303,899 records. Seven percent (20,843) were entered into the database through COPLOG, the reporting system for citizen online reporting and Store Security Officers in the Retail Theft Program, and are omitted from the analysis. Another 108,763 cases contained non-NIBRS codes (i.e., missing persons, collisions, natural death), and are also omitted from the analysis.93 The remaining 174,341 records were analyzed. A case may have multiple records if there are multiple offenses, and according to NIBRS protocol, the first-listed offense in a particular case is the most serious. Restricting the database to those NIBRS-coded records listed as most serious creates an analysis of 155,202 individual records to match with the IAPro database. Following a similar winnowing process with the King County Jail records, the Monitoring Team found 23,423 individual arrests during the same period. As will be discussed later in this report, it is important to note upfront a known limitation of the current RMS. It was designed as a NIBRS reporting tool, not a records management tool. For NIBRS, an event is only counted as an arrest if charges are sent to a prosecutor. Therefore, the “arrest” count in the RMS currently excludes custodial arrests that do not lead to filed charges. This leads to an undercounting of arrests and, when force is compared to arrests, an overstatement of the frequency of force. Most of this analysis relies on the IAPro database, with CAD and RMS information providing important context, and in some instances additional information, to use of force information logged in the IAPro database. The report identifies the specific statistical techniques or analytical approaches used when summarizing the various types of inquiries conducted with the data. B. A Note on the Data’s Limitations The Monitoring Team’s goal is to provide an aggregated snapshot of the Department’s uses of force by combining these data into a single analytical dataset. The complexities of the data, the methods in which the data are stored, and the absence of critical information, however, make it difficult to achieve this goal. For example, SPD has a practice of attaching to the IAPro database a PDF document listing the findings of each use of force review, but the information contained in the PDF document is not extracted or summarized and included as data in the IAPro database. That information is therefore NIBRS codes refer to standardized codes used by the FBI’s Uniform Crime Reporting program. See Federal Bureau of Investigation, Criminal Justice Information Services Division, “National Incident-Based Reporting System (NIBRS) User Manual” (Jan. 2013), available at https://www.fbi.gov/aboutus/cjis/ucr/nibrs/nibrs-user-manual. 93 The reason for the omission is that the FBI codes refer to actions or activity that might be considered as squarely related to the law enforcement of actual or potential criminal misconduct. It is in these contexts that use of force might be reasonably expected occur. It is dramatically less likely that a use of force would occur in the context, for example, of a missing person incident or a traffic collision response. 92 28 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 33 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 unavailable and cannot be extracted for data analysis. Similarly, while the Monitoring Team is able to calculate how many times an individual officer reports using force, it cannot analyze the propensity of an individual officer to use force because the databases do not track individual officer activity. The ARREST table of RMS contains the badge number of the booking officer but not the badge number of the arresting officer. The GO_REPORT contains the badge number of the officer who wrote the report, but not the badge number of the officer who made the arrest. While this information is contained in the narrative reports of the PDFs, it is absent from the data in the databases making it difficult, if not impossible, to analyze in a timely or thorough fashion. These issues and associated recommendations are detailed further in the report. C. A Note on the Availability of SPD Data With respect to data on use of force, SPD routinely posts data about its activities and performance to the City’s open data site – data.seattle.gov – including 9-1-1 response times, crime statistics and maps, in-car video logs, CAD events, and the like. On its website, SPD posts public-facing “dashboards” by which members of the public and researchers across the country can extract and explore SPD data across multiple dimensions. In January 2017, SPD posted its use of force data. The Monitor understands that SPD will shortly be publishing its data on stops, as well. Thus, the data that the Monitoring Team summarizes here is available to anyone with an interest in accessing it. This represents another highly commendable step in SPD’s evolution from a Department that did not systematically collect and analyze information about officer performance to one that collects and publicly disseminates such data as a standard way of doing business. II. Overview of Force Data & Trends The overall incidence of force is in no way, by itself, singularly dispositive of whether SPD is complying with the Consent Decree. For one thing, not all force is the same. A control hold – a lower-level force technique not normally associated with actual or possible subject injury when applied in a manner consistent with training – is not the same as an officer-involved shooting, of which there were just fifteen during the period that this report studied. Second, the fact that SPD officers used force in a given number of incidents does not reveal anything about whether that force was reasonable, necessary, and proportional under the circumstances. Nor does it show whether officers engaged in affirmative techniques and tactics to de-escalate the situation and resolve the situation with less significant, or no, force. Part III of this report considers the Monitoring Team’s discussion of just these issues. Nevertheless, decreasing numbers of use of force over time is a positive indicator that SPD officers are deploying force at least less frequently – which might correlate positively with officers using sounder tactics and enhanced de-escalation skills to resolve incidents without needing to use force. 29 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 34 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 With these important limitations in mind, the bottom line: Between July 1, 2014 and October 31, 2016, officers reported using force of some level in a total of 2,385 incidents, or an average of about three incidents per day.94 The number of use of force incidents decreased by nearly 11 percent in the more recent (September 2015 through October 2016) study period. Again, for the purposes of this report, the term “incident” refers to one event or case, regardless of the number of officers using force. Viewed in the context of overall officer and SPD activity, this number indicates that of the 759,383 unique incidents reported in the CAD database and to which officers were either dispatched or which they on-viewed in the field during the time period selected, less than half of one percent (0.3 percent) involved any degree of force of any type. There are some important caveats with respect to this statistic. Specifically, it must be observed – as outlined above – that the number of unique incidents (759,383) necessarily includes a host of incidents in which it would be unlikely that officers would encounter force (e.g., missing person response, etc.). To that end, the number of unique incidents may be overly inflated as far as establishing an appropriate context for considering how much force officers are using per total number of encounters in which force may have been used. On the other hand, the 759,383-incident number could be under-counting the number of overall contacts that SPD officers made during the time period because it does not include voluntary and consensual encounters with residents and members of the public. In short, although it is useful to note that, by one measure, SPD officers used force in 0.3 percent of overall incidents, potential imprecision with respect to calculating the overall number of incidents in which SPD officers engaged in potential force might have been used means that this statistic alone cannot be dispositive in any particular direction. Likewise, even if this measure were more precise, the Consent Decree does not require or expect that the number of use of force incidents decrease. Indeed, the risks of unconstitutional policing are not necessarily ensured by reducing policing95 or the use of force. Over time, it is at least possible that officers might use force across fewer incidents but use such force in a manner contrary to the Department’s policies when they do. Part III of this report explores whether the force that officers do use is employed in a manner consistent with SPD policy. A more detailed breakdown of the character and type of arrests resulting in force, as well as the overall prevalence of force given officer activity, is included later in this report. 95 Fourth Semiannual Report at 11. 94 30 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 35 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 A. Use of Force by Incident Type, Severity of Force, and Across Time Table 1. Use of Force Incidents by Type Use-of-Force Incidents July 1, 2014 - October 31, 2016 July 2014 to August 2015 989 78% September 2015 to October 2016 909 81% Total 1,898 80% Type II 249 20% 199 18% 448 19% Type III 14 1% 10 1% 24 1% OIS 8 1% 7 1% 15 1% Total N 1260 1125 2,385 UOF Type Type I The vast majority (approximately 80 percent) of force incidents involved only low-level, Type I force. As Table 3 sets forth, of these Type I incidents, more than one-half (58 percent) involved handcuffing, while another more than one-quarter (30 percent) involved the pointing of a firearm. This means that a majority of force is low-level force – and a majority of that low-level force is related to handcuffing. There were nearly 11 percent more reported use of force incidents in the first half of the study period (July 1, 2014 to August 31, 2015) than in the second half (September 1, 2015 to October 31, 2016). While the overall difference is due in large part to an 8 percent decline in Type I cases, there was also a 20 percent decline in the number of Type II cases across these periods. By considering the varying force levels of types, we can – for the first time since the Department of Justice’s investigation in 2011 – begin to see whether and to what extent, in the aggregate, officers have begun to use less force than before the implementation of the Consent Decree. As the DOJ stated in its 2011 findings letter, SPD previously reported the use of physical force only if such force “causes an injury, could reasonably be expected to cause an injury, or results in complaint of injury.”96 This definition approximates the current standard for Type II force, which also turns, in part, on the presence of injury; i.e., Type II force is force causing harm greater than transitory pain (the marker for Type I) but less than great or substantial injury (the marker for Type III force). DOJ received 1,230 use of force reports covering the period between January 1, 2009 and April 4, 2011, a period of 28 months.97 As Figure 1 above shows, there were 487 Type II, Type III and OIS uses of force during the 28-month period of study since the policy became effective and training 96 97 2011 Findings Letter at 15 (citing SPD’s prior Department Policy & Procedures Manual 6.240.I.D). 2011 Findings Letter at 4. 31 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 36 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 began in earnest. This represents a decrease of 743 force incidents, or a 60 percent reduction in the use of moderate- to higher-level force, between the 2014–2016 period studied here and the time period addressed by DOJ’s investigation. This is a notable statistic. It must, of course, be observed that force reporting policies and practices between 2009 and 2011 were markedly different and less rigorous than SPD’s current systems – giving no reliable way to say with any degree of certainty whether the 1,230 reports from 2009 through 2011 actually and appropriately reflected all uses of force. The true number of force incidents for the pre-Consent Decree period could well have been higher, as SPD did not have rigorous policies, processes, and systems in place for ensuring the uniform reporting of all force – which would increase the comparative drop. Of the 2,385 incidents in which force was reported, only 1.6 percent (or 39 incidents) involved Type III use of force – the most significant use of force incidents – a number that includes the fifteen officer-involved shootings. In the context of overall police activity, albeit with the caveats previously discussed as to the CAD data, this means that of the 759,383 unique incidents to which officers responded or on-viewed during this review period, less than 0.00003% involved a greater than moderate (Type II) level of force. Figure 1. Use of Force Trends 32 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 37 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Figure 1 displays information about the trends in force incidents by month and type – making clear that use of force across types, from the most serious to the comparatively least serious, have generally been decreasing over time. The Monitoring Team observes that the number of reported Type I incidents has varied systematically over time, rising from July 2014 to September 2015, and declining thereafter. Type I cases did rise and fall during the period under study, peaking in August 2015 (93 cases). From July 2014 through August 2015, Type I cases increased at a rate of approximately 3 percent per month. Over the subsequent 14 months, Type I cases decreased at a statistically significant rate of approximately 4 percent per month, ending the study period with consecutive months of 55, 54, and 56 cases.98 The Monitoring Team hypothesizes that this early rise and subsequent fall in Type I incidents is at least partially due to changes in reporting this type of force, which was generally not reportable prior to the Consent Decree. In order to understand how the variation in different types of force affect the overall variation in Type I force reports, we analyzed the three most common types: handcuffs, pointing a firearm, and personal force. Handcuff-related uses of force increased by approximately 4 percent per month during the period from July 2014 through August 2015, and decreased by approximately 5 percent per month thereafter. In addition, reports of officers pointing firearms at individuals did not increase over time in the first period, but did decrease significantly from September 2015 through October 2016.99 The use of personal force did not vary over time in any systematic fashion. All three of these are highly correlated with the overall use of Type I force, and the frequency of handcuff-related uses of force, firearm pointing uses of force, and personal uses of force explains nearly all of the fluctuation in Type I incidents during the study period.100 To understand how policies and training influence the frequency of Type I force it would be wise to start by studying how it affected the application of these three types. Meanwhile, intermediate, Type II cases declined slightly over the entire period, at an average rate of about 1% per month; the decline is not statistically significant.101 Type III incidents tapered off over time, with no Type III incidents reported at all in four of the last twelve months of the study period. Training to the new force and force-related policies, including the new Crisis Intervention policies, accelerated during 2014 into 2015. However, with correlation insufficient to These were calculated using a linear regression model with an interaction (R2 = .37). The estimates are significant at p<.01. 99 A linear regression with an interaction demonstrated that Pointing a Firearm uses of force vary significantly with time in the latter half of the period (R2 = .31, p<.05). 100 A linear regression leads us to conclude that changes in the frequency of Handcuff-related uses of force, Firearm Pointing uses of force, and Personal uses of force explain the changes of Type I uses of force during the study period (R2 = .90). Once these are accounted for, seasonal fluctuations are not significant predictors of Type I cases. 101 This was calculated using a linear regression model (R2 = .13). The p-value of the estimate is .06. 98 33 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 38 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 establish causation, the extent to which the provision of training effectuated the decrease cannot be determined definitively. B. Use of Force by Location/Precinct As the Monitoring Team has previously observed – and as common sense would suggest – use of force incidents are not spread equally across Seattle’s precincts. Areas and neighborhoods with the greatest number of arrests, reported crime, or the highest numbers of residents living or otherwise spending time there tend to correlate to those areas where officers use force with greater frequency. The number of use of force incidents reported over the study period ranged from a low of 155 incidents in Southwest Precinct to a high of 619 incidents in West Precinct (Table 2).102 Table 2. Use of Force by Precinct and Type Use-of-Force Incidents by Precinct and Type July 1, 2014 - October 31, 2016 Precinct Type I Type II Type III OIS Total West 447 160 9 3 619 North 439 80 5 7 531 East 390 93 6 1 490 South 406 61 1 4 472 Southwest 132 22 1 0 155 84 32 2 0 118 Other As discussed elsewhere in this report, the number of force incidents occurring within each precinct appeared roughly proportional to the number of arrests made in that area. The Monitoring Team notes, however, that the West and North precincts were more likely to report more serious uses of force (Type II or Type III) than the other precincts. Type II and III use of force incidents made up 28 and 20 percent, respectively, of incidents for West and North precincts and between 14 and 17 percent of incidents for other precincts. It bears noting at this point that West and North Precincts include Seattle’s active nightlife areas, are the locations of centralized social service providers, and have the highest numbers of calls for service. North and South Precincts had the highest number of OIS incidents, however, with three incidents in each precinct.103 An additional 118 incidents involving SPD officers were classified in IA Pro as non-precinct specific: Outside of the City (53), Special Ops/SWAT (47), Investigations (10), Demonstration (5), and Traffic/PEO (3). 103 One of the incidents attributed to North Precinct, given the significance of North Precinct officers to the events that unfolded, was a multi-officer incident that originated in West Precinct, leaving open the question of whether the incident is most properly attributed to either West or North. 102 34 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 39 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 III. Force Instruments & Techniques Used by SPD Officers Table 3 contains the rates at which various force instruments, techniques, or options were used during use of force incidents. This data aggregates separate uses or applications of specific force techniques across all incidents, including the separate application of different force instruments or techniques in the same incident. Thus, the total number of individual applications of force summarized in Table 3 exceeds the total number of force incidents reflected in Tables 1 and 2 because, in some force cases, officers used multiple techniques or force instruments during the same incident. According to SPD’s data, and perhaps unsurprisingly, the most common type of force used was that of restraints, particularly handcuffing. (Although handcuffing is not considered reportable force on its own, the use of handcuffing may be reported alongside other types of force or, for Level I cases, where the use of the handcuffs is reported by the subject to cause transitory pain.) More than half (55 percent) of Type I cases included the use of restraints (handcuffing, as use of the hobble qualifies as a Type II restraint), and 44 percent included no force other than handcuffing. Although it was not immediately clear from the data whether such restraints resulted in complaints of pain, the Monitoring Team observed in its qualitative review, discussed in Part III, that a large proportion of Type I cases appear to encompass instances where a subject reported that the handcuffing caused pain. Table 3. Applications of Force Instruments/Options. Applications of Force Instruments/Options by Type of Force July 1, 2014 - October 31, 2016 All Handcuff/Hobbl e Type I Type II Type III OIS 1322 55% 1101 58% 208 46% 12 50% 1 7% Control Hold 826 35% 462 24% 343 77% 19 79% 2 13% Firearm - Point 628 26% 562 30% 53 12% 2 8% 11 73% Personal Weapons 316 13% 114 6% 188 42% 14 58% 0 0% Verbal 309 13% 163 9% 134 30% 10 42% 2 13% Other 72 3% 27 1% 39 9% 2 8% 4 27% Taser 73 3% 5 0% 63 14% 2 8% 3 20% Chemical Agent 34 1% 0 0% 31 7% 1 4% 2 13% Baton/OC/Balls 23 1% 1 0% 21 5% 1 4% 0 0% Firearm - Fire 17 1% 1 0% 1 0% 0 0% 15 100 % After handcuffing, the next most common type of force, accounting for more than onethird (35 percent) of force applications, was the use of a control hold, which includes both 35 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 40 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 “soft” and “hard” takedowns. These uses of force were particularly prevalent in Type II and III incidents (used in more than three-quarters of those incidents). Next, about one-third (30 percent) of Type I incidents involved the pointing of a firearm at a person. One-quarter (25 percent) of Type I cases included no force beyond the pointing of a firearm. When a firearm was in fact fired, the pointing of a firearm by another officer present at the scene was reported in a vast majority (about four-fifths) of instances. Those use of force techniques or instruments generally considered to be more significant, severe, or injurious,104 such as the use of personal force (hard strikes, kicks, and the like) or lesslethal weapons such as the Taser, were used less often, making up a small proportion of overall uses of force. The Monitoring Team found that the use of less-lethal weapons was relatively rare. With respect to one such tool, the baton, the decline in use has been dramatic. The Department of Justice’s 2011 investigation concluded that “SPD officers too quickly resort to the use of impact weapons, such as batons and flashlights.”105 The DOJ investigation could not provide specific numbers that could serve as a reliable baseline for how frequently officers were using the baton because it found that SPD did not reliably report force incidents. However, the investigation did conclude that a single officer had used his baton 12 times in a 14-month period. The Monitoring Team has found that, for the 28-month period studied for this report, all of Seattle’s officers combined used their batons just 23 times. This is a noteworthy finding. The other major less-lethal instrument available to officers is the Taser. As with the apparent sharp drop in SPD’s use of the baton, the numbers for Tasers point to a significant decrease from the pre-Consent-Decree period. “From January 2001 through December 2010,” the DOJ observed, “Tasers have been used in 1,707 incidents, averaging 14 incidents per month. During 2010, Taser use averaged 7 incidents per month, well below the overall average.”106 In the 28 months between July 2014 and October 2016, Tasers were used by 80 officers107 during 73 Type II and Type III incidents – an average of 2.9 deployments of the Taser during force incidents per month.108 There is a correlation between the use of the Taser and the presence of individuals either (a) experiencing a behavioral crisis (as defined by policy) or (b) exhibiting signs of drug or alcohol Indeed, it is on the grounds that some techniques are usually more significant, are more severe, or carry more inherent risk of injury or death even when applied consistent with training and best practice that SPD’s three-level classification scheme (Type I, Type II, and Type III force) for reporting, investigating, and reviewing force is based. 105 2011 Findings Letter at 4. 106 “Taser Use Update,” Seattle Police Department (May 2011). 107 Individual officers may have used the Taser in multiple incidents; each incident is counted separately. 108 Five Taser incidents were classified as Type I force because, although the officer pointed the Taser, the officer did not activate or deploy it. 104 36 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 41 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 impairment, though not in crisis. Specifically, in nearly all of the incidents in which a Taser was deployed (67 of 73, including Type I incidents), the subject was assessed either to be experiencing a behavioral crisis event (42 incidents) or impaired by alcohol or drugs (25 incidents). As explored in the Monitor’s assessment on crisis, use of force in true crisis events is rare (less than 1-2 percent). However, it is worth noting that SPD officers seem to be using the Taser almost entirely when a subject either was affected by a behavioral crisis, under the definition of that term under SPD policy and training, or – even if not experiencing a crisis – identified to be under the influence of drugs or alcohol. About four-fifths of Taser applications occurred during encounters initiated through a 9-1-1 call, with the others occurring during encounters initiated due to officer observation (11 percent), booking (7 percent) or another type of call (3 percent). In about 79 percent of instances where the Taser was used, the force encounter resulted in the arrest of the subject. Using records from SPD’s separate RMS database, the Monitoring Team determined that for the incidents involving Taser applications in which the encounter ultimately resulted in the subject’s arrest, the majority resulted in arrests for assault (57 percent), and that the percentage of assault arrests was the same (56 percent) among those who were classified as CIT/Behavioral Crisis Event. IV. Use of Force by Officer Just as force is not applied uniformly across Seattle, the frequency with which individual SPD officers use force varies. Some officers use a great deal of force on many occasions; others little or none at all. The DOJ investigation previously concluded that “a minority of officers account for a disproportionate number of use-of-force incidents,” and that during a calendar year, “just 20 officers accounted for 18 percent of all force incidents.”109 According to SPD data, there were a total of 4,272 applications of force during the study period by 826 unique officers.110 The Monitoring Team found that some officers used force much more often than others. Across the 4,272 individual uses of force by officers, this means that an average officer used reportable force, of some level, in approximately 3.3 incidents over the course of the 28-month evaluation period. 2011 Findings Letter at 4. Calculated by badge number, with each officer counted once per incident. Three officer entries were not properly entered and were excluded from the calculation. 109 110 37 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 42 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Table 4. Frequency of Officer Force. Use of Force Incidents per Officer July 1, 2014 to October 31, 2016 # Incidents # Officers Percent Cumulative 1 198 24% 24% 2 127 15% 39% 3 95 12% 51% 4 71 9% 59% 5–6 106 13% 72% 7 - 10 120 15% 87% 11 - 20 96 12% 98% 20 - 30 11 1% 99% 34 1 >0% 99% 49 1 >0% 100% Total 826 Of the 826 officers who did use force during the 28-month period studied, the median number of times force was used was 3. About one-quarter (24 percent) of officers were involved in just one incident (Table 4). Another quarter (27 percent) of the officers were involved in two or three incidents. Overall, 87 percent of those who used force were involved in ten or fewer incidents over the 28 months of this study. The remaining 13 percent are spread across a range from 11 to 49 incidents, with most of them clustered in the 11-20 category. Two officers are outliers, with 34 and 49 incidents respectively. Officer use of force is not very consistent over time. The number of times an officer used force in the first 14 months of this study was a not a good predictor of the number of times an officer used force in the second 14 months.111 One obvious question is why the number of force incidents per officer averaged 3.3 incidents when half of all SPD officers used reportable force in 3 or fewer incidents. The answer is two-fold. First, although the sworn membership of the Department hovered around 1,300 during the time period selected, a much smaller number of those employees held positions where they might more foreseeably be involved in circumstances that involved a use of force. For example, during this time period, fewer than 500 officers were assigned to patrol positions at any one time, which would foreseeably be involved in the vast majority of force incidents. Moreover, even within patrol operations, some officers, by virtue of their particular assignments, encounter a greater number of situations in which some level of force may be necessary. Both of these realities are reflected in the data, which show that that a relatively small number of SPD officers – 109 officers, or about 8 percent of all SPD officers and 13 percent of all SPD officers who used force at least once – 111 This is based on a bivariate linear regression model (R2 = .01, b=.10, p<.01, N=825). 38 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 43 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 were involved in eleven or more use of force incidents each. This included the officer who reported being involved in some 49 use of force incidents. In all, these 109 officers accounted for 40 percent of all force reported by SPD during the 28-month period.112 Table 5. Frequency of Officer Force by Type. Use of Force Incidents per Officer by Type of Force July 1, 2014 to October 31, 2016 # Incidents Level 1 Level 2 Level 3 OIS N 1 146 50 1 0 197 2 177 70 2 3 252 3 212 68 2 2 284 4 200 79 3 2 284 5-6 414 145 5 4 568 7-10 715 256 9 10 990 11-20 986 333 17 7 1,343 20-30 198 57 2 1 258 34 21 13 0 0 34 49 41 7 0 1 49 Total 3,110 73% 1,078 25% 41 1% 30 1% 4,259 100% Looking at the type of force used by officers, these 109 officers who used force more often are not distinguishable from those who used force less frequently. That is, in the aggregate, the relatively small group of SPD officers who used force the most did not use different, or more serious, force than SPD officers who used force less. Three-quarters (73 percent) of all officer use of force is categorized as Type I, and 25 percent is categorized as Type II (Table 5).113 There are minor differences between those who use force more frequently and those who use it less, but these differences are not statistically significant.114 Just three officers were involved in more than one Type III incident in the 28-month period. Of the six incidents involving these three officers, four occurred in 2014, one in 2015, and one in 2016. Three of the incidents involved control holds, two involved personal weapons, and one involved handcuffing. No officers were involved in more than one officer-involved shooting in the 28-month period. Note that some unique incidents or events may be counted more than once, depending on the number of officers using force during the incident. 113 These numbers are based upon the maximum use of force reported by each officer in each incident, as opposed to the maximum use of force reported by all officers in each incident. 114 We tested statistical significance using a 2x4 table (2= Officers with 10 or fewer incidents/more than 10 incidents, 4 = the 4 UOF Type categories). N= 4259, Chi-square = 3.32, df=3, p=.35. 112 39 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 44 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Overall, of the 495 individual officers responsible for the 1,149 officer uses of Type II or Type III force (in about 487 unique incidents), 35 officers – about 7 percent of the officers who used Type II or greater force at least once during the 28-month period – used this type of force more than 5 times each. All told, these 35 officers accounted for approximately 23 percent of all Type II-or-greater uses of force. This finding might appear, at a superficial glance, to identify a population of officers who have a preference for serious force. Breaking the data into two 14-month periods tells a different story, however. Twelve officers used Type II or greater force at least 5 times during the period from July 1, 2014 to August 31, 2015, and 6 officers did so in the period from September 1, 2015 to October 31, 2016. These two groups do not overlap. Indeed, 6 of the 12 officers from the first period reported 0 or 1 incidents of Type II or III force in the second period, and 3 of the 6 from the second period reported 0 incidents in the first period. In short, then, we can conclude that a small group of officers is responsible for a substantial portion of SPD’s force – but that there is no one group of officers that appears systematically and systemically more inclined to use more serious force. What this aggregate analysis cannot determine, however, is whether there are individual officers who do use force more regularly than their peers and/or who use more significant force during their more frequent force applications. On the one hand, it may be expected that some officers would be more regularly involved in force incidents by nature of their assignment. That is, it might well be the case that some SPD officers are more likely to find themselves in situations in which they must lawfully and reasonably use force – because their assignment or beat is especially demanding or high-volume or because the officers themselves are particularly active and are more likely to handle more significant or demanding incidents than their peers. Indeed, that would appear to be the case with the officer who reported being involved in 49 force incidents in 28 months (a rate of 1.75 per month) and is assigned to a particularly active watch and beat. On the other hand, it may well be the case that an officer who uses force 10 times in a year might be using more force than necessary or failing to appropriately de-escalate situations in a manner consistent with SPD policy and, crucially, the officer’s SPD peers. Accordingly, the numbers presented here cannot, in isolation, establish whether any of the officers with higher levels of force than their peers have or have not been engaged in force contrary to SPD policy – and care should be taken not to derive any such conclusion. The Monitoring Team has previously emphasized this point: [C]onsider an officer who uses force much more often than his peers. The officer may be following the letter of the law and Department policy but may nonetheless be placing himself in unnecessary risk by failing to adhere to basic tactics or by unnecessarily escalating confrontations. Alternatively, closer scrutiny may show that the officer has been making tactical, sound decisions – but could nonetheless 40 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 45 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 benefit from some additional support from fellow officers and his or her supervisor. Likewise, the scrutiny may merely identify an officer who is exceptionally productive and proactive.115 Consequently, it is crucial that the SPD continue its efforts to determine which officers are using force disproportionately and whether their performance presents a need for additional training or other intervention. Indeed, DOJ previously used this data “to highlight . . . the kind of enhanced supervision and oversight that SPD needs to provide officers . . . . ”116 DOJ did not argue “that the pattern or practice of the use of excessive force is attributed only to those officers who use force more than one time in a year.”117 Instead, DOJ was troubled by the fact that “[o]f all officers who used force 10 or more times, only one officer received administrative review of any kind” and that, among the 1,230 use of force reports DOJ received, only five use of force packets were referred at any level for further review.” 118 For these reasons, it is relevant that written policies have improved,119 the reporting, investigation and oversight of force largely met the Consent Decree’s standards,120 and that SPD’s Early Intervention System is geared toward “track[ing] various aspects of an officer’s performance,” including an officer’s use of force, “and provides a mechanism for notifying supervisors that an officer’s performance may require greater scrutiny – and, if necessary, correction and support.”121 The Monitoring Team is aware of internal, peer-to-peer analyses that have identified “outlier” officers. The upcoming rollout of the Data Analytics Platform will permit the SPD to even more easily, and in real time, determine which officers are responsible for a higher number of force than others in light of activity levels, and be able to take steps to address that disparity, for the safety of both the public and officer. Although it may produce some uncomfortable conclusions from time to time, SPD will need to meaningfully embrace this type of analysis to conduct active risk management. V. Use of Force by Officer Characteristics The characteristics of individuals who are the subjects of force can be both delicate and important. Indeed, in Seattle, there continue to be concerns that police officers may treat individuals of certain Third Semiannual Report at 79. 2011 Findings Letter at 20. 117 Id. at 21. 118 Id. at 19, 21. 119 See Dkt. 115. 120 See First Systemic Assessment; Second Systemic Assessment. 121 Third Semiannual Report at 79. 115 116 41 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 46 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 races, ethnicities, or backgrounds and experiences differently.122 This section explores the characteristics of individuals who are the subjects of use of force incidents. A. Use of Force by Perceived Subject Sex The majority of uses of force, over 75 percent, involved subjects who were male, with female subjects comprising a much smaller proportion of those against who force was used.123 Table 6. Use of Force Subjects by Gender. Use of Force Subjects by Gender July 1, 2014 to October 31, 2016 Gender Total Type I Type II Type III OIS Not Entered 120 5% 93 5% 20 4% 1 4% 6 40% Female 475 20% 430 23% 44 10% 1 4% 0 0% 1783 75% 1372 72% 380 85% 22 92% 9 60% Other 5 0% 3 0% 2 0% 0 0% 0 0% Unknown 2 0% 0 0% 2 0% 0 0% 0 0% 2385 100% 1898 100% 448 100% 24 100% 15 100% Male Total Female subjects were much more likely to be involved in a Type I use of force than in a Type II-or-greater use of force. While they made up nearly a quarter of subjects in a Type I use of force, they comprised just ten percent of subjects in the more serious types of force. B. Use of Force by Race/Ethnicity Overall, about 42 percent of force subjects were people of historically underrepresented racial or ethnic groups, including Hispanics, Latinos, African-Americans, Asians, and Native Americans (Table 7).124 Thus, historically-represented racial or ethnic groups are Dkt. 235 at 9–10 [hereinafter “2015 Community Survey”] (finding that 54 percent of Seattleites believe that Seattle police treat people differently because of their race, with more than two-thirds (69 percent) of African-American residents saying that SPD engages in excessive physical force). 123 This excludes 120 cases where no sex or gender was entered. It should be noted that this information is based on the police officer’s documentation of a subject’s sex for purposes of force reporting – which itself would be based either on external observations, by some official documentation (such as on a subject’s identification), or information within a law enforcement database. It is possible that the biological sex of some individuals might have been different from the gender identity expressed or observed by an officer. 124 The demographics of the use-of-force subjects were derived from what officers reported in IAPro. The Monitoring Team has so far focused its analysis on the incident-level, which could pose problems for assigning sex or race to an incident that has more than one subject, or if officers reporting the use of force in a single incident enter conflicting information. In other words, there might be confusion or misassignment of sex or race at the incident level. One solution, if the proportion of incidents with mixed sex or race was large, would be to report sex and race as “mixed” in those incidents in which there may be ambiguity due to multiple subjects. Another solution is to randomly select one of the subjects as 122 42 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 47 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 represented in the population of force subjects at a rate disproportionate to that of the overall Seattle population. Approximately 41 percent of all use-of-force incidents where a race was entered on a use of force reporting form involved a subject who was White, followed by 33 percent of incidents that involved a Black subject. Approximately 16 percent of subject races were unknown, including 12 (or 31 percent of) subjects involved in Type III incidents.125 Table 7. Use of Force by Type and Subject Race/Ethnicity. Use of Force by Type and Subject Race/Ethnicity July 1, 2014 to October 31, 2016 Use of Force Type Race/Ethnicity Pop. Am Indian 1% 42 2% 34 2% 7 2% 1 4% 0 2% API 14% 89 4% 70 4% 18 4% 1 4% 0 4% Black 8% 794 33% 638 34% 149 33% 4 17% 3 33% Hispanic 7% 109 5% 92 5% 16 4% 1 4% 0 5% 371 16% 299 16% 60 13% 5 21% 7 16% Unknown Total Type I Type II Type III OIS White 70% 980 41% 765 40% 198 44% 12 50% 5 41% Total 100% 2385 100% 1898 100% 448 100% 24 100% 15 100% This potential disparity – analyzed in crude, Census population terms rather than more sophisticated analyses for the sake of simplicity – is notable if found to be true using more sophisticated statistical techniques and would need to be meaningfully unpacked, evaluated, and scrutinized going forward. However, the Monitoring Team concludes that, although there may be some disparate impact established by aggregate data with respect to use of force, there are no statistically significant disparities with respect to the type or severity of force used. Put differently, in terms of the overall Type (Type I/Type II/Type III) or severity of force used, there were no statistically significant differences noted among subjects of different races in terms of what Type of force was applied.126 Thus, although non-white subjects may be overrepresented vis-à-vis the population, a subject’s race does not appear to predispose him or her to more or less serious force. representative of the incident, if the proportion of incidents with mixed sex or race is small. The Monitoring Team chose the second option, as there is ambiguity in only a small percentage of cases. The subjects were reported as solely male in 71 percent of the incidents, and as solely female or unknown in 23 percent. Similarly, subjects were reported as solely White in 39 percent of incidents, and solely persons of color (Black, Native American, Hispanic, Asian, etc.) in 57 percent of incidents. 125 The Monitoring Team finds the omission of this basic information in relatively higher-level uses of force to be problematic and recommends that SPD devote resources to ensuring that such basic data is uniformly collected going forward. 126 Chi-square = 19.9, df=15, p=.18 43 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 48 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Table 8. Use of Force Techniques/Instruments Used by Subject Race/Ethnicity Use of Force Techniques/Instruments Used by Subject Race/Ethnicity July 1, 2014 to October 31, 2016 Race/Ethnicity of Subject Type of Force Total Native American Asian/PI Black Hispanic Unknown White Verbal 13% 12% 16% 14% 7% 12% 13% Handcuff 55% 52% 49% 53% 47% 55% 58% Firearm - Point 26% 24% 34% 32% 39% 28% 19% Personal Weapons 13% 12% 17% 14% 11% 11% 13% 1% 0% 0% 1% 0% 2% 1% Baton/Balls 1% 0% 1% 1% 2% 2% 1% 35% 31% 33% 35% 23% 31% 37% Taser 3% 2% 3% 3% 5% 3% 3% Firearm - Fire 1% 0% 0% 1% 0% 2% 1% Other 3% 0% 4% 2% 5% 6% 3% Hobble 0% 0% 0% 1% 0% 1% 0% 2,385 42 89 794 109 371 980 Chemical Agent Control Hold N Nevertheless, within types of force or levels of severity, the Monitoring Team does note that it found that SPD officers are more likely to point firearms at historically-underrepresented than White subjects but are more likely to go hands-on with White subjects. Specifically, Black, Hispanic/Latino, and Asian/Pacific Islander subjects were significantly more likely to have firearms pointed at them than were White subjects of a use of force (Table 8).127 The differences are stark (19 percent for White subjects versus 34, 32, and 39 percent for Asian/PI, Black, and Hispanic subjects, respectively). In contrast, White subjects were more often the subject of control holds than were Hispanic subjects128 but were not significantly different from subjects in other race or ethnic categories. Neither this data nor the Monitoring Team’s qualitative use of force assessment, summarized in Part IV, reveal any systematic trends about the actions, activities, or behavior of Black, Latino, and Asian subjects that would explain this disparity. That is, nothing about the circumstances of the interactions suggested that pointing a firearm at Black, Latino, and Asian subjects was more reasonable or necessary than for White subjects. This phenomenon deserves more study by SPD, CPC, and the anticipated Inspector General. 127 128 Chi-square = 49.3, df=3, p<.001 Chi-square= 8.9, df=1, p<.01 44 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 49 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 C. Subject Impairment The Monitoring Team has previously cited research that “suggest[s] that as many of 7-10% of U.S. police contacts involve persons with mental illnesses.”129 At the time of the DOJ’s 2011 investigation that led to the Consent Decree, SPD “estimate[d] that 70% of use of force encounters involve . . . populations” of “persons with mental illnesses or those under the influence of alcohol or drugs.”130 For this reason, the Consent Decree’s requirements regarding crisis intervention were geared expressly toward the provision of information, training, and services to individuals experiencing mental or behavioral health challenges.131 The Monitor’s Fifth Semiannual Report found SPD in “initial compliance” with the Decree’s crisis intervention requirements. The Monitoring Team notes that the data discussed in this section come from officer-provided information about a subject’s observed behavior on the use of force report. As the Fifth Semiannual Report summarized, SPD has other mechanisms for logging and tracking information about individuals experiencing behavioral crisis. This discussion should not be confused to be perfectly equating behavioral crisis with short-term drug or alcohol impairment. Many individuals can be impaired by drugs or alcohol in a given situation but not be experiencing a crisis event; likewise, many individuals who have mental health or behavioral health issues may be in crisis but not under the influence of drugs or alcohol. Nevertheless, these observations about subject characteristics are related to the extent that both are recognitions by an involved officer that subjects may respond in a distinctive manner when compared to individuals not in crisis and not impaired. During the 28-month period evaluated for the present assessment, more than half (57 percent) of the 2385 force incidents involved a subject determined by at least one SPD officer to be exhibiting some sign of impairment, whether due to apparent “behavioral crisis” (often associated with mental health issues) (n=450) and/or intoxication by alcohol or another substance (n=920) (Table 9).132 This factor was particularly pronounced during Type II and Type III force, for which a substantial proportion of subjects – 63 and 72 percent, respectively – were determined to be either impaired by drugs or alcohol or, alternatively, experiencing a behavioral crisis. Of the fifteen OIS incidents, six incidents involved a subject who appeared to be experiencing a behavioral crisis. Fourth Semiannual Report at 76 (quoting Stephanie Franz & Randy Borum, “Crisis Intervention Teams May Prevent Arrests of People with Mental Illnesses,” 2010 Police Practice & Research 1, 1 (2010)). 130 Id. (quoting 2011 Findings Letter at 4). 131 Dkt. 3-1 ¶¶ 130–37. 132 For the purposes of this chart, subjects exhibiting signs of both behavioral crisis and intoxication were placed into the “behavioral crisis” category. Impairment is calculated on a per-incident basis, meaning that at least one subject involved in the incident was determined to be impaired. 129 45 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 50 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 When SPD officers interact with an individual who appears to be experiencing a behavioral crisis, they enter information about that interaction and individual into what is referred to as a “crisis template.” The Monitoring Team attempted to link this crisis template data with information logged in the Department’s IAPro force database. After doing so and analyzing the data, it appears as though a relatively small portion of SPD interactions with individuals experiencing mental illness, substance abuse, or other behavioral health challenges involve a reportable use of force. Specifically, of the 9,271 crisis templates entered during the first twelve months of reporting, only 149 – or 1.6 percent – involved a reportable use of force. Of these 149 crisis intervention incidents in which force was reported, the vast majority (113, or 76 percent) involved low-level, Type I force. Thirty-two (22.8 percent) involved a Type II use of force, and only 2 (1.3 percent) involved Type III force. Table 9. CIT/Behavioral Crisis or Impaired Subjects in Use of Force Cases by Type Crisis-Eligible or Impaired Subjects by Type of Force July 1, 2014 to October 31, 2016 Type of Force All Type I Type II Type III OIS CIT/Behavioral Crisis 19% 17% 27% 25% 40% Impaired by Alcohol or Drugs 39% 39% 37% 54% 20% Unimpaired 39% 41% 34% 17% 33% 3% 4% 2% 4% 7% 2,385 1,898 448 24 15 Unknown N The use of force data show statistically significant differences in perceived behavioral crisis or impairment between White subjects and other large groups.133 At least according to force reports reviewed for this report, 67 percent of White subjects were reported to be impaired, either due to a behavioral crisis or alcohol/substance intoxication (Table 10). In contrast, 49 percent and 51 percent of Black and Latino subjects, respectively, were reported to have been impaired. Ultimately, the data show that White subjects were 1.4 times as likely to be determined by SPD officers to be experiencing behavioral crisis than were Black subjects, and that Black subjects were 1.6 times as likely to be unimpaired than White subjects.134 The use of force data show statistically significant differences in perceived behavioral crisis or impairment between White subjects and other large groups.135 At least according to force reports reviewed for this report, 67 percent of White subjects were reported to be impaired, Chi-square p < .001 for comparisons between White subjects and all other subjects, as well as between White subjects and Black subjects. 134 Chi-square = 62.8, df=1, p<.001 135 Chi-square p < .001 for comparisons between White subjects and all other subjects, as well as between White subjects and Black subjects. 133 46 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 51 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 either due to a behavioral crisis or alcohol/substance intoxication (Table 10). In contrast, 49 percent and 51 percent of Black and Latino subjects, respectively, were reported to have been impaired. Ultimately, the data show that White subjects were 1.4 times as likely to be determined by SPD officers to be experiencing behavioral crisis than were Black subjects, and that Black subjects were 1.6 times as likely to be unimpaired than White subjects.136 Table 10. Crisis-Eligible or Impaired Use of Force Subjects by Race. Crisis Eligibility/Impairment by Subject Race July 1, 2014 to October 31, 2016 Race/Ethnicity of Subject Total Native American Asian/ Pacific Black Hispanic Unknown White CIT/Behavioral Crisis 19% 7% 20% 14% 9% 22% 23% Impaired by Alcohol or Drugs 39% 60% 27% 35% 42% 31% 44% Unimpaired 39% 26% 47% 49% 48% 38% 30% 3% 7% 6% 2% 1% 9% 2% Total 100% 100% 100% 100% 100% 100% 100% N 2385 42 89 794 109 371 980 Unknown The use of force data show statistically significant differences in perceived behavioral crisis or impairment between White subjects and other large groups.137 At least according to force reports reviewed for this report, 67 percent of White subjects were reported to be impaired, either due to a behavioral crisis or alcohol/substance intoxication (Table 10). In contrast, 49 percent and 51 percent of Black and Latino subjects, respectively, were reported to have been impaired. Ultimately, the data show that White subjects were 1.4 times as likely to be determined by SPD officers to be experiencing behavioral crisis than were Black subjects, and that Black subjects were 1.6 times as likely to be unimpaired than White subjects.138 It must be noted here that the assessment of whether a subject is experiencing a behavioral crisis or is otherwise under the influence of alcohol or drugs is an assessment that SPD officers are themselves making and recording on their use of force report. Without the ability to link the assessments of non-SPD officers of the same subjects, the Monitoring Team cannot definitively determine if, in fact, a far higher proportion of Black subjects of use of force incidents are not experiencing a behavioral crisis or are unimpaired than White subjects or if, instead, SPD officers are more likely to classify Chi-square = 62.8, df=1, p<.001 Chi-square p < .001 for comparisons between White subjects and all other subjects, as well as between White subjects and Black subjects. 138 Chi-square = 62.8, df=1, p<.001 136 137 47 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 52 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 White subjects as experiencing a crisis or impaired than they are Black subjects. Either may be plausible explanations.139 For those force incidents that involved the use of a less-lethal device against subjects experiencing a behavioral crisis or some type of impairment, the Monitoring Team found that the Taser was used almost exclusively. Of the 73 force cases where a Taser was used, about 58 percent involved subjects were reported to be in behavioral crisis, with another 34 percent of subjects impaired due to alcohol or drug intoxication. In all, just 8 percent – six uses of the Taser – involved a subject who was reported to be unimpaired or not experiencing a crisis. D. Subject Injury Most subjects in force incidents emerged uninjured as a result of the force, with approximately one-quarter of all reported use-of-force incidents resulting in some kind of injury (Table 11). However, and perhaps predictably, more serious uses of force were more likely to result in injury. Indeed, 69 percent of all Type II uses of force resulted in some sort of injury, while 79 percent of Type III force and 80 percent of officer-involved shootings resulted in subject injuries. Table 11. Subject Injury and Hospitalization Rates by Type of Force Subject Injury and Hospitalization Rates by Type of Force July 1, 2014 to October 31, 2016 Type of Force All Type I Type II Type III OIS % Injured 24% 12% 69% 79% 80% % Hospitalized 17% 14% 26% 71% 67% N 2,385 1,898 448 24 15 Even though injuries resulted from about one-quarter of force incidents, 17 percent of force cases overall resulted in the subject being hospitalized due to injury or CIT status. As of 2014, the prevalence of serious mental illness among U.S. adults is higher among Whites (4.4%) than Blacks (3.1%). See National Institute of Mental Health, “Serious Mental Illness (SMI) Among U.S. Adults, http://www.nimh.nih.gov/health/statistics/prevalence/serious-mental-illness-smi-among-usadults.shtml (last visited June 20, 2016). As of 2013, the rate of illicit drug use was slightly higher among Blacks (10.5%) than Whites (9.5%). See U.S. Department of Health & Human Services, Substance Abuse & Mental Health Services Administration, Center for Behavioral Health Statistics and Quality, “Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings” at 26 (Sep. 2014), http://www.samhsa.gov/data/sites/default/files/NSDUHresultsPDFWHTML2013/Web/NSDUHresults2 013.pdf. At the same time, Whites are “more likely than other racial/ethnic groups to report current use of alcohol,” with 57.7 percent of whites reporting such use and 43.6 percent of Blacks reporting use. Id. at 38. 139 48 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 53 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 V. Incident Characteristics This section reports on the circumstances, beyond the characteristics of the subject, in which SPD officers use force. To take a deeper look at the specific incident characteristics that are associated with uses of force, the Monitoring Team collected additional data about the SPD’s activities from outside IAPro, the Department’s main use of force database. Specifically, the Team considered information from the Department’s Computer Aided Dispatch (CAD), arrest, and Crisis Intervention Team (CIT) databases. Using common identifiers – generally the General Offense or “GO number” – information from these various sources was joined to provide additional context and detail about the incidents during which force occurs.140 For the purposes of this report, the Monitoring Team was able to obtain contextual CAD and RMS data from July 1, 2014 through October 31, 2016. Not all GO numbers were present across databases and so they could not be perfectly matched for comparison, thus the total numbers in the tables below may differ from those above. A. How Force Incidents Were Initiated Overall, 57 percent of force encounters were associated with a call for service, with 30 percent initiated through officer observation or “on-view activity.” When broken down by force type, , it appears that Type II uses of force were more likely to be initiated by the officer (on-view) than were Type I or Type III cases. In aggregate, officers tended to use more significant force when they affirmatively elected to initiate the contact than when they were dispatched to an incident. This finding is consistent across both the first and second halves of the study period. Table 12. Use of Force by Source Activity & Force Type Use of Force by Source Activity & Force Type July 1, 2014 to October 31, 2016 Type of Force How Initiated 911 Alarm Call History Call Type I Type II Type III OIS Total 59% 48% 58% 60% 57% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% On-View 27% 43% 29% 33% 30% Telephone, Not 911 14% 9% 13% 7% 13% Total 100% 100% 100% 100% 100% N 1,888 445 24 15 2,372 As shown in the charts, not all of the data could be matched, leaving some cases without additional detail. 140 49 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 54 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 The CAD database contains 1,064,073 records for the 28-month period of this study. About onethird of these (361,510) are initially classified as types of calls that are unlikely to put the officer in a situation in which force might be necessary (118 UOF incidents in 361,510 of these calls, or a rate of 0.03 percent). This leaves 702,563 records in which there is an a priori probability that force will be required. The overall rate of force incidents in this group is 0.42 percent, or one use of force in every 240 calls. However, seven types of calls have a higher probability, and they account for about twothirds of all uses of force (N=1484): Crisis, Domestic Violence, Narcotics, Assault, Robbery, Warrant Services, and Weapons (Table 13). Table 13. Percentage of Initial Call Types in Which Force is Used Percentage of Initial Call Types in Which Force Was Used July 1, 2014 to October 31, 2016 Number Percentage Initial Call Type of Calls UOF Crisis 22,841 0.63% Domestic Violence 24,584 1.14% Narcotics 12,896 1.30% Assault 21,972 2.14% Robbery 4,432 2.28% 10,366 2.36% 1,884 3.93% Warrant Service Weapons As noted above, while a large percentage of reported uses of force occur when the subject is in behavioral crisis or under the influence, previous studies by SPD and the Monitoring Team have found that a very small percentage of cases involving individuals in crisis end up in a reportable use of force. These data further confirm those findings. Crisis calls result in a use of force in less than one percent of incidents. By comparison, calls in which a weapon is reported are six times more likely to result in a use of force (3.93 percent). B. Ultimate Incident Type The original classifications of an incident often change as events unfold or more information is gained. Accordingly, the CAD system provides information on the “disposition” code for each incident, regardless of how it was initially coded. According to these data, about one-third of all incidents and 38 percent of Type II-or-greater incidents were coded as either domestic violence or another type of assault. Cases involving a warrant also made up a significant category. However, perhaps because SPD does not have a “warrant squad,” it appears that some of these “warrant” encounters were initiated, in at least some instances, for other reasons. Another notable category was that of Crisis, which made up six percent of all cases and seven percent of Type II-or-greater cases. 50 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 55 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Table 14. Use of Force by Ultimate Incident Type and Force Type Use of Force by Ultimate Incident Type and Force Type July 1, 2014 to October 31, 2016 Ultimate Incident Type Type of Force Type I Type II Type III OIS Total Assault 17% 30% 50% 20% 20% Domestic 13% 8% 8% 7% 12% Warrant 11% 9% 17% 0% 10% Traffic 8% 4% 0% 0% 8% Narcotics 7% 8% 0% 0% 7% Crisis 6% 7% 8% 0% 6% Autos 6% 2% 0% 7% 5% Other 5% 3% 8% 33% 5% Theft 5% 4% 0% 0% 5% Robbery 4% 4% 0% 7% 4% Burglary 5% 3% 4% 7% 4% Prowler 3% 4% 0% 0% 3% Weapons 3% 3% 0% 20% 3% Disturbance 2% 5% 4% 0% 3% Property 2% 3% 0% 0% 2% Suspicious Circumstances 2% 1% 0% 0% 2% Assigned 0% 1% 0% 0% 0% Intoxication 0% 0% 0% 0% 0% Arson 0% 0% 0% 0% 0% Vice 0% 0% 0% 0% 0% Alarm 0% 0% 0% 0% 0% Demonstration 0% 0% 0% 0% 0% Total 100% 100% 100% 100% 100% N 1,888 445 24 15 2,372 Other categories of minor offenses included traffic offenses, narcotics, property damage, and incidents involving “suspicious circumstances.” These incidents were more likely to result in Type I force than other types of force; specifically, while they made up approximately 18 percent of all incidents involving force, they made up just 15 percent of Type-II-or-greater force incidents. C. Incident Disposition “Incident disposition” refers to the ultimate action that SPD officers took in the incident – including arrest, a police report written but no arrest effectuated, and the like. As might be expected, about three-fourths of use-of-force cases were associated with incidents resulting in a physical 51 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 56 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 arrest of a person, while about 20 percent were associated with incidents where only a report was written (Table 15).141 Table 15. Incident Disposition by UOF Type Incident Disposition by UOF Type July 1, 2014 to October 31, 2016 Type of Force Disposition Type I Type II Type III OIS Total 76% 82% 79% 40% 77% 21% 15% 17% 53% 20% 1% 2% 4% 0% 1% Other 2% 1% 0% 7% 2% Total 100% 100% 100% 100% 100% N 1,888 445 24 15 2,372 Physical Arrest Report Written (no arrest) Assistance Rendered Depending on the situation, such reports were likely related to a psychiatric hold, a crime report where no subject was identified or arrested, or – for OIS incidents – a fatal use of force. About 29 percent of cases where only a report was written included the pointing of a firearm, likely indicating – in at least some cases – the investigation of a potential crime (including 911 calls or on-view contacts) that did not result in an arrest.142 Approximately 39 percent of incidents not associated with an arrest involved the use of restraints, possibly indicating a psychiatric hold or the temporary detention of a subject.143 With the exception of the use of a firearm, which may have been associated with the death of a subject or a subject who was not apprehended, all other use-of-force types were more likely to be associated with incidents resulting in a physical arrest. There were no significant differences in arrest rates between the types of force used. VII. Data Limitations & Recommendations144 In the process of working with SPD’s data for this assessment, the Monitoring Team encountered three difficulties that hampered our ability to analyze and contextualize the use of force in Seattle. The first was the difficulty of matching across various databases. The second was the lack of reliable Approximately 15 percent of cases could not be matched with incident data using the identifiers provided. 142 The data indicate that a firearm was more likely to be pointed during an incident where no arrest occurred than in incidents resulting in arrest. 143 As noted in earlier sections, the documenting of handcuffing does not appear to be consistent, as it is documented in only 60 percent of cases where an arrest was made. 144 These and all other recommendations contained in this report, do not create new obligations under the Consent Decree, or otherwise modify existing obligations that must be complied with to reach initial or full and effective compliance. These recommendations are given in the spirit of the technical assistance: ways in which the Monitoring Team believes that SPD can improve its oversight through data analysis. 141 52 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 57 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 arrest data – again, as a function of the RMS being built as a NIBRS reporting tool. The third was the absence of data necessary to situate uses of force in the context of officer activity. A. The Difficulty of Matching Databases In the most general terms, there are two classes of issues that can be encountered when analyzing data: substantive issues (questions of measurement and interpretation) and logistical issues (accessing information or combining data from different sources to create a single analytical dataset or to analyze a single issue). The first difficulty that the Monitoring Team encountered relates to this second class of logistical issues. Specifically, SPD’s IAPro database contains information on uses of force. However, to use that information in context, it is necessary to place it within the larger universe of police department activity. This currently requires combining databases that were not designed to be combined. At the simplest and most concrete level, matching multiple databases requires uniformity in the fields to be matched. In this report, IAPro was matched with the RMS and CAD databases using the GO Number field. However, this field has different names and different formats in each database. It is known as the GO Number only in the RMS database. In IAPro it is called CASENUM, and in CAD it is the EVENT_ID number. The number is formatted differently in the various databases. In RMS the GO Number is a field of varying lengths, from 5 to 10 characters, a four-digit year followed by a serial number unique to each incident (i.e., 20143646). In IAPro CASENUM is a 10-digit number with a hyphen separating the year from the serial number, with leading zeroes filling in serial numbers smaller than 6-digits (2014-003646). And, in CAD, the EVENT_ID number is an 8-digit number, with a two-digit year separated by a hyphen from the serial number (i.e., 14-003646). Standardizing these fields is the first step toward making it possible to understand uses of force in context, and should be integrated into the data collection protocols so that analysts within and without the SPD can spend their time on the substantive interpretation of information rather than the substantial amount of effort required to link the RMS, CAD, and IAPro systems together. The Monitoring Team hopes that SPD’s ongoing work on its Data Analytics Platform (“DAP”) and its interest in replacing its outdated RMS will provide a long-term solution to these, and other, data challenges. B. The Lack of Reliable Arrest Data An arrest is a significant event. It requires physical contact between a police officer and a suspect in a high-stress situation – and constitutes a core law enforcement and public safety activity. It is troubling, then, that Seattle lacks adequate technology for tracking and analyzing “arrests” as a custodial detention, aside from the more limited universe of bookings. 53 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 58 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 The Monitoring Team sought to understand simply how often uses of force occur in relation to arrests. Unfortunately, none of the sources of data on arrests was comprehensive enough to allow us to investigate this question as deeply as necessary. Specifically, the CAD database contains one record per incident and allows only one entry into the Final Disposition field. Thus, if the incident resulted in two dispositions (i.e., an arrest and assistance rendered), then one of these two dispositions will not be represented in the data. The RMS database is even less comprehensive. It contains information on bookings into King County Jail, but not all people who are arrested by the SPD ultimately end up in King County Jail. C. The Absence of Police Activity Data The Monitoring Team had hoped to establish the propensity of a given officer to use reportable force in a particular situation. To do this the Monitoring Team would need to know how many calls each officer responded to, how many so-called “on-views” the officer initiated, and the circumstances surrounding each. This would require a database that identified the officers responding to each call, their role in the call, whether they had contact with the subject, who the arresting officer was, and various facts related to the incident (time of day, location, offense, race and sex of subject, foot pursuit, etc.). There were two purposes to gathering such data. First, the Monitoring Team could estimate both under- and over-reporting of the use of force by identifying situations in which force was likely to take place, and examining officers’ behavior. This type of analysis would be a useful training tool, and also a mechanism for understanding how reportable uses of force might be reduced through changes in deployment. Second, these data would allow the department to identify officers who use force disproportionate to their overall activity level or assignment. It would reduce the current reliance on the number of reportable uses of force in favor of the proportion of activity which results in a reportable use of force. The Monitoring Team notes again the Department’s interest in replacing its RMS with a more advanced system that would allow the Department to more systematically capture and query data in forms that are more readily comprehensible and able to be aggregated. 54 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 59 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Part III. Officer Safety Part II established that, overall, SPD uses force less often. This includes both a 60 percent reduction in the use of moderate- to higher-level force between the 2014–2016 period studied here and the time period addressed by the DOJ’s investigation and a 10 percent decrease in force from the first part of the July 2014 through October 2016 study period to the latter part. An important concern raised about the changes to policies and procedures mandated by the Consent Decree is that alterations to how force is used by the Department would put officers at greater risk of injury. Increased injuries could occur, for example, if officers are discouraged from using effective force or if the public is emboldened to resist arrest by the knowledge that officers have moderated their practices. This section briefly explores whether, in the period where use of force against subjects was down, officers were at greater risk of being harmed. Overall, we find that officer injuries in the context of use of force incidents were flat to slightly down over the study period, although the decrease was not statistically significant. Consequently, officer force has gone down, as reported in Part II, without any increases in officer injury. I. Methodology Officer injuries and hospital visits related to uses of force are entered by the reporting officer into the centralized IAPro database. As with the time-series use of force analysis in Part II, we aggregated or collapsed the data by month for analysis – so that each month’s worth of data includes the number of officers who reported an injury or a hospital visit. If an officer was injured or hospitalized in more than one use of force incident in a month, each incident was counted as a separate event and contributed to the month’s total. Injuries of any severity were included in the month’s tally. Accordingly, a month’s worth of data could include anything from an officer’s self-report of injury to a long-term hospitalization.145 Incredibly fortunately, no SPD officers were killed in the line of duty during a force incident in the July 2014 through October 2016 time period. II. Findings There were 597 reported officer injuries in the 28-month period of this study, with a median of 20.5 per month. The most and fewest injuries were reported in the first and last months of the study, respectively. Thirty-nine were reported in July 2014, and five were reported in October 2016. The Information was not readily available about the seriousness of the injury, the complexity of medical treatment, or the length of a hospital stay. 145 55 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 60 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 bookending of these values suggests that injuries might have declined steadily over time, but that is not the case (Figure 2.). The number of injuries is flat over time during the July 2014 through October 2016 study period, with a slight downward slope that is not statistically significant.146 Figure 2. UOF Incidents and Officer Injuries by Month Importantly, the most significant predictor, in the aggregate, of the number of injuries is the number of uses of force. A statistical analysis indicates that one officer was injured for every three use of force incidents during the study period.147 This would run contrary, although admittedly not outright refute, the hypothesis that officer safety depends on using force more regularly. This also refutes the possibility that officers being more likely to be injured now in force incidents than they once were is a viable explanation for the phenomenon of flat officer injury and decreased officer force in the aggregate. Put most simply, officers are more likely to be injured when it is more The results of a bivariate linear regression suggests that there is no relationship between the passage of time and the number of injuries (R2 = .08, b=-0.29, p=.15). 147 A bivariate linear regression produces estimates that every use of force incident results in 0.34 officer injuries (R2 = .25, b=0.34, p<.01). 146 56 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 61 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 likely that they need to use force – and the number of force incidents has trended down significantly. Because the number of incidents in which officers are injured goes up and down in a cyclical manner, in the same seasonal manner as crime and other incident rates, and the percentage of incidents in which officers are injured has remained relatively constant across the study period, it does not appear to be the case that anything in the SPD’s force policy is leading to officers being injured as a result of adhering to that policy. Instead, it appears that simply being involved in the type of incident that requires officers to use force puts officers at some, relatively predictable from the statistical sense, risk of injury. Here, too, there is no support for any assertion that the new force policy is leading to officer injury. Figure 3. UOF Incidents and Officer Hospital Visits by Month. There were 200 reported hospital visits related to use of force during the 28-month study period, with a median of 6 per month. The findings are similar with regard to hospital visits by officers related to use of force incidents. The highest number (19) occurred in December 2015, and the 57 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 62 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 lowest number (0) occurred twice, in August and October 2016. As with officer injuries, the incidence of hospital visits appears to have declined slightly over time, but this slope is not statistically significant (Figure 3).148 Also similar to officer injuries, the best predictor of officer hospitalizations is the number of use of force incidents, although the relationship is not as strong. A statistical analysis indicates that there was one hospital visit for every six uses of force incidents during the study period.149 These findings further shed doubt on the contention that officer safety depends on greater or more regular use of force. Force incidents and officer injuries are positively and significantly related. The more force is used, the more officers will be injured. In some ways, this is unsurprising, given the dynamic and physical nature of many force incidents. It also consistent with the idea that methods designed to avoid the use of force (de-escalation, calling for backup) are contributing beneficially to officer safety overall. We note here that this officer safety analysis is limited by a few factors. First, because force was not routinely and regularly reported in the area before the Consent Decree, reliable SPD data on officer injury resulting from force was not readily available in the pre-Consent-Decree era. Accordingly, we cannot easily compare the pre-Consent Decree reality to more recent, post-Consent Decree trends. Second, we cannot at this time explore the size or extent of potentially colluding variables. For instance, the data do not allow us at this time to explore whether Seattle subjects might be more likely to resist arrest than previously and whether that might put Seattle officers at greater risk of personal injury. The results of a bivariate linear regression suggests that there is no relationship between the passage of time and the number of hospital visits (R2 = .07, b=-0.16, p=.17). 149 A bivariate linear regression produces estimates that every use of force incident results in 0.14 hospital visits (R2 = .13, b=0.14, p=.06). 148 58 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 63 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Part IV. Public Safety The Consent Decree seeks to ensure that constitutional policing occurs in a context where both officers and the public are safe. The previous sections of this report have concluded that the overall incidence of use of force by SPD officers has gone down over time. This has occurred without an increase in officer injuries – strongly suggesting that the failure to use force is not occurring at the expense of officer safety or well-being. With respect to public safety, nothing about the Consent Decree or adherence to principles of constitutional policing detracts from, or need be at odds with, effective law enforcement. Through the Consent Decree itself, both the City and United States have rejected the notion that constitutional policing and effective policing are somehow mutually exclusive concepts. Nevertheless, some critics of the Consent Decree raised concerns early on that it would lead to an increase of crime in Seattle. For one thing, they argue that preventing officers from using necessary force would discourage them from being proactive (the “de-policing” explanation). For another, they say that it encourages criminals by changing their risk-reward calculus in favor of more criminal activity (the “rational criminal” explanation). We tested whether evidence of either of these explanations can be found in patterns of crime reported during the 28 months of the present assessment. We started by estimating crime in Seattle using the Department’s RMS database of reported offenses, limiting our study to reports in which the most serious offense was a NIBRS (National Incident-Based Reporting System) Group A crime.150 The data were aggregated into three tables of monthly incidents: all Group A crimes, crimes against persons, and crimes against property. The monthly incident rates were then compared with the monthly UOF rates to test for evidence of de-policing or rational criminals. The month with the lowest number of all reported Group A crimes was February 2015 (5,498), and the month with the highest was May 2016 (6,900). The month in which reported property crime was the lowest was February 2015 (3,052) and the highest month was July 2014 (3,870). Similar incident rates for personal crimes were December 2014 (750) and July 2016 (1,003). It should be NIBRS, a system coordinated by the Federal Bureau of Investigations (FBI), provides a uniform method of categorizing crimes across jurisdictions with different laws and municipal codes. NIBRS Group A Crimes are predominantly crimes against persons or property. Group B crimes are predominantly crimes against society. See https://ucr.fbi.gov/nibrs/2012/resources/crimes-against-persons-property-andsociety. 150 59 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 64 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 observed that there is some evidence of seasonality in these data consistent with observations in other cities and social science literature – with the colder months of shorter days seeing the lowest crime rates and the warmer months with more daylight hours seeing the highest crime rates.151 Thus, the comparison of crime rates in a cold winter month (February) with those in a hot summer month (July) is generally problematic from an analytical perspective. Although the trend lines point to the number of Group A crimes having gone up slightly in numerical terms over the study period after a drop from July to December 2014, Group A crimes at the end of our study period, in October 2016, were nearly identical to the crime levels at the same time two years prior, in October 2014. Thus, simply looking at the aggregate monthly totals, crime in Seattle appears relatively flat when comparing mid-Fall in 2016 to mid-Fall in 2014. Further, there is no obvious correlation between the number of Group A crimes reported and changes in the use of force in Seattle (Figure 4). As noted above, crime dropped from July to December 2014 and then began a very shallow climb to its present level. Uses of force rose from July 2014 to September 2015 and steadily declined thereafter. There is, in fact, a significant statistical correlation between the two, but contrary to predictions, it is positive: as crime goes up, so does use of force.152 If we take into account the seasonal variation noted above, analyzing the months from May through September separately, we find that the correlation between use of force incidents is much stronger, but it is still positive.153 Consequently, not only does it not appear that decreased use of force has been associated with increased crime, but it is actually the opposite: crime is highest when officers have used the most force. Therefore, we conclude that there is no evidence to support either the depolicing or the rational criminal explanations, which would predict a negative correlation (as uses of force decreases, crime increases).154 See, e.g., Janet L. Lauritsen & Nicole White, Office of Justice Programs, Bureau of Justice Statistics, United States Department of Justice, “Seasonal Patterns in Criminal Victimization Trends” (June 2014), https://www.bjs.gov/content/pub/pdf/spcvt.pdf (noting that “[s]easonal patterns are a long-standing topic in both popular and scholarly literature on crime and show how environmental factors, such as temperature changes and daylight hours, might be associated with crime throughout the year”); Jacqueline Cohen, et al, United States Department of Justice, National Institute of Justice, “Estimation of Crime Seasonality: A Cross-Sectional Extension to Time Series Classical Decomposition” (2003) (outlining academic findings on connection between seasonality and crime victimization rates), available at http://www.mv.helsinki.fi/home/amoaning/movies/EBENoriginal/forecasting%20crime/retrievePDF_id= 2003-18.pdf. 152 A bivariate regression of Group A crime and UOF Incidents is significant and positive, but not strong (R2=.15, b=9.7, p<.05). The prediction is that the ratio of force to crime is 9.7 crimes to every use of force, but the uncertainty around that estimate is very large, with the lower bound near zero (0.26). 153 A multivariate regression predicting Group A crime using UOF Incidents and a binary variable representing the months May to September is significant and positive (R2=.43, bUOF=9.7, p<.01). 154 Similar analyses were conducted for each precinct, and our conclusions from these tests are the same. There is no evidence to support either theory that decreasing the use of force increases the rate of crime. 151 60 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 65 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Figure 4. Use of Force Incidents and Group A Offenses by Month It could be argued that the overall, aggregate crime rate – including both personal and property crime – is, to some relevant extent, masking the effects of the Consent Decree if the incidence of force were not to affect all types of crime equally. That is, one might argue that the incidence of personal crimes is more affected by the de-policing or rational criminal phenomena, as personal crimes are more likely to involve subjects that have engaged in personal, physical violence. One might also argue that at least property crimes would be more affected by the rational criminal phenomena, with criminals incentivized to engage in property crimes if or when they know that officer response will be more minimal. To test the association of the overall crime numbers and SPD’s use of force numbers, we divided the Group A crimes into two groups: personal crime and property crime.155 We re-ran the same analysis as above but, this time, used the monthly number of personal and property crimes. In terms of aggregate numbers, property crimes were flat to slightly down between time periods from identical seasonal spans – July 2016 and July 2014, as well as October 2016 and October 155 These are the NIBRS categories. The remainder category is social crime. 61 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 66 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 2014. Personal crimes, comparing the same July 2016 and 2014 and October 2016 and 2014 data points, show somewhat of an uptick. However, and as with the overall Group A crime numbers, there is no obvious correlation between the use of force and crime incidence (Figure 5). The overall crime rate in both domains does not systematically increase as the use of force decreases. Specifically, a statistical analysis of both property and personal crime indicates that neither is a significant correlate of the use of force.156 Further, when the regression analyses include a seasonal variable, only personal crime correlates with the use of force157 – and the correlation is positive. That means that SPD officers are using more force when personal crime is higher. This runs contrary to the predictions of both the de-policing and rational crime explanations, which would predict that crime would be higher where use of force levels were lower. The estimated ratio of uses of force and personal crime is 1:2, such that every use of force is accompanied by a two-incident increase in crimes against persons. It seems unlikely that the causal arrow points in that direction, however, and it is more likely that when personal crime increases, so does the use of force. Put in those terms, this is a more intuitive and natural relationship: increases in criminal activity involving physical violence and personal crime might be requiring that SPD officers interact with a higher number of individuals who might use force somewhere during those interactions. Ultimately, an analysis of SPD crime data and use of force data lead to the conclusion that the decreases in force that have occurred over time have not been associated with increases in crime. Although the concept of public safety can be measured in many different ways, the failure of the data to establish a relationship of force going down while crime goes up – and, indeed, establishing the opposite relationship of more force occurring when more crime happens to be occurring in the City, all in a context where overall crime is stable – gives the Monitoring Team confidence that the reforms of the Consent Decree are not compromising community welfare and public safety. Indeed, as the next section of this report illustrates, the Monitoring Team suspects – though cannot definitively prove at this time with the available data – that the decreasing numbers of use of force over time have been driven by a reduction in inappropriate, unnecessary, or unconstitutional force rather than a reduction in lawful and necessary force vital for crime-fighting. The reason, as this report turns to now, is that the Monitoring Linear regression estimates, R2=.10, p=.09 (property crime) and R2=.08, p=.14 (personal crime). A multivariate regression of Personal Crime per month on the number of uses of force and a seasonal variable is consistent with the finding above that crime and UOF incidents are positively correlated (R2=.50, bPERSONAL=2.1, p<.05) 156 157 62 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 67 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Team has also seen, by reviewing force cases across two distinct time periods – an earlier period soon after implementation of and training on the force policies and a later period running to close to the end of 2016 – that officers, when they use force, are more likely now than they have been since the DOJ investigation to use force only when it is reasonable, necessary, proportional, and de-escalation techniques that would be safe and feasible under the circumstances have been deployed. The Monitor finds that at least a reasonable explanation for fewer force incidents occurring and more force incidents that do occur being appropriate is that SPD officers are refraining from using force where it would be unlawful, unreasonable, and inappropriate in the first place. Figure 5. Use of Force Incidents by Property and Personal Crimes 63 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 68 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Part V. Qualitative Assessment of SPD Use of Force Part II of this report evaluated overall, aggregate data on SPD’s use of force between July 1, 2014 and October 31, 2016. It found that overall officer force is down across that time, with moderate to serious force also down substantially when compared to the Department of Justice’s 2011 findings. Part III found that this decrease of force has occurred even as there have been no increases in officer injury. Part IV likewise found that this reduced use of force has occurred without any corresponding increases in crime. Taken together, these data and factors are encouraging signs that SPD and its officers are using force more appropriately than they did in the past. However, as this section discusses in some detail, a reduced overall incidence of force does not, by itself, establish that SPD has eliminated the pattern and practice of unconstitutional force that DOJ’s 2011 investigation found reasonable cause to believe had existed within the Department. The following section of this report summarizes the findings of the Monitoring Team’s in-depth qualitative assessment of all Type III force incidents and a random, statistically significant sample of all Type II and Type I force incidents involving SPD officers that occurred between July 1, 2014 and October 31, 2016. I. Methodology The methodology that the Monitoring Team used to analyze and evaluate uses of force by SPD officers conforms closely to analytical approaches used in the Monitoring Team’s previous assessments and to approaches used to analyze force investigations and reports elsewhere.158 This section describes what force incidents were evaluated and how the Monitoring Team’s experts reviewed the cases. First Systemic Assessment at 20-24; Second Systemic Assessment at 7-8; Fourth Systemic Assessment at 18-20; see generally Denise Rodriguez King, et al, Community Oriented Policing Services (COPS) Office, U.S. Department of Justice, Collaborative Reform Model: A Review of Use of Force Policies, Processes, and Practices in the Spokane Police Department (2014) at 9, 12 (describing random sampling of use of force reports for analysis “using a 95 percent confidence level and a confidence interval of 5 percent”); George Fachner & Steven Carter, Collaborative Reform Initiative: An Assessment of Deadly Force in the Philadelphia Police Department, Community Oriented Policing Services (COPS) Office, U.S. Department of Justice (2015) at 16 (describing “investigative quality evaluation” of officerinvolved shootings of “randomly selected . . . case files” using a survey instrument “of ‘yes/no’ and Likert scale (1–5 items)” evaluated by “expert, experienced investigators”); U.S. Department of Justice, Letter to Mayor Richard J. Berry re: Albuquerque Police Department (Apr. 10, 2014) at 3 (“review[ing] a random sample of the department’s use of force reports completed by officers and supervisors”). 158 64 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 69 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 A. What Force Incidents Were Reviewed The population of force cases reviewed included all use of force incidents – also referred to within SPD as “force cases” – that occurred between July 1, 2014 and October 31, 2016 (the “study period”). The study period was divided into two parts. An earlier period included cases that occurred between July 14, 2014 and August 31, 2015 and for which any required force investigation had been completed as of December 15, 2015.159 The later time period included cases that occurred between September 1, 2015 and October 31, 2016 and for which any required force investigation had been completed as of November 7, 2016. The bifurcation of the larger study period into halves allowed for a meaningful gauge of progress over time – especially in the context of active use of force training and policy implementation still occurring during the earlier period. The Monitoring Team has previously described the extent to which one force “case” or “incident” might involve multiple uses of force: “[C]ases” and “incidents” refer to investigations of applied force in a given encounter or instance. It does not refer to individual applications of force within those instances. Accordingly, one force “case” or “incident” may involve multiple types or applications of force. For example, a single traffic stop that involved three discrete applications of force by Officer A and two separate applications of force by Officer B would be . . . a single “case” or “incident.”160 When discussing force cases or aggregate force trends, the Department and this report sometimes refer to a force incident or case as a “Type III,” “Type II,” or “Type I” force incident. SPD policy dictates that, “if a case involved more than one level of use of force, it is ‘assigned’ the highest level of force used by an officer for purpose[s] of the investigation.”161 Accordingly, a case in which one officer applied a Taser, and the Taser did not cause injury (Type II force) and another officer applied a “soft” takedown (Type I force) would be classified as a Type II case or incident. Thus, when this report associates a force type with an overall incident or encounter, the type simply refers to the most significant or severe level of force that was used in that incident. A “completed” case reflects that “the chain of command has certified the investigation as complete, and the case has been accepted for review by the Force Review Unit,” which oversees the Force Review Board (“FRB”).162 Thus, the investigation for a case that occurred between July 1, 2014 and August 31, 2015 needed to have been completed as of December 15, 2015 to be included in the population of reviewed cases and the investigation for a case that occurred between September 1, 2015 and October 31, 2016 needed to have been completed as of November 7, 2016 to be included in See First Systemic Assessment at 21 (outlining similar methodology and terms). Id. 161 Id. 162 Id. 159 160 65 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 70 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 the population of reviewed cases. Although this process necessarily excluded a small number of cases that occurred during the time period but did not have completed investigations as of the designated cut-off date, the Monitoring Team has no reason to believe that the characteristics of those technically-excluded cases introduced any systemic bias to the population or sample. For the qualitative review discussed in this report, the Monitoring Team’s experts reviewed all Type III incidents, including all officer-involved shootings. For Type II and Type I incidents, the Monitoring Team considered a significant, random sample of reports. Consistent with the approach used to evaluate the quality of the Department’s internal investigations and reviews of force incidents in the First Systemic Assessment: [W]e reviewed a randomly selected subset of Type II and Type I cases that included a number of cases large enough to ensure, within generally accepted levels of confidence within social science, that the subset was unbiased and representative of the whole set of cases. This random-sampling approach is the best way to ensure that the selected sample represents the population of all use of force [incidents] . . . that occurred during the studied period and that the findings in the sample of reviewed cases can be generalized to the population of all of the force cases . . . . 163 Ultimately, the Team reviewed 75 intermediate-level, Type II cases (46 from the first half of the study period and 29 from the second half). It reviewed 67 low-level, Type I cases (34 from the first half of the study period and 33 from the second half).164 Again, the Team reviewed all 40 completed Id. (internal citations omitted). The necessary sample sizes were determined using a process consistent with the approach summarized in the Monitor’s First Systemic Assessment, First Systemic Assessment at 21-22 & n. 143, in which the desired confidence level for the first half of the assessment was 95% with 10% confidence interval for the Type I and Type II incidents only. As all Type III and officer-involved shooting incidents were examined, the Monitoring Team reviewed all of these most serious incidents, so there was no need to sample. For the latter half of the study period, Type I and Type II cases were sampled using a finite population correction (“fpc”), which was necessary because the population sizes of those incidents are, in the statistical sense, not large. As in the earlier period, Type III cases were not sampled such that all Type III incidents meeting the population definition occurred. The samples were stratified, meaning that incidents within each Type of force were sampled separately. The Type I and Type II samples were designed to achieve a 90% confidence with an interval of 14%. This design is based on a common assumption that the measured outcomes would be in the range of 50% (i.e., a coin flip). However, the further the outcomes are from the 50-50 range, the smaller the confidence interval becomes. When the outcomes are in the tails of the distribution the confidence intervals are much smaller. The tails of the distribution are where the probabilities are less than 10% or more than 90%, similar to the probability of flipping four coins and getting four heads. In this study the outcomes are in the tails, and the reported confidence intervals reflect that and thus diverge from the 14% in the original design. All of these design characteristics (finite population correction, stratification, and weights) were used to create the analytical framework to analyze the data in the statistical software Stata. Finally, it should be noted that, because a sample size was pulled independently per type of force, instead of all force events for the entire universe of force events, the Type I and II cases were sampled at a 163 164 66 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 71 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Type III incidents, including officer-involved shootings (25 from the first half of the study period and 15 from the second half). It should be noted here that this qualitative review did by definition, then, focus on analyzing those instances in which some force was, in fact, applied. Instances in which force perhaps could have been used but in which force was not in reality employed are difficult to regularly or systematically capture and challenging to identify in a rigorous or reliable way. The Monitoring Team is, however, aware of reports, public commendations, and incident summaries in which officers appeared to have used sound tactical de-escalation skills in incidents where no force ultimately needed to be used. Although this necessarily represents anecdotal evidence, here are a few examples of instances where officers were affirmatively commended by persons who wrote to the SPD for their de-escalation skills: De-Escalation Example 1. A community member wrote the following: “I work with outstanding officers from all precincts dealing with difficult cleanups throughout the City. We had a particularly challenging cleanup . . . with RV’s and tent campers that had settled in for months on [a Seattle] Street, which [another agency] had attempted to clear two months earlier which didn’t go well. We rescheduled this cleanup utilizing several departments and resources prepared for what we expected could again be quite contentious. SPD had a large presence and was prepared to do whatever was necessary to clear the area. Officer 1 made several visits days before preparing occupants for what was coming and a few campers moved on. The several campers that remained I was certain would have to be arrested in order to move them. Officer 1 took the lead and clearly stated this area was going to be cleared that day one way or another. The campers were very vocal about their rights and continued to argue, but Officer 1 persisted that they had to move that day. And as long as they kept gathering their things he allowed them time to pack up. As expected there still were numerous heated conversations, but Officer 1 stayed calm keeping his voice down and continued to press them. Several people pushed back to what might have resulted in an arrest if not for [the officer’s] continually reasoning with them. Working with homeless encampments is challenging on many levels . . . . I felt it important to bring the good work . . . to your attention.” De-Escalation Example 2. A community member called to make sure to thank the officers who had responded when she was in crisis and wanted to kill herself. She stated that the officers had come to her location twice. She said that she would not be alive today if it had rate not directly proportional to their overall incidence, particularly as compared to DOJ’s 2011 investigation. The Parties agreed in advance on these methodological approaches. 67 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 72 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 not been for those officers and now having been released from the hospital, she is now nonharming. The woman went on to say that she was very impressed with how the officers interacted with her and ensured that she would be taken care of. In her words, “they didn’t have to care about me the way they did but they treated me like a very deserving person and with respect. I remember one of the officers telling me that I deserved to live and to have a good life, and I could tell he wasn’t faking it but that he meant what he said. The officers were like old friends who knew what to say to me and they did a job from the heart.” De-Escalation Example 3. Community members wrote to express their appreciation for the exemplary work of the Seattle Police Department in de-escalating a potentially lifethreatening situation for one of our patients. A confused patient had used an oxygen tank to break through an 18th floor window at which point he stepped out onto the window ledge. This patient requested the presence of the Seattle Police. Five officers quickly responded and were able to reassure the patient and convince him to step safely back into the building where much-needed treatment could occur. The members commended five East Precinct Officers who provided appropriate, professional and timely assistance. B. How the Force Incidents Were Evaluated Four members of the Monitoring Team reviewed Type III cases. Every Type III case was reviewed by two team members. As with the Team’s previous qualitative reviews, “[t]his two-tiered re[viewing] structure sought to ensure that any unduly outlying determinations would be identified or ‘checked’ by another equally comprehensive review.”165 Six members of the Monitoring Team reviewed Type II and Type I cases. Because the numbers of both Type II and Type I cases were substantially higher than the number of Type III cases and six independent reviewers participated in the review of these intermediate- and low-level force cases, the two-tiered review structure was less necessary. Each Monitoring Team reviewer independently examined the incidents. The Monitoring Team’s experts considered the full investigative file of the force review incidents, which is often referred to as the “packet.” These files were again supplied to the Monitoring Team by the SPD and “included written material, such as officer reports, investigator logs, and supervisor evaluations; video material, including in-car video and private video footage; other images, including incident photographs or pictures of subject or officer injuries; and audio material, such as audiotapes of recorded officer interviews.”166 165 166 First Systemic Assessment at 22. Id. 68 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 73 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 For each incident, reviewers used an electronic qualitative review instrument in which they logged basic information about the incident and nature of force applied and made determinations about the extent to which each individual officer’s performance, and application of force, was consistent or inconsistent with various aspects of SPD’s force policy. Most broadly, reviewers were prompted to make a determination, with respect to each individual application of force applied by each individual officer, about whether the force was objectively reasonable, necessary, proportional, consistent with the de-escalation policy, and otherwise consistent with all force-specific policies. Where reviewers indicated that force was inconsistent with some element of SPD policy, the basis for that determination needed to be thoroughly explained. The specific nature of any policy issues noted needed to be specified, as well as any training, tactical, or equipment concerns noted, regardless of whether the concerns rose to the level of a clear policy violation or not. If the nature of the evidence contained within the force packet did not allow the reviewer to make a definitive determination in one direction or another, reviewers were required to explain what investigative deficiency rendered them unable to determine whether officer use of force was or was not consistent with SPD policy. Again, one objection that might be raised to the Monitoring Team’s approach in the present report is that it focuses on evaluating those instances where SPD officers did indeed use force – and not the potential universe of cases in which SPD officers successfully de-escalated or resolved situations without needing to use any force. However, the Consent Decree requires that SPD’s force policy be effectively implemented across time and types of incidents – and most principally those instances where force is used. Even if SPD officers are performing in a manner consistent with SPD policy in a number of instances in which force is never used, a fundamental inquiry remains whether, when officers use force, that force is or is not consistent with the force principles and requirements set forth in the Consent Decree and codified in SPD policy. Another objection may be that the Monitoring Team’s experts simply want officers to avoid using any force whatsoever and not engage in smart and proactive policing. The Monitoring Team has previously, and repeatedly, emphasized that the Consent Decree, and the specific reforms to which the United States and City of Seattle agreed, is concerned with simultaneously ensuring effective, accountable, and constitutional policing. For instance, when the Monitor described how SPD will look and function when it has fully and effectively complied with all of the Consent Decree’s provisions, it noted that “SPD’s provision of services and internal accountability mechanisms [must] effectively promote public safety and public confidence”: The SPD’s core law enforcement and policing activities [must] promote public safety and officer safety. SPD must reflect[] a commitment to proactive, safe policing consistent with constitutional demands – and an aversion to suggestions that unconstitutional policing should be reduced by reducing policing.167 167 Fourth Semiannual Report at 11. 69 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 74 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Even police use of force that is definitively consistent with constitutional and state laws and sound departmental policy can have fact patterns or video associated with it that are hard to read or watch. In some instances, “[p]olice work is not pretty and when you see situations where officers have to use force, it is not a pleasant video to look at – but it is the reality that police officers do sometimes have to use force.”168 The Use of Force policy itself acknowledges that “[s]ometimes the use-of-force is unavoidable.”169 The Monitoring Team’s six reviewers of force incidents have reviewed, during the course of their extensive careers, thousands of force incidents and investigations. All have substantial background in policing, and two experts have significant experience working as sworn law enforcement professionals. Likewise, the Team’s reviewers understand that the relevant analysis is not what the officer should have done based on the benefit of the “20/20 vision of hindsight.”170 Analyzing police performance, whether to apply constitutional standards or SPD policy, must allow for an appreciation of the fastmoving and high-risk situations in which police may find themselves. As the Supreme Court has outlined: Not every push or shove, even if it may later seem unnecessary in the peace of a judge’s chambers, violates the Fourth Amendment. [The analysis of force] must embody allowance for the fact that police officers are often forced to make splitsecond judgments – in circumstances that are tense, uncertain, and rapidly evolving – about the amount of force that is necessary in a particular situation.171 Consequently, the Monitoring Team considered all determinations about whether a given application of force by a given officer was reasonable, necessary, proportional, and adequately applied strategic de-escalation tactics as safe and feasible from the perspective of what a reasonable officer, under all of the circumstances that the involved officer encountered. The standard, then, was not whether the Team’s force experts believe that they would have done the same thing or whether they personally believed that force was appropriate based on all of the facts available after a full investigation had been completed. Instead, the Monitoring Team considered officer performance and decision-making in light of what a reasonable officer – under the totality of the circumstances that the involved officer confronted or were known to the officer at the time – would have done. “D.C. Police Chief Defends Violent Arrest Caught on Video,” NBCWashington.com (Oct. 27, 2015) (quoting Washington, D.C. Chief of Police Cathy Lanier). 169 Dkt. 107-1 at 2. 170 Graham v. Connor, 490 U.S. 386, 396 (1989). 171 Id. at 397. 168 70 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 75 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 This approach has a few important implications. The first is that it forces reviewers to evaluate the incident not in light of perfection but based on the realistic and reasonable options that an officer would have under the circumstances. This does not mean, however, that the Monitoring Team’s analysis reduced the whole of SPD’s policy down to “reasonableness.” It did not. Instead, the decisions that SPD officers made in the cases that the Team reviewed were analyzed not according to what officers should have done in light of all of the facts that became apparent after the incident or what the Team’s individual reviewers believe that they would, or the involved officers should, have done – but, instead, considered what a reasonable officer would have done under the circumstances presented to the officers involved in each case. The approach likewise does not mean that a reviewer, who was not involved in the incident and thereby not experiencing the moment-by-moment and typically fast-moving situations as the officer did as they unfolded, cannot make definitive determinations as to whether officer performance was or was not reasonable, necessary, proportional, or sufficient with respect to de-escalation. They can. In the Fourth Amendment context, the Supreme Court has repeatedly advanced a mode of analysis in which reviewers consider the nature of the circumstances suggested by the evidence and consider what a reasonable officer would have done in such circumstances.172 The fact that a reasonableness standard in the context of police use of force may not be “capable of precise definition or mechanical application” does not mean that police performance cannot be considered unreasonable or inappropriate in a given case – instead, a reviewer must give “careful attention to the facts and circumstances in each particular case.”173 The second major implication of this approach is that it allows, in the same way that the Supreme Court has instructed with respect to the Fourth Amendment analysis, the Monitoring Team to analyze officer performance in force cases “without regard to” an officer’s “underlying intent or motivation.”174 Just as an officer’s malicious intentions do not transform policy consistent with SPD policy into a violation of law or policy, an officer’s “good intentions” do not justify or excuse force that violated law or policy.175 Finally, application of the standard allows the Monitoring Team to examine the tactics, strategies, decision-making, and performance of officers not merely at the narrow moment at which force is deployed (e.g., a Taser fired, a baton swung, a control hold applied, or the trigger of a firearm pulled) but, rather, “from the time the involved officer(s) begins to engage in police activity relating to the See, e.g., Scott v. Harris, 127 S.Ct. 1769, 1775, 1779 (2007) (finding that, because a “videotape” of a car chase “tells quite a different story” than other evidence in the factual record, no reasonable jury could conclude that an officer’s actions were unreasonable); Graham v. Connor, 490 U.S. 386 (1989). 173 Graham v. Connor, 490 U.S. 386, 396 (1989). 174 Id. at 397. 175 See id. 172 71 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 76 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 incident . . . until the completion of the enforcement activity.”176 Indeed, this is consistent with the force policy’s guidance177 and is the same mode of analysis in which the Monitor has previously indicated that the Department’s FRB has become increasingly skilled.178 Monitoring Team reviewers made separate judgments as to discrete applications of force by the same officer during an incident, as well as to discrete applications of force made by different officers during an incident. Each application of force – whether separate applications of force by the same officer or separate uses of force by multiple officers – must separately and independently conform to constitutional, legal, and SPD policy standards. *** The Monitoring Team’s qualitative evaluations of force incidents to determine if officers acted in accordance with SPD policy are not an exercise in second-guessing or “Monday morning quarterbacking.” This Monitor, and this Monitoring Team, have unwavering respect for the men and women of this police force who often face tense and dangerous situations – and who are routinely called upon to interact with individuals who have been let down, left behind, or forgotten by the social service, mental health, educational, and criminal justice systems. When confronted with a threatening or resisting subject, police officers do not have the option that others in different spheres of public life might to shuffle subjects along, look the other way, or pass the buck. Especially given the high numbers of force incident subjects who were identified as experiencing some behavioral crisis (including mental health, substance abuse, or other behavioral issues) in the time period studied, it cannot be reasonably disputed that SPD officers often bear the burden of encountering the human costs and effects of the tears in the social service fabric elsewhere. At the same time, laws, courts, and SPD policy entrust officers with substantial and necessary discretion:179 Most police activity occurs in private, away from the public’s view. This creates a situation that allows police officers discretion in the way they think about what they see and how they handle those with whom they come in contact.180 Fourth Semiannual Report at 44. Dkt 107-1 at 3 (e.g., “Officers should take reasonable care that their actions do not precipitate an unnecessary, unreasonable, or disproportionate use of force, by placing themselves or others in jeopardy, or by not following policy or training.”) 178 Second Systemic Assessment at 3 (“[FRB] must consider the tactics of all involved officers from the time that they were dispatched or initiated activity through the use of force and securing of the scene, until the time when the involved officers completed their statements,” to determine what elements of officer performance were in or out of SPD policy.”). 179 Calvin J. Larson & Gerald R. Garrett, Crime, Justice, and Society 274 (2d ed., 1996) 176 177 72 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 77 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 No law, court, or policy can prescribe specific rules that would apply to every conceivable circumstance involving all possible subjects of police encounters under any possible permutation of facts.181 Indeed, as former Chief Justice Warren Burger is reported to have observed, “[t]he officer working the beat makes more decisions and exercises broader discretion affecting the daily lives of people everyday and to a greater extent than a judge will exercise in a week.”182 Communities and police departments call on their law enforcement officers to regularly exercise the discretion to: [M]ak[e] choices in light of policy norms, [which] involve[] routine but adaptive choices. In the act of discretion, although the decision maker accepts a framework of values and goals, some aspects of the decision process are unspecified or contingent on circumstances and thus up to the judgment of the individual.183 Because “the effective limits on a” police officer’s “power leave him or her free to make a choice among a number of possible courses of action” in any given situation,184 it is the job of any community and police agency – and, with respect to the Court and this Consent Decree process, the Monitor – to ensure that officers are exercising their discretion in a manner consistent with the guidelines and norms of a department’s policy. With the Consent Decree’s focus on SPD’s use of force policy and training, the Monitoring Team turns its attention to examining whether officers are performing in a manner consistent with the Decree’s requirements and the provisions of the Department’s policies. C. How the Findings Are Summarized After reviewing all cases in the manner described above – again, all Type III and officer-involved shooting incidents and random samples for both the earlier and latter parts of the 28-month study period – aggregate statistics were determined. Importantly, the statistics presented in this section of the report are “weighted” to reflect the actual incidence of various types of force incidents. Because the Monitoring Team reviewed all serious force incidents but instead used representative, random samples of intermediate Geoffrey P. Alpert, et al, National Institute of Justice, “Police Officers’ Decision Making and Discretion: Forming Suspicion and Making a Stop” (Oct. 2004), https://www.ncjrs.gov/pdffiles1/nij/grants/213004.pdf. 181 Linda S. Miller, et al, Community Policing: Partnerships for Problem Solving 46 (11th ed. 2011) (“Police use discretion because no set of policies and procedures can prescribe what to do in every circumstance.”). 182 Karen M. Hess, Police Operations: Theory and Practice 523 (5th ed. 2006). 183 Michael K. Brown, Working the Street: Police Discretion and the Dilemmas of Reform 25 (1988). 184 Larry K. Gaines & Victor E. Kappeler, Policing in America 161 (8th ed., 2015) (“Discretion is at the heart and soul of policing . . . Discretion is when the effective limits on a public official’s power leave him or her free to make a choice among a number of possible courses of action.”). 180 73 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 78 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 and low-level force, summarizing results in terms of what the Monitor reviewed (e.g., “X percent of cases reviewed had a given feature”) would “over-represent” the Type III and officer-involved shooting cases and “under-represent” the lower-level force incidents given how comparatively less common serious force was within the study period. To account for this intentional “over-representation” of more-serious incidents in the group of cases evaluated in a manner that allows for valid extrapolation to all SPD activity throughout the study period, the Monitoring Team’s aggregate results across all reviews were statistically weighted to reflect the actual incidence of each force type within the SPD for the study period. By using this approach, the results can be presented in terms of a percentage of all SPD cases across the time period. II. Findings The Monitor concludes that officers, in an overwhelming majority of instances, are affirmatively and actively implementing the requirements of, and principles embodied in, the Consent Decree and SPD’s revised force policy. Indeed, officer force appeared necessary, proportional, and objectively reasonable under the circumstances – with a number of incidents featuring superior examples of officers strategically de-escalating situations in order to minimize the nature of the threat while potentially mitigating the severity of force that needed to be used. Specifically, in the latter, 14-month half of the 28-month study period, more than 99 percent (99.27 percent) of force used was consistent with SPD policy. The 95 percent confidence interval around this estimate is 0.1 percent to 1.4 percent185 – meaning that the Monitoring Team can say that, if it evaluated every single force case during the period rather than all Type III force and a statistically-significant sample of Type II and Type I force, there would be a 95 percent chance that the percentage of force consistent with policy would fall roughly within 98.6 and 100 percent. This overall number is weighted to reflect the phenomenon outlined in Part II of this report: namely, that a vast majority of SPD force is low-level, Type I force. Focusing only on intermediate-level Type II and serious Type III force from the more-recent 14-month period, which is analogous to the moderate and serious uses of force identified and analyzed for numerical purposes in the 2011 DOJ investigation, nearly 96 percent (95.7 percent) were consistent with SPD policy.186 Again, although the samples for Type I and Type II force were designed to achieve a 90% confidence with an interval of 14%, the further that the actual outcomes are from the 50-50 range (i.e. a case having an equal propensity to be out of policy as in policy), then the confidence interval becomes smaller. Indeed, when the outcomes, as here, are at the tails of the distribution, the confidence interval is significantly smaller. In this study, the reported confidence intervals here reflect the fact that the outcomes are in the tails of the distribution. See n. 159, supra. 186 The 95 percent confidence interval around this estimate is 0.3 percent to 8.2 percent. 185 74 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 79 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Further, where officer force was contrary to policy, SPD’s internal mechanisms for critical selfanalysis and review are regularly identifying them as such. Indeed, in every case that it determined officers to have not complied with SPD’s use of force policy during the later half of the study period, the Force Review Board identified the force as out of policy. Officer compliance with SPD’s use of force policies can be exemplified by the following: • Case 1 (Type II). Officer 1 responded to a call of a naked man running in the street. The call was modified to a man entering a group home and assaulting a resident. Officers 2, 3, and 4 responded as backup. Upon reaching the scene, Officer 1 identified a civilian victim bleeding from the ear and appearing to have a neck abrasion. Officers 1 and 2 encountered the subject in a living room – seated, still nude, and shouting erratically. The two officers backed up and proceeded to talk with the subject. At some point, the subject advanced on Officer 2 with closed fists. Officer 1 applied his Taser and ran a single, five-second cycle. The subject was arrested without further incident. Monitoring Team reviewers tended to find the application of force here reasonable, proportionate, necessary – and thereby consistent with SPD policy. • Case 2 (Type I). Officers approached a suspect in a parked, stolen vehicle. They attempted to block the car and pointed their weapons, believing it to be a high-risk stop. The suspect drove the car at the officers. The officers removed themselves from the car’s path and did not fire their weapons. A pursuit followed but is terminated, per SPD policy, quickly. The subject is eventually arrested after being detained pursuant to a 9-1-1 call, with the male subject reported as cooperative while being taken into custody. The Type I pointing of the firearm was consistent with SPD policy and training regarding high-risk felony stops. The Monitoring Team was particularly impressed by the involved officers refusing – in a manner consistent with SPD policy – to shoot at the moving vehicle and, indeed, affirmatively moving out of the path of the vehicle as it proceeded toward them. • Case 3 (Type I). The involved officer conducted a traffic stop of the mother of a suspect who had an outstanding felony warrant for unlawful possession of a firearm. The officer called for backup and went to the house of the subject’s mother, who had indicated that the subject was inside. The subject’s mother gave officers permission to enter. The officers proceeded upstairs to arrest the subject, holding their firearms at low ready. When the subject opened the bedroom door, the involve officer could not see the subject’s hands. Consequently, she raised her weapon and pointed it at the subject. The subject complied with officer commands and was arrested without incident. 75 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 80 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 This Type I use of force was consistent with SPD’s force policy and with training. The Monitoring Team notes that, early on, some raised concerns that the requirement to report pointing a firearm at a subject might discourage officers from un-holstering their firearm when necessary and as dictated by best practices in law enforcement training. This was one of many incidents that affirmed that officers are not being precluded from performing according to their training – but are, instead, routinely documenting any point of a firearm at a subject. Importantly, the proportion of cases in which all officers complied fully with SPD policy increased – and, from a statistically perspective, at the least remained the same – from the first half of the study period to the later half. As noted elsewhere in this report, force used during the more recent half of the study period is somewhat more probative and reflective of where SPD currently stands with respect to compliance. However, there remain some instances in which the Monitoring Team noted concerns that the performance of at least one officer in a given incident acted inconsistently with SPD’s use of force policy. The following sections summarize the assessment’s findings with respect to compliance with the core elements of SPD’s use of force policy. A. Necessity SPD policy requires that officers use force only when it is necessary – that is, “only when no reasonably effective alternative appears to exist, and only then to the degree which is reasonable to effect a lawful purpose.”187 In the more recent time period, the Monitoring Team found, or could find no documented reason to dispute that, all officers use only the force necessary to perform their duties in more than 99 percent of force cases. This was an increase on the 85 percent of cases where all involved SPD officers complied with SPD policy to use necessary force in the earlier time period. For instance: Case 4 (Type III). One early afternoon, officers were investigating an attempted auto theft. Officers were searching the area when Officer 1 spotted a possible suspect. Before Officer 1 was able to address the subject, the subject turned and advanced on the officer, producing two knives. The subject continued to advance as Officer 1 retreated. The subject, arriving at the driver’s-side door of the officer’s patrol car, dropped his center of gravity slightly, leading officers to believe that he was about to charge him. At the same time, the subject reached for the car door handle and began to open the door. Officer 1 fired his pistol, causing the subject • 187 SPD Manual Section 8.100(1) (2014), Dkt. 107-1 at 7. 76 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 81 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 to fall down. However, the subject then got back up (still in possession of at least one knife) and entered the patrol vehicle. Meanwhile, additional officers responded. The officers gave multiple commands for the subject to exit the vehicle. Officer 2 twice discharged a Taser at him. During this time, the subject attempted to manipulate the patrol car’s patrol rifle mount, and (based on presence of blood recovered after the incident), touched the rifle. Officers approached the subject, pulled him from the car, and took him into custody. Officer 1 in this instance appropriately tried to preserve distance between himself and a subject in possession of two edged weapons. In addition to backing away from the threat, he used verbal commands in an attempt to get compliance. Based on the evidence available, and given the subject’s physical movements, proximity to the patrol vehicle, and ultimate proximity to the patrol officer (approximately 15 to 20 feet), it would not be erroneous to conclude that a reasonable officer would deem deadly force to be necessary at the time that Officer 1 employed it. Officers at the scene communicated well during their repeated, postshooting attempts to get the subject to exit the vehicle. • Case 5 (Type III). Officers responded to a park around midnight pursuant to a 9-1-1 call reporting a prolonged fight. Officers arrived and observed a victim unconscious on the ground and the subject appearing to forcibly remove the victim’s purse. The subject fled on foot. Officers gave chase and caught up to him as he was climbing a fence to escape. Officers pulled him down from the fence. The subject subsequently struck both officers in the face. The involved officers both then punched the subject in the face. The use of force on the subject was within a reasonable officer’s discretion if it was necessary and proportional. • Case 6 (Type I). Officers contacted the subject after a domestic violence report. The subject threatened to jump from an elevated position. Officers persuaded him to climb over a fence and surrender. When officers went to arrest the subject, he pulled away and back toward the fence. Officers used control holds, a soft take-down, and body weight to effect his arrest. The subject was uninjured. Under the circumstances, it was not unreasonable to conclude that officers used the degree of force necessary to get the subject into custody. Indeed, such force was only used when the subject attempted to go back toward the fence – and the associated danger of self-harm from a more elevated position came into play. When deployed, the force was used only for the period of time necessary to effectuate control of the subject and take him into custody. • Case 7 (Type II). Officers 1 and 2 responded to a report that a male subject had threatened to murder someone at a store and said he would be back with something. When he quickly returned, there was concern that the man had armed himself. The officers contacted the subject and asked him to take his hands out of his coat pocket. He refused. Officer 1 grabbed 77 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 82 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 the subject’s wrist. The subject did not comply and attempted to flee the store. A struggle ensued, which ended when Officer 1 handcuffed the subject after a hard takedown. The incident, classified by SPD as a Type II force incident, was captured on the store’s surveillance video. Audio of the incident was captured on Officer 2’s in-car video. The subject sustained a minor facial abrasion and complained of left shoulder pain. When a third officer arrived, the subject complained of handcuff pain. Officer 3 readjusted the cuffs. Assuming that a hard takedown was not avoidable, one could reasonably determine that the force used was necessary in light of the nature of the call and the subject’s refusal to remove his hand from his coat pocket. In a very small number of instances across the 28-month period, officers used force that a reasonable officer under the circumstances would not have determined to be necessary. • Case 8 (Type II). Officers responded to a disturbance call at a hospital where a subject was disruptive and had box cutters and a skateboard as potential weapons. Officers talked the subject into voluntarily relinquishing both. Later, while the subject was being seen by hospital personnel, Officer 1 contacted the Department of Corrections, which requested a detainer. Officer 2, a Taser trainer and Field Training Officer, told the subject that he was under arrest. The subject asked why. Officer 2 ignored the subject and threatened to put 50,000 volts into the subject if he did not comply. The subject, reclining on a gurney, placed one hand in a pocket and purportedly may have balled the other hand into a fist. Officer 2 fired a Taser probe, with one 5-second cycle registering on the device. SPD’s Force Review Board noted that Officer 2 unnecessarily escalated the incident and did not need to deploy his Taser, as the officers were dealing with a subject reclining on a hospital gurney with his hand in his pocket and no access to a firearm. The subject’s actions simply did not present a threat to officers, others, or the subject that warranted the use of intermediate-level force. Rather than explaining to the individual why he was being arrested, the officers made it into a confrontation. This affirmative escalation was unnecessary, disproportionate, and unreasonable. The FRB aptly saw this as an unnecessary Taser deployment precipitated by the inappropriate actions of an officer who should have known better. However, the Board failed to make findings that supervisors in the chain of command missed the issue. • Case 9 (Type III). The subject, reportedly using “spice” (synthetic marijuana), kicked a Seattle Fire Department employee. Multiple SPD officers responded to a call to assist the firefighter. A protracted use of force then occurred that ultimately involved seven officers applying reportable force to the subject. 78 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 83 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 An officer punched the subject in the face and delivered multiple knee strikes to the subject’s abdomen. This force was classified as Type II force. The necessity of the knee strikes was not established by available evidence. Further, the decision by the officer to continue delivering knee strikes after they proved ineffective in achieving the purported goal of causing the subject to release his arms did not appear necessary. It should be noted that, although the officer claimed that he continued to strike the subject because he feared the subject might be accessing a weapon with his hands, video of the incident did not capture the officer communicating to other officers on the scene as to this belief. B. Proportionality SPD policy requires that an SPD officer’s force be proportional to the “totality of circumstances surrounding the immediate situation, including the presence of an imminent danger to officers or others.”188 That is, “[t]he more immediate the threat and the more likely the threat will result in death or serious injury, the greater the level of force that may be proportional, objectively reasonable, and necessary to counter it.”189 Put differently, the quality and severity of the threat posed by the subject dictates the quality and severity of the force response that an officer may deploy. A force response or application that is substantially more significant or severe than a more minimal threat posed by a subject may not be proportional under the circumstances. Looking at cases from the later 14-month portion of the 28-month study period, all officers were determined to have used force that was proportional to the threat and/or resistance posed by the subject in 99 percent of cases. This showed continued improvement over the earlier time period, where the Monitoring Team found that all SPD officers complied with SPD policy not to use disproportionate force in 88 percent of cases. For example: • Case 10 (Type II). Officers responded to a domestic violence call, where a female subject had injured a male victim. Officers approached the subject, who was initially sitting in the middle of the street. The subject attempted to flee on foot as a team of officers began to surround her. Officers initially attempted to gain voluntary compliance. The subject then assaulted an officer by punching him. Officers then effectuated a hard takedown, with an officer who apparently bore the brunt of the landing sustaining a broken rib. Following a brief struggle on the ground, the subject was handcuffed. The subject sustained minor cuts to her arm. One could conclude the responding officers acted as a team and seemed well-coordinated and professional in their handling of the incident. 188 189 SPD Manual Section 8.100(1) (2014), Dkt. 107-1 at 7. Id. 79 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 84 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • Case 11 (Type II). Officers observed an unattended stolen car. The subject attempted to enter the car but fled when officers approached. The officers gave chase on foot, encountering the subject a short time later in a backyard. Officer 1 pointed his gun at the subject because he could not see the subject’s hands. At this point, the subject ran toward officers, elbowing them as he tried to break through between them. As he proceeded to do so, he grabbed Officer 1’s equipment belt –heightening the risk that the subject could seize control of Officer 1’s firearm. Officer 1 responded by punching the subject in the face, with Officer 2 striking the subject in the ribs. The subject continued to struggle and remain actively aggressive. Officer 2 fired his Taser twice, which incapacitated the subject and allowed officers to take the subject into custody. The Monitoring Team cites this case here as an illustration of officers managing to resolve a situation with an actively aggressive subject, who reached for Officer 1’s equipment belt, through the use of intermediate-level, Type II force – effectively using less and less serious force than they otherwise might have. The use of a less-lethal instrument and personal weapons was proportional to the assaultive and potentially life-threatening actions of the subject. Having seen many fact patterns in which a subject’s movement for the equipment belt set the occasion for an officer-involved shooting, the Monitoring Team was duly impressed by the officers’ decision-making in this case. In a few instances across the 28-month overall study period, the Monitoring Team did conclude that SPD officers used a type or level of force that was disproportionate to the nature of the threat posed by the subject under the circumstances because they involved head strikes to the head or neck: • Case 12 (Type III). Officers were dispatched to a call involving a man vandalizing cars as he walked down the sidewalk and streets. Multiple patrol cars converged upon the subject. Officer 1, a Field Training Officer, approached and contacted the subject. The subject attempted to assault an officer. The officer counterstruck and attempted a takedown. Officer 1 reported that the subject wrapped his arms around him and tried to take his sidearm. Witness officer reports refer to the attempted gun grab. In-car video captured a struggle and a statement about “my gun” shortly after Officer 1 called the subject an “asshole.” The officer responded with four head strikes with his wooden baton. In-car video from another officer shows two strikes that appear to be full, top-down strikes. This is clearly a dangerous situation. Subjects attempting to dispossess an officer of his weapon are real occurrences and can lead to justified use of deadly force. However, by his own account, Officer 1 did not believe that he was in imminent peril and said that he was not in fear for his life or that of others yet used deadly force (head strikes) four times on an 80 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 85 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 unarmed subject who was, by then, surrounded by several officers.190 Because Officer 1 used head strikes despite not believing that such deadly force was warranted, the force may have been contrary to SPD policy because it was not proportional to the nature of the threat. • Case 13 (Type III). Two bike officers contacted two subjects sitting on a sidewalk. A Seattle ordinance prohibits sitting on sidewalks, and the officers informed the subjects of this. Subject 1 was asked for identification and provided a fake name and date of birth. After investigating, the officers identified the suspect and attempted to arrest him for a felony escape warrant. The subject attempted to escape, fighting and struggling with the officers. The officers and one of seven civilian witnesses claimed that the subject was assaultive. Five agreed that the suspect was hard to control. Indeed, the officers claimed that the subject came after them multiple times. After being brought under control and handcuffed, the subject complained of neck and shoulder pain. Medical care was summoned, with the subject placed in a cervical collar. An examination showed no skeletal or spinal injuries. Part of the incident was captured on two smartphone cameras and a bank surveillance camera. The video, in aggregate, shows both officers intentionally kneeing the subject in the face multiple times. Available video evidence and witness statements tended to establish that the subject’s efforts were aimed at flight rather than assault of officers. It should be noted that there is a moment on one of the phone videos where the subject appeared to potentially make some kind of contact with an officer’s belt or holster. Although there was no attempt to take any of the officers’ guns, reviewers understood that officers might have believed that, during the incident, an increase in the level or severity of force was necessary in light of the subject potentially going after the officers’ guns. The involved officers displayed poor tactics in attempting to restrain a subject who they outweighed by nearly 200 pounds. Initially, officers were working at cross-purposes, trying to physically pull the subject in opposing directions – indicating that officer planning and/or communication could have been improved. Even if the officers did believe that the subject was potentially in possession of a knife, the knee strikes delivered to the subject’s face are arguably contrary to policy because the force was not reasonable, proportional, or necessary under the totality of the circumstances. The FRB’s finding to the contrary lacks persuasive evidence. See Dkt. 3-1 ¶ 70(e) (noting that “a hard strike to the head with any impact weapon, including a baton, could result in death, and any strikes to the head should be consistent with policy and training”). 190 81 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 86 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Although the incidence of head strikes was low, and therefore does not pose an impediment to a finding of initial compliance, the Monitoring Team urges that the Department provide ongoing training to officers – whether in in-class, in-service training or via more informal roll-call or electronic platforms – that emphasizes that any strikes to the head are only justifiable if deadly force is justified under the circumstances. C. Reasonableness SPD policy requires that an officer “use only the force reasonable . . . to effectively bring an incident or person under control.”191 As outlined in Part I of this report, “[t]he reasonableness of… force is based on the totality of circumstances known by the officer at the time of the use of force” – and focuses on evaluating an officer’s force “from the perspective of a reasonable officer on the scene” and under the circumstances.192 The policy provides factors that are part of the reasonableness inquiry: • • • • • • • • • • • • The seriousness of the crime or suspected offense; The level of threat or resistance presented by the subject; Whether the subject was posing an immediate threat to officers or a danger to the community; The potential for injury to citizens, officers or subjects; The risk or apparent attempt by the subject to escape; The conduct of the subject being confronted (as reasonably perceived by the officer at the time); The time available to an officer to make a decision; The availability of other resources; The training and experience of the officer; The proximity or access of weapons to the subject; Officer versus subject factors such as age, size, relative strength, skill level, injury/exhaustion and number of officers versus subjects; and The environmental and/or other exigent circumstances.193 Across the more recent time period, the Monitoring Team found, or found no reason to dispute that, all involved SPD officers complied with SPD policy to use reasonable force 99 percent of the time in the most recent time period, up from 88 percent in the earlier time period. Examples of reasonable force, in addition to some of the incidents cited above, included: SPD Manual Section 8.100(1) (2014), Dkt. 107-1 at 6. SPD Manual Section 8.100(1) (2014), Dkt. 107-1 at 7. 193 SPD Manual Section 8.100(1) (2014), Dkt. 107-1 at 7. 191 192 82 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 87 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • Case 14 (Type I). Two officers stopped a car after reports that the occupants had been exchanging gunfire with another car. The officers pointed their guns at the suspects while executing the felony stop. This was reasonable and consistent with SPD policy given the evidence that the occupants were armed. • Case 15 (Type II). Officers responded to a male subject in the street who was threatening others and apparently suffering from a mental health crisis and alcohol intoxication. The subject had a bleeding injury from another individual having struck him in response to the subject’s threats. Officers decided to detain the subject for an involuntary mental health hold. The subject did not submit to the detention, and his increasingly erratic behavior concerned the officers. To restrain the subject in handcuffs for the safety of the subject, officers, and others, officers used a soft takedown. Given the subject’s aggressive and erratic behavior, and the associated potential for an aggressive or violent response, it seems clear that officers were reasonable to use a soft takedown to secure the subject. • Case 16 (Type II). Officers arrested a subject for harassment after he reportedly threatened a victim. The subject struggled against the officers’ attempts to place him in handcuffs, and the officers took the subject to the ground. A protracted struggle ensued involving the use of body weight and grips to control the subject. Based upon security footage and in-car video, it appears that officers attempted to seek voluntary compliance throughout their interaction with the subject – and only used force when the subject began to physically resist. The body weight and grips were reasonable to overcome the subject’s vigorous resistance. Because the concepts of necessity, proportionality, and reasonableness are all related, in some incidents discussed above in which the Team’s experts found that force was not proportional or was unnecessary under the circumstances, they also concluded that a reasonable officer, under the same circumstances, would not have used force. In addition to those cases, the Monitoring Team had concerns about the reasonableness of force in some other incidents, as well, including the following: • Case 17 (Type III). Officers placed a subject under arrest for an outstanding warrant. The subject struggled against officers’ attempts to handcuff him. After being handcuffed, the subject was moved to a police vehicle. At the vehicle, the subject was verbally protesting. Officers lifted the subject’s handcuffed arms high above his waist and back, which presents a high risk of injury and pain. When officers got the subject to the station, station lockup video depicts similarly problematic treatment of subject’s arms. When the subject resisted efforts to remove his eyeglasses, more pressure was applied to the subject’s raised arms. An officer removing the glasses threw them with substantial force to the ground. 83 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 88 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Later, after the subject kicked the interior of a holding cell door, Officers 1 and 2 entered the cell, presumably to remove the subject’s sneakers. In the process, Officer 1 performed a hard takedown. Officer 2 pinned the subject to the ground. The door closed behind the two officers, temporarily locking them in the cell with the resistive subject. After the officers left, the subject repeatedly kicked the door and, somehow, it opened. The subject exited the cell. A soft takedown and restraint by multiple officers was required to get the subject back into a cell. Raising the handcuffed subject’s arms was unreasonable, as the subject was under control. It appears likely to have been done to inflict discomfort or pain on a verbally resistant subject. The force subsequently used in the cell by officers was unreasonable and punitive, as it lacked a legitimate purpose, was entirely avoidable, and inadvertently placed the officers in substantial danger when the cell door closed behind them. Here, SPD’s Force Review Board identified the potential misconduct issues and was willing to find the force as prima facie violations of SPD’s Use of Force and De-Escalation policies. The fact that the Department appropriately identified and addressed policy violations increases confidence that SPD has in place the mechanisms of critical self-analysis that allow for it to systematically identify and meaningfully address officer misconduct, including improper use of force. • Case 18 (Type III). A subject acted in a bizarre manner while armed with handguns, prompting multiple calls to police. The armed subject then engaged in two car-jackings and drove dangerously. SPD officers initiated a pursuit, during which the subject discharged multiple rounds from a handgun. Multiple units ineffectively “rammed” the subject’s vehicle as the pursuit progressed. The pursuit terminated when a plain unit rammed the subject’s vehicle, causing it to also collide with other vehicles. As the subject again began to drive/maneuver his vehicle, multiple officers opened fire, collectively discharging in excess of 100 rounds. Of these rounds, one caused a penetrating gunshot wound and another caused a graze wound. SWAT responded and approached the vehicle in order to move the deployed airbags concealing the subject. During this process, a SWAT Officer fired five rounds at the subject, striking and killing him. The Monitoring Team had substantial concerns about the risk to public safety potentially created by the police response to the incident. In some ways, it was only a matter of luck that the involved officers did not strike a member of the public or a fellow officer with one of their more than 100 rounds. The Monitor had questions about the nature and volume of the force in this incident. 84 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 89 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • Case 19 (Type III). Officers responded to multiple calls of a person shouting at park patrons, smoking marijuana, and masturbating. They arrived sometime after the calls and observed the subject sitting on the ground near discarded cigarettes that he was taking apart to roll into a cigarette. Officers informed the subject that he could not smoke in the park and offered to give him a cigarette. He appeared cooperative but at some point became, according to the officers, non-responsive, with a distant stare and clenched fists. Officer 1 decided to preempt any aggressive behavior by handcuffing him but was unable to do so because his handcuffs were double-locked. He then reached over and around the front of the subject and brought him to the ground. The officers struggled with the subject for some period of time until he gave up his arms, which were under him, and they handcuffed him. During the initial investigation, three witnesses reported that Officer 1 had placed the subject in a choke hold. Two stated that the subject appeared to have difficulty breathing. The involved officer, in his interview, indicated that his arm likely did go across the subject’s neck inadvertently. There was inadequate FIT inquiry into this point. His partner said that he could not see what was happening. There was little FIT inquiry into this point. The Force Review Board would appropriately render administrative disapproval of the investigation on the basis that it was not thorough. It appears that the force used did not relate to anything that the subject did or any set of external circumstances but, rather, because the officer did not have his handcuffs ready to use. In proceeding down this path, officers affirmatively escalated the situation and used force disproportionate to the threat. Likewise, it was not clear that the initial handcuffing or takedown was warranted based on articulable facts – especially given that no commands or directions were given by the officers to the subject that would have allowed the subject to comply. Further, while the issue as to whether the officer in fact used a neckhold to control the subject remains under OPA investigation, if the officer used a neck hold, that would constitute deadly force and would not be objectively reasonable in the face of the subject’s resistance, threat posed, or the severity of any of the subject’s underlying offenses. D. De-Escalation 1. General Findings SPD policy provides that “when safe under the totality of circumstances and time and circumstances permit, officers shall use de-escalation tactics in order to reduce the need for force.”194 Even when force must be applied, officers must “assess and modulate the use-of-force as resistance decreases.”195 194 SPD Manual Section 8.100(3) (2014), Dkt. 107-1 at 8 (converted to sentence case). 85 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 90 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 The policy outlines a number of strategies and tactics that an officer may consider, including the use of distance, cover, and concealment; calling extra resources or officers to assist; moving from a position that exposes officers to threats; containing a threat or combative subject; communication from a safe position to gain the subject’s compliance; and others.196 In the earlier time period, shortly after implementation of the force policies, at least one officer in a given incident failed to reasonably employ de-escalation tactics when it was safe and feasible to do so in about 19 percent of all incidents that resulted in force – and about 26 percent of those force incidents where the facts or circumstances of the case made the duty to de-escalate applicable.197 Examples of incidents in which the Monitoring Team found a problematic failure to de-escalate included: • Case 20 (Type III). Officers 1 and 2 were patrolling on an “emphasis patrol.” They spotted a group of five males with open containers of alcohol. They approach and told the males to dispense with the beer. Subject 1 was initially non-compliant and escorted to a police car. Subject 2 approached officers and physically intervened, slapping Officer 1’s arms away. A melee ensued. Monitoring Team reviewers concluded that, even if the force was reasonable, necessary, and proportional, Officers 1 and 2 had failed to appropriately use de-escalation skills and techniques. Indeed, they may have escalated the situation. The Team agrees with the Force Review Board that the incident was very likely avoidable had the officers followed sound tactics and not engaged a boisterous group of five intoxicated males without additional resources. Officer 1, who was a crisis intervention specialist officer, should have been more focused on de-escalation efforts. Given the nature of the situation, conflict was predictable – and the involved officers failed to adequately plan for addressing it. Because SPD’s use of force policy requires de-escalation, the failure to de-escalate appears contrary to SPD policy. The point here is that the officer put himself at unnecessary risk in the situation, increasing the potential that force would need to be used to resolve the situation. Although limited, the Monitoring Team did observe a few instances where officer actions appeared to have affirmatively escalated the incident. For instance: Case 21 (Type II). Officers 1 and 2 responded separately to a felony violation of a domestic violence no-contact order. While en route, dispatch put out that the suspect had used meth and had a history of carrying weapons. After arriving and making unsuccessful attempts to • SPD Manual Section 8.100(4) (2014), Dkt. 107-1 at 9. SPD Manual Section 8.100(3) (2014), Dkt. 107-1 at 8-9. 197 For instance, a subject’s complaint of pain as a result of handcuffing is a Type I use of force. The duty to de-escalate is not directly implicated by the typical factual circumstances surrounding such force application. 195 196 86 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 91 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 persuade the suspect to submit to cuffing, Officers 1 and 2 began to grapple with him as he resisted. In-car video showed the subject resisting and reaching his right arm into the open window of Officer 1’s radio car as both officers grappled with him. Officer 2 stepped back, drew her baton, and struck the subject once on the lower leg. The subject braced and resisted the officers but was not actively aggressive with officers. Officer 2 slid her baton under the subject’s right arm and tried to pry him away from the radio car. The subject was handcuffed and taken into custody. Officer 1’s force (grappling), was found by the FRB to be within policy. On the other hand, FRB found Officer 2’s baton use contrary to policy and training and her general tactical approach was a failure to de-escalate. • Case 22 (Type III). Officer 1 and a sergeant contacted the subject, who had been kicking and throwing no-parking signs into the street. The subject did not comply with initial commands to stop. The sergeant cut him off, using his police vehicle. The subject continued not to respond to verbal commands. The officer and sergeant used control holds in an attempt to apply handcuffs. The subject struggled against their attempts. Officer 2 arrived on the scene and, as he approached on foot from his vehicle, Officer 1 and the sergeant took the subject to the ground. The take-down appeared to be an uncontrolled, collective fall on the part of the subject, Officer 1, and the sergeant, who went down to the ground with the subject at the bottom of the pile. The fall resulted in the subject striking his head on the roadway. Officers 1 and 2 and the sergeant struggled on the ground with the subject, who was unwilling to relinquish his hands for cuffing. Eventually, handcuffs were successfully applied. Officers 1 and 2 both applied a Type I wrist lock, Type II strike/takedown, and (classified out of caution) Type III force because the takedown resulted in apparent unconsciousness. The fall to the ground, which did result in a forceful landing for the subject, appeared to have been a consequence of all involved falling to the ground in an uncontrolled manner rather than a deliberate effort to achieve this effect. Under these circumstances, the force appeared consistent with SPD policy. Officer 3 pointed a firearm at the subject and applied a control hold to assist in handcuffing. These Type I force applications were also consistent with policy. The Monitoring Team did, however, have some concerns about the extent to which the involved sergeant’s verbal techniques during the initial commands to the subject to stop may have affirmatively escalated the situation. Despite eight hours of CIT training and his stated belief that the suspect was in crisis, the in-car video of the incident captures an exchange in which the sergeant demanded that the subject “put your hands on your head,” the subject asked “why,” and the officer in an aggressive tone said “because I told you to do so.” Although the subject still may not have complied with officer commands to stop, “because I told you to” seemed to invite a more confrontational exchange than a communication 87 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 92 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 strategy rooted more in SPD’s procedural justice and problem-solving communication training initiatives would counsel. However, SPD’s performance improved notably in the latter half of the 28-month study period that this report considers. Among the more recent cases that the Monitoring Team reviewed, officers complied with the duty to de-escalate in approximately 99 percent of those cases where that duty was applicable (i.e. in fact patterns where a reasonable officer would have believed that it was safe and feasible under the circumstances to employ de-escalation strategies or tactics) – up from 81 percent of cases in the earlier time period. The following exemplify those instances in which SPD officers did admirably fulfill their duties to apply strategic and tactical deescalation skills when safe and feasible to do so – and, indeed, in situations where the efforts to deescalate did not ultimately mitigate the officer’s need to use force consistent with SPD policy: • Case 23 (Type II). Officers responded to reports of an attempted assault and that the subject was following the victims. Officers arrived at the scene and positioned themselves between the subjects and the victims. The officers told the subject to relax. They asked his name. They offered their names. One officer used calming hand gestures and reassuring body language. Ultimately, the subject charged toward the victims. Officers performed a takedown. In taking the subject into custody, officers placed weight on the subject and attempted a cross-face pain compliance hold. The incident was captured on in-car video. Despite not ultimately being successful in terms of brining the subject into compliance, Monitoring Team reviewers were nevertheless strongly impressed by the responding officers’ verbal communication and de-escalation skills. • Case 24 (Type III). An initially-responding officer to a call involving a subject walking down a street with two butcher knives in his hands soundly, reasonably, and pragmatically waited for backup rather than affirmatively confronting the subject by himself. • Case 25 (Type II). Officers responded to a fight in progress and located the subject outside of the building where the incident occurred. Officers developed probable cause to arrest the subject for property damage. Officers waited for additional officers to arrive before attempting to arrest the subject. SPD personnel ultimately needed to use intermediate-level, type II force (a takedown, body weight, and control holds) in response to the subject’s active resistance while being taken into custody. Although officers nonetheless needed to use force while bringing the subject into custody, the initially-responding officers’ decision to wait for additional resources to arrive at the scene before initiating an arrest constituted prudent de-escalation tactics under the circumstances (e.g., a subject outside of a venue where a fight was known to have just occurred). 88 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 93 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • Case 26 (Type II). Officers detained and subsequently arrested a subject for an assault in which the victim was knocked unconscious in an apparently random attack. As officers sought to handcuff the subject, he tensed and pulled his arms away. Officers effectuated a soft takedown of the subject to the ground, where he was successfully handcuffed. No injuries to the subject were identified at the scene. A minor abrasion or bruise was identified later. This Type II use of force was straightforward, reasonable, necessary, and proportional. Officers communicated calmly with the subject throughout the encounter and used a low degree of force to overcome his resistance – and, indeed, did not use any more force than was necessary to effectuate the arrest. The Monitoring Team found reason to believe that the strategic communication tactics employed may well have prevented the situation from escalating further. 2. The Issue of Solo, Affirmative Action It has been well-established that solo pursuits by officers of subjects can elevate the risk of injury to the officer and the likelihood that the officer will need to use force. Thus, even in those instances where force used by officers is reasonable, necessary, and proportional, solo foot pursuits can constitute dangerous and unsound tactics under a number of circumstances. Especially in its review of cases in the earlier time period covered by this assessment, the Monitoring Team identified more instances than it would like to see of SPD officers pursuing fleeing subjects on their own rather than containing the situation and waiting for backup to arrive. Not all of these instances necessarily involved the application of force that was unreasonable or counter to SPD policy at the time it was applied. Instead, some cases involved tactics during the course of an unfolding incident, and before force was used, that the Monitoring Team found to be inappropriate or unsound. • Case 27 (Type III). Officers responded to a burglary-in-progress call at a reportedly empty residence. Upon arrival, officers ascertained three subjects were still onsite. The officers split up. Officer 1 went to the rear yard alone. He encountered a subject, who was on top of a garage. Telling the subject to come down or he would “beat [the subject’s] ass,” Officer 1 pursued the subject alone as the subject unsuccessfully attempted to jump over a gate. The two collided. The subject claimed that the officer just ran into him. Officer 1 alleged that he was attacked. Regardless, a struggle between the two ensued. Officers 2 and 3 reached Officer 1 to assist, although they were delayed in reaching him due to darkness and obstacles that separated Officer 1 from the others. The subject was eventually brought under control and into custody. Officer 1’s pursuit of a subject, on foot and alone and for however long it lasted, was dangerous, tactically unsound, and inconsistent with SPD policy. Not only did Officer 1’s 89 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 94 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 splitting from the other officers put him in the position of confronting the subject alone, but Officer 1 actively pursued the subject from there without the benefit of multiple other officers on-site. Under the circumstances, it was unsafe and unreasonable for the officer to attempt to apprehend the suspect alone. • Case 28 (Type III). Officers 1 and 2 responded to a report of a subject in a victim’s vehicle. On arrival, Officer 1 encountered the suspect in the bushes and told him to come out and show his hands. The subject emerged holding a knife. Officer 1 backed off. As the subject walked away, the officers began to track the subject. Officer 2 discharged a Taser at the subject, but this was ineffective. A foot pursuit ensued during which officers became separate, with Officer 2 falling behind. Officer 1 turned a corner and confronted the subject at a distance of approximately 10 feet. The subject shined a flashlight at Officer 1, interfering with the officer’s ability to clearly see his surroundings. Believing that he would be stabbed by the non-compliant, armed subject, Officer 1 fired four rounds at the subject, striking and injuring him. A knife was recovered at the scene. Although it can be argued that at the time that force was applied, it was reasonable, necessary, and proportional under the circumstances, the tactical issue of concern in this case was the manner in which the officers pursued the subject. Officers 1 and 2 became separate, which resulted in Officer 1 being alone at the time of the officer-involved shooting – greatly increasing the risk that the subject posed to the officer. The question of whether or how Officer 1 maintained awareness of his partner’s position, or whether he could or should have taken steps to coordinate with him, was not explored by the FIT investigation. The partners split raised significant concerns that the officer's actions elevated the risks involved to the officer and made it more likely that force would need to be used. • Case 29 (Type II). A witness saw the subject breaking into a car and called the police. The responding officer, while knowing it was a property crime, chased the subject through a neighborhood, found her on a porch, and immediately went “hands-on” with the subject, using a Type II strike/takedown. The officer did not wait for backup, nor did she attempt use any strategic communication skills in an effort to gain compliance without force. Given the officer’s affirmative decision to pursue a subject without waiting for backup and failure to deploy de-escalation techniques, a finding that force was contrary to SPD policy would have been reasonable. SPD’s chain of command and the Force Review Board both identified the tactical issues associated with the officer searching for the subject alone and immediately going hands-on upon locating the subject. Per these findings, the case was referred to OPA. 90 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 95 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • Case 30 (Type II). Officer 1 responded to a report of a burglary in progress that involved a subject who was nude and shouting Bible verses. The subject fled on foot. Officer 1 initiated a foot pursuit. As additional SPD units converged on the subject, Officer 1 pushed the subject, causing him to fall to the ground. Once on the ground, Officer 1 was assisted by Officer 2 and Officer 3, who each applied control holds, as the subject was handcuffed. It would have been safer overall if Officer 1 had not engaged the subject until Officers 2 and 3 arrived. In some other instances not involving foot pursuits, SPD officers affirmatively closed the distance between themselves and potentially dangerous subjects when doing so was unnecessary and placed themselves in substantially higher danger. For example: • Case 31 (Type III). A homeless man had illegally lit a fire in a makeshift fire pit area. A park security guard reported that the subject had attempted to attack him. Two officers responded and called for CIT backup. Those officers tried to persuade the subject to leave the park, but he refused and threatened them. The officers then called and waited for a supervisor to arrive. While waiting for the supervisor, the officers prepared a contact and cover plan. A Sergeant arrived, briefly spoke with the officers, and then started to contact the subject himself. The subject held up a gas can to the fire, broke a bottle, and threatened the now-five officers who had responded to the scene. Officer 1 and Officer 2 deployed their Tasers unsuccessfully. The subject advanced toward Officer 2 with the broken bottle and gas in hand. Officer 2 slipped and fell. The subject advanced toward him. Officer 1 transitioned from Taser to firearm, opening fire. Another officer on the scene, Officer 3, also opened fire. The subject died. In this instance, the use of the firearm (Type III force) was consistent with SPD’s use of force policy. With a subject with a sharp object and gasoline advancing toward a temporarily incapacitated officer, it was reasonable, necessary, and proportional under the circumstances for Officers 1 and 3 to use deadly force. The deployment of the Taser by Officer 1 and 2 to a subject carrying gasoline is inconsistent with SPD’s Taser policy which mandates that Tasers “not be used in any environment where an officer knows that a potentially flammable, volatile, or explosive material is present.”198 The deployment of the Taser under such circumstances increased the likelihood of injury or death and, thereby, the severity of the force. At the moment that the Tasers were applied, it was clear that the subject may pose a danger but, before and until the subject affirmatively 198 Seattle Police Manual 8.200-POL-3(8) (2014), Dkt. 107-2 at 8; see Dkt. 3-1 ¶ 79(d). 91 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 96 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 advanced toward officers, it does not appear that the nature of the threat posed by the subject would justify the use of deadly force. However, the most significant issue in this instance was the poor tactics of the sergeant in this case. The initially-responding officers acted appropriately when initially responding to the scene – calling for CIT backup, using verbal techniques with the subject, calling for a supervisor, not approaching or initiating contact with the subject, and forming a contact and cover plan. Ironically, this superior performance decisively ended when the sergeant arrived, spoke only briefly with the officers, and closed the distance with the subject – initiating contact and escalating the situation. The sergeant’s actions set in motion the events that put all the SPD personnel at risk and culminated in an officer-involved shooting. Rather than closing the distance, the officers had plenty of time to keep their distance from a subject who did not have a firearm and to try to use verbal de-escalation techniques as well as consider the deployment of a beanbag or less-lethal projectile option if others were unsuccessful. The sergeant’s comprehensive violation of SPD’s policy in this instance was troubling, as the FRB found. The FRB did not refer findings to OPA because the sergeant had retired. • Case 32 (Type III). A Field Training Officer (“FTO”) and student were patrolling when they observed the subject riding a bicycle on an I-5 ramp. The FTO recognized the subject and knew him to be the subject of a felony warrant. After verifying that the subject had an outstanding warrant for burglary, the officers decided to stop the subject and requested backup. The subject appeared to flee from officers on his bike and eventually crashed his bike. The student officer approached the subject by herself and a struggle ensued. The subject was later found to have a dislocated left shoulder. The subject also experienced abrasions to his face. It could not be definitively established whether officers or the subject’s bicycle crash caused the injuries. The Force Review Board properly determined that the FTO should have affirmatively come to the assistance of the student officer rather than passively observing the struggle. • Case 33 (Type III). Two officers responded to a burglary call alleging that two male suspects were stealing bicycles from an apartment complex. They arrived, investigated, and ultimately contacted the two subjects. One subject fled on foot and fell to the ground either (a) when Officer 1 shoved him to the ground from behind, or (b) Officer 2 bumped into the subject with his shoulders. According to both officers, the subject fell onto bark dust and struck his head only when he rolled off a curb. This movement, the officers reported, caused him to strike his head on the ground and sustain a hematoma to his thigh. The subject, a hemophiliac, required four days of hospitalization due to internal bleeding. The second subject escaped. 92 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 97 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 The involved officers should have called for backup before conducting a building search for multiple subjects. Moreover, the officers parked directly in front of the building where the burglary was occurring and did not request backup despite clearly needing to check a building and parking structure area for two suspects. • Case 34 (Type III). Officers 1 and 2 responded to a report of a violent female with a mental illness at a hospital. Although the officers arrived together, Officer 1 immediately approached and contacted the subject. The subject swung a bag at Officer 1, prompting the officer to pull the subject’s arm in an effort to control her. This caused the subject to fall. The subject injured her chin as a result. She was subsequently admitted to a hospital for treatment of injuries apparently unrelated to the use of force. The admission to the hospital nonetheless triggered classification of this force incident as a Type III use of force. The officer’s failure to coordinate with his partner and make affirmative, unilateral contact with the subject foreclosed the ability for the officers to apply affirmative or strategic deescalation tactics and appeared to contribute to the need to use force. The officer gave up the advantages of time, distance, and cover by immediately approaching a subject who he knew to be suffering mental illness and exhibiting violent behavior. Because the officer escalated the situation and put himself at far greater risk than necessary under the circumstances, the use of force was arguably inconsistent with SPD’s force policy and its requirement to de-escalate situations when safe and feasible to do so. • Case 35 (Type III). Two officers responded to a call of a possible domestic disturbance in a car. The car was parked with the engine running. A female driver and male front passenger gave suspicious answers as to their residence and purpose for being there. The car, with outof-state plates, came back as stolen. A third officer arrived as backup. Officer 2 stuck his hand in the passenger window of the vehicle and placed a cuff on the subject’s wrist. The female driver then tried to roll up her window and put the car in gear. Officer 1 pulled the car door open and pulled the female driver to the ground, where the officer began handcuffing her. Meanwhile, the male passenger exited the car with one hand cuffed and attempted to run past Officers 2 and 3, who ultimately tackled the subject to the ground, where he struck his head on the pavement. The incident was partially captured by two in-car videos which show residents in an adjacent building holding out their cell phones. The investigatory file did not suggest a rigorous effort to obtain such civilian video during the investigation. Officer 1, who was clearly the officer in charge of the stop, made several errors. First, he did not ask the driver of the car to remove her car keys – which he should have done at the outset, preventing the vehicle from being either a means of flight or a deadly weapon. Second, when the car came back as stolen, he should have summoned additional officers and moved back to carry out a regular felony stop. As it turned out, the driver had a firearm in 93 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 98 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 her purse, which meant that the failure to call for back-up and treat the encounter as a serious and dangerous one might have put all officers involved in grave danger. Third, the officer should have summoned additional officers before signaling Officer 2 to initiate the arrest of the passenger. Officer 2’s takedown of the male passenger appears to be justified. Nonetheless, Officer 2 appears not to have taken the usual tactical precautions once the car came back as stolen, including requesting that the driver turn off the vehicle, attempting to handcuff the male subject while he was still seated in the vehicle, and not separating the parties involved in the reported domestic violence. Officer 2 exacerbated the situation by ratcheting his handcuffs loudly in the presence of the passenger and then reaching into the vehicle to cuff one wrist before the subject was secured. Later, after both parties were cuffed and seated, Officer 1 grabbed the male passenger by the rear of the neck in a C-clamp hold. The use of the C-clamp push down on the neck was unwarranted, unjustifiable, and punitive. Both SPD’s Force Review Board and the Training Section agreed on the inappropriateness of applying this maneuver on a handcuffed subject under the officers’ control. OPA sustained the findings of a policy violation as to Officer 1, which was upheld by the Department. In the same case, an involved officer appeared to kick the subject in the abdomen while he was on the ground and as other SPD officers were in the process of restraining the subject. This force appears unreasonable under the circumstances, as the subject was on the ground and being brought under control by several SPD officers. Likewise, it seemed unnecessary and disproportionate to the nature of the threat. Importantly, the incidents recounted above notably contrast to some more recent cases where it appeared that officers appropriately used Departmental resources to effectuate the necessary law enforcement objectives: • Case 36 (Type II). Gang officers spotted a subject with open warrants for robbery and burglary. Prior to initiating with the subject, who was situated in a parking lot, officers place units in the area and then attempted verbal contact via detectives. The subject fled on foot. Given the outstanding felony warrants, three officers gave chase. Officer 1 grabbed the subject around the waist and pulled him down into an area of bushes. (The incident was conservatively classified as a Type II because of scratches to the subject’s arm from the bushes). Officers 2 and 3 used minor force – appropriately classified as de minimis – to keep the subject on the ground, restrain his arms and legs, and allow for handcuffing. In the Monitoring Team’s view, this is a good example of officers using advantages in terms of superior numbers and resources to take a fleeing felony suspect into custody. Prepositioning resources in the vicinity entailed sound decision-making that increased the 94 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 99 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 likelihood that the officers would have the tactical advantage. Such a coordinated team approach left no officer disproportionately exposed to risk and likely mitigated the severity of the force used. • Case 37 (Type I). Officer 1, in plainclothes, was patrolling a back alley when the subject stepped in front of the unmarked police unit and pointed a handgun at the officer. Officer 1 backed out of the alley and called for backup officers. Responding officers located the subject nearby, pointing their firearms at the subject (which constitutes Type I force) to get him to comply and be taken into custody. He was arrested without further incident. The Monitoring Team found Officer 1’s actions to be quick, decisive, and prudent under challenging circumstances. By increasing distance between himself and the subject, calling for additional resources, and waiting to engage with the dangerous subject until that backup arrived, Officer 1 ensured both the officer’s personal safety and public safety. The Monitoring Team understands that officers in some instances must affirmatively confront subjects in order to effectuate their law enforcement duties. However, mindful that reducing space between a subject and officer reduces the scope of the de-escalation tools available to officers in the event that a subject becomes threatening or becomes more of a threat during an interaction, the Monitoring Team’s experts would like to ensure that all officers continue to systematically assess all available strategic options in all encounters – including remaining at a distance from an officer while a threat is assessed more closely or as additional resources are dispatched. Although highly encouraged that it saw far fewer problematic instances of solo foot pursuits or problematic solo action in the more recent time period studied for this assessment, the Monitor recommends that SPD continue its comprehensive training initiative – featuring electronic, roll call, and in-class elements – that further focuses on pursuits and appropriate single-officer tactics to ensure that these positive trends continue and endure. E. Specific Force Prohibitions The application of physical force in some circumstances is prohibited in situations where force is presumed to be necessarily unreasonable, unnecessary, disproportionate, and contrary to using strategic or tactical de-escalation skills to mitigate the likelihood that force will be used. SPD Policy Section 8.100(2) outlines five specific instances in which an officer is not authorized to use force. These nearly categorical prohibitions stem from the DOJ’s 2011 findings letter. This section considers force incidents reviewed by the Monitoring Team to determine if these prohibitions are effective in practice. 95 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 100 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 1. Force Used to Punish or Retaliate Officers are prohibited from using force in a manner that is retaliatory or punitive.199 The Monitoring Team identified only isolated instances in which force seemed sufficiently unnecessary and disproportionate to the threat so as to appear potentially punitive in nature: • Case 38 (Type III). This case was described in detail, above. The use of the C-clamp to push down on the neck of the handicapped subject appeared to be unwarranted, unjustifiable, and punitive. The force was referred to OPA. • Case 39 (Type I). The involved officer was working undercover in a sting operation in the early morning hours. One female suspect fled the scene. The involved officer broke cover and assisted a uniformed officer in a foot pursuit of the subject. After some distance, the subject stopped fleeing and collapsed to the ground. She kept her hand on what appeared to be a gun in the waistband of her pants. The involved officer, who had not previously identified himself as a police officer, told the subject to let go of the gun and present her hands. The subject did not respond. The involved officer subsequently drew his gun and placed the barrel of his gun against the subject’s head, telling her that he would shoot her if she did not comply. The other, uniformed officer was then able to grab the gun from the subject. The involved officer holstered his weapon. The Monitoring Team had concerns about the officer placing the gun squarely against the subject’s head and threatening to kill her. If something had gone wrong, an officer-involved shooting may have constituted excessive force. Indeed, the tactic seemed punitive and hence inconsistent with policy, whether or not a shooting took place. 2. Force Against Individuals Only Verbally Confronting Officers The Monitoring Team did not identify any force incidents, across either the early or later study periods, that implicated the prohibitions against using force on individuals who had solely verbally confronted officers. This constitutes noteworthy and commendable compliance on the part of the Department and its officers with respect to ensuring that force used is necessary and proportional under the circumstances. 199 PD Manual Section 8.100(2) (2014), Dkt. 107-1 at 7. 96 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 101 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 Other sections of this report detail a few instances in which verbal confrontation, in the context of other circumstances, seemed to set the occasion for the use of force even when options to deescalate were available, feasible, and safe. 3. Force on Handcuffed or Otherwise Restrained Subjects SPD officers may not use force against subjects who are handcuffed or otherwise restrained. The exception is only in “exceptional circumstances when the subject’s actions must be immediately stopped to prevent injury, . . . escape, [or] destruction of property.”200 The Monitoring Team found only isolated instances where the application of force to a handcuffed or otherwise restrained subject was an issue, which were primarily restricted to the earlier time period evaluated. • Case 40 (Type III). In this case, also descried in another example, supra, a handcuffed subject, in SPD custody in a holding cell, was being taken out of the cell to place a spit-sock on him after he spat on a sergeant. In doing so, the subject lost his balance, resulting in the still-handcuffed subject taking an essentially unbroken fall and landing on his face. The subject sustained a bleeding injury, and the force was classified as a Type III use of force. The incident was captured on video. There is also evidence to suggest that the sergeant was angry and upset – allegedly snapping at a witness officer who offered assistance, “I’m here and I will handle this.” The use of force seemed, at minimum, unnecessary. The sergeant more appropriately should have closed the cell door rather than entering and putting his hands on the subject. The Force Review Board, without documented explanation, found that the de-escalation policy was adhered to, but the Board cited the sergeant’s performance as inconsistent with training and the force used as unnecessary. Because the sergeant retired prior to the Board considering the incident, the matter was not referred to OPA. • Case 41 (Type III). (The facts in this case are discussed supra.) An officer used a C- clamp on a handcuffed subject. The incident was partially captured by two in-car videos which show residents in an adjacent building holding out their cell phones. The investigatory file did not suggest a rigorous effort to obtain such civilian video during the investigation. FIT, the Force Review Board, and the Training Section agreed on the inappropriateness of applying this maneuver to a handcuffed subject under the officers’ control. OPA sustained the policy violation on the use of force, which was upheld by the Department. 200 SPD Manual Section 8.100(2) (2014), Dkt. 107-1 at 7. 97 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 102 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 4. Force to Prevent Subject from Swallowing a Substance or to Extract a Substance or Item from Inside the Body of a Suspect without a Warrant. The Monitoring Team did not identify force incidents that implicated the prohibitions against using force to prevent a subject from swallowing a substance or to extract an item from an individual’s body. F. Use of Deadly Force: Neck Maneuvers This report elsewhere outlined some instances where officers applied a maneuver to the head or neck that a reasonable person would conclude to be inconsistent with SPD’s policy that such a maneuver constitutes deadly force. Specifically, deadly force is “[t]he application force through the use of firearms or any other means reasonably likely to cause death, Great Bodily Harm, or serious physical injury,” which include “[a] hard strike to a person’s head, neck, or throat”; “[s]triking a person’s head into a hard fixed object”; and “[n]eck and carotid restraints.”201 Blunt trauma to an individual’s head and the restriction of blood and oxygen to the brain caused by impacts or pressure to the neck; strikes to the head, neck, or throat; or maneuvers that would restrain the neck can only be used when the nature of a subject’s threat and the totality of the circumstances would justify the use of deadly force. Further, in SPD’s policies regarding intermediate weapons, the Department requires that even “[u]nintentional or mistaken blows to these areas [head, neck, and throat, among others] must be reported to ensure that all reasonable care was taken to avoid them.”202 Incidents that involved unreasonable and impermissible neck restraints dropped precipitously in the latter half of the 28-month period considered here, which was noteworthy. Nevertheless, SPD should reinforce the significance and gravity of neck restraints, as well as head strikes, in ongoing and future use of force training such that instances of the following are addressed: • Case 42 (Type III). A sergeant was working a plainclothes “buy/bust” detail when he heard a broadcast of a pursuit of a robbery subject. The sergeant broke cover and pursued the subject, bringing him to the ground. Another SPD officer assisted in the arrest. That officer reported that he placed his shin on the subject’s “lower neck and left shoulder blade.” SPD did not categorize this force as a Type III; however, the application of pressure to the neck constitutes significant, Type III force, and the incident should have been categorized as such even if it was consistent with SPD policy and training. SPD Manual Section 8.050 (2014), Dkt. 107-1 at 4; accord SPD Manual Section 8.200-POL-10(1) (“Officers May Only Use Neck and Carotid Restraints When Deadly Force is Justified.”). 202 SPD Manual Section 8.200-POL-5-(4) (2014), Dkt. 107-2 at 11. 201 98 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 103 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 • G. Case 43 (Type II). In a struggle with a subject, one officer reported putting a knee on a subject’s neck. However, the Monitoring Team’s reviewers were unable to determine whether that force application was consistent or inconsistent with SPD policy because it was not adequately explored by the chain of command investigation. It is true, however, that the subject did not appear to have been injured by that application of force. Force Instrument-Specific Guidelines Under the Court-approved use of force policy, “uniformed officers are required to carry at least one less-lethal tool,” such as a Taser, OC (pepper spray), or baton.203 As that policy observes, “[l]esslethal devices are used to interrupt a subject’s threatening behavior so that officers may take physical control of the subject with less risk of injury to the subject or officer than posed by great force applications.”204 Because the nature of each less-lethal force instrument entails certain elements and sometimes unique considerations that officers must keep in mind, SPD’s Use of Force Policy Manual on Force Tool-Specific Policies sets forth a number of requirements relating to the use of each force instrument.205 These guidelines do not supplant or replace the general requirements. Instead, they provide specific guidance for officers to ensure that their use of less-lethal force instruments adheres to the general force policy. Reviewers identified a number of instances in which less-lethal instruments were employed effectively and in a manner consistent with policy, including in many of the following cases, as well as those involving application of the Taser and OC spray outlined in examples above: • Case 44 (Type II). Officers responded to a radio call of a man screaming in his room in a hostel. Officers encountered the subject intoxicated and in apparent mental health crisis. The subject was armed with knives, was self-inflicting injuries, and stated his desire to be shot by police. Protracted negotiations were conducted in an attempt to persuade the subject to surrender. These efforts were unsuccessful. Specialized resources, including SWAT, were also requested. After approximately 30 minutes of negotiations, a single TASER discharge was used. The subject was taken into custody to be placed on an involuntary hold. The use of one Taser discharge to resolve a dangerous situation, involving a subject armed with knives, constituted appropriate, necessary, and proportional force under the SPD Manual Section 8.200(2) (2014), Dkt. 107-1 at 11 (converted to sentence case). SPD Manual Section 8.200(2) (2014), Dkt. 107-1 at 10. 205 SPD Manual Section 8.200-POL-1–10 (2014), Dkt. 107-2 at 2-11. 203 204 99 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 104 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 circumstances – and allowed for resolution of the situation without more significant or severe force needing to be employed and without subject or officer injury. • Case 45 (Type II). Patrol officers responded to a domestic violence incident in which a subject was alleged to have stabbed his father in the face with a fork. The subject was noncompliant with officers and retreated into the house. Believed to be armed with a knife and refusing to exit, the subject was considered “barricaded.” SWAT and HNT responded. Various attempts were made to engage the subject in negotiation. None were effective, with the subject remaining non-responsive. Ultimately, the decision was made to use OC canisters to cause the subject to exit. Multiple OC canisters were deployed. The subject exited following the introduction of OC spray and was taken into custody without further incident. The use of less-lethal instruments here was reasonable, necessary, proportional – and effective in bringing a non-compliant subject into custody without requiring officers to more affirmatively close the distance between themselves and the subject. • Case 36 (Type II). A mentally ill individual was heavily drunk on Everclear and slashing himself with a knife while asking officers to shoot and kill him. After substantial but ultimately unsuccessful verbal de-escalation efforts, officers tased the subject. The subject was subsequently taken into custody without incident. In some other incidents, problems were identified in the deployment of less-lethal instruments. It must be noted that a good majority of these problematic less-lethal applications were identified in the earlier review timeframe rather than the more recent timeframe – suggesting that officer decision-making and performance has improved with respect to the use of less-lethal weapons. Nevertheless, the Monitor would be remiss not to describe briefly some of the incidents that concerned the Team: • Case 37 (Type III). Officers responded to a complaint that a subject was behaving erratically and had threatened to kill his roommate with a knife. Officer 1 was the primary officer on a six-officer contact team that went to the subject’s apartment door. The subject, who officers had established was alone in the apartment, answered the door but refused to show his right hand. Officer 1 pointed the Taser at the subject, who eventually held the door with his previously-hidden hand. The subject then went back into the apartment. Aware of two warrants of the subject’s arrest, Officer 1, with sufficient grounds under Washington state law to enter the premises, proceeded into the apartment to make an arrest. Upon entering, the subject was in possession of a lock-blade knife and was trying to open it. Officer 1 applied the Taser for an 11-second cycle. The subject was disarmed, handcuffed, and taken into custody. Officers 2 and 3 had their guns pointed at the subject and then assisted during cuffing. The subject sustained a laceration to a finger, likely during either his effort to open the knife or while being Tased. 100 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 105 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 The application of a Taser to a knife-wielding subject was, in this instance, likely proportional, necessary, and reasonable under the circumstances at the time that it was applied. However, Officer 1 applied the Taser for eleven seconds. SPD policy allows only a five-second cycle: When a CEW [Taser] is used against a subject, either in probe or drive stun mode, it shall be for one standard discharge cycle of five seconds and the officer using the CEW must reassess the situation. Only the minimum number of five second cycles necessary to place the subject in custody shall be used.206 The officer’s flippant responses to various questions during an interview with FIT also suggested a troubling lack of professionalism. In short, although Officer 1’s application of the Taser may have been reasonable. Yet the application of the Taser for more than twice as long as sanctioned by SPD policy constitute violations of SPD policy. The FRB’s decision to find the 11-second application of the Taser as justified was ill-advised and ran contrary to SPD policy. • Case 38 (Type II). An officer was in a supermarket getting food for himself. A store employee approached and told him that a customer had assaulted a store manager and was refusing to leave. The officer approached the subject, who was talking on his cell phone, and asked him to leave. When the subject refused to acknowledge the officer, the officer picked up the subject’s bag, which was sitting on the floor, and attempted to escort the subject out of the store. The subject became agitated, slapped the officer’s hand and arm, and tried to grab the officer’s hands to stop him from escorting the subject out of the store. The officer and subject made their way, while in the middle of a “hand fight,” outside the store. The officer reported that the subject pulled up his pants and crouched in a fighting stance. Believing that the subject was about to assault him, the officer pulled out his OC (pepper) spray, put his arm up behind him to create some distance between him and the subject, and applied the OC spray to the subject. The officer stated in his report that he did not believe that he had time to issue a warning that he was going to deploy the OC spray because he thought the subject was about to assault him. Upon application of the OC spray, the subject stopped resisting. Arriving backup officers handcuffed the subject. Like the chain of command that reviewed this Type II force and the Force Review Board, Monitoring Team reviewers found that the officer had failed to follow SPD’s policy on OC 206 Seattle Police Manual 8.200-POL-3(4) (2014), Dkt. 107-2 at 8; Dkt. 3-1 ¶ 79(c). 101 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 106 of 108 Seattle Police Monitor Ninth Systemic Assessment April 2017 spray by failing to warn of its use. Although the officer contended that it was not feasible to provide such a warning, SPD’s own review of the incident concluded that a reasonable officer would have been able to yell “pepper spray” before applying it. Likewise, the involved officer failed to address why he made no efforts to try to decontaminate the subject after using the OC spray as required by policy. A failure to render medical aid by the police at the earliest feasible opportunity should be subject to substantial discipline. The consequences of failure to decontaminate the subject here led to unnecessary and avoidable pain for which non-trivial discipline was appropriate. A Note on “Blast Balls” Some community attention has focused on SPD’s use of “blast balls,” which the Department sometimes uses as a crowd control tool. The devices, which give off a loud sound when used, are intended to distract and direct large crowds. The Monitoring Team’s sample of cases happened to include a large “use of force” that, in actuality, was a combined or collapsed array of uses of force related to a crowd-control incident. The incident, then, included numerous applications of various types of intermediate- and low-level force, including applications of the “blast ball.” Although the use of blast balls is not overly widespread – and has been primarily limited to occasional, large protests under certain circumstances – the Monitor urges SPD to ensure that appropriate protocols for the use of the instrument are included in its policies. The current Monitoring Plan calls for updated force policies to be submitted to the Monitor and the Court for approval again, and the Monitor will look forward to reviewing new standards and protocols regarding blast balls. 102 Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 107 of 108 Monitoring Team Staff Merrick Bobb Monitor Matthew Barge Deputy Monitor Peter Ehrlichman Deputy Monitor Ronald Ward Assistant Monitor Chief Hassan Aden (ret.) Senior Police Expert Brian Center Senior Consultant Julio Thompson Marnie Carlin MacDiarmid Karlene Goller Andrea Yang Esq. Joseph Doherty Ellen Scrivner Ph.D. Jeffrey Yamson Tim Shugrue Executive Assistants Case 2:12-cv-01282-JLR Document 383 Filed 04/06/17 Page 108 of 108 1 2 CERTIFICATE OF SERVICE I hereby certify that on this date I caused to be served the foregoing on the following 3 counsel of record by the method indicated: 4 J. Michael Diaz michael.diaz@usdoj.gov 5 Kerry Jane Keefe kerry.keefe@usdoj.gov Rebecca Shapiro Cohen rebecca.cohen@usdoj.gov Puneet Cheema puneet.cheema2@usdoj.gov 7 Timothy D. Mygatt timothy.mygatt@usdoj.gov 8 Christina Fogg christina.fogg@usdoj.gov Annette L. Hayes annette.hayes@usdoj.gov Peter Samuel Holmes peter.holmes@seattle.gov 10 Michael K. Ryan michael.ryan@seattle.gov 11 Andrew Thomas Myerberg andrew.myerberg@seattle.gov Brian G. Maxey brian.maxey@seattle.gov Gregory C. Narver gregory.narvar@seattle.gov 13 John B. Schochet john.schochet@seattle.gov 14 Rebecca Boatright rebecca.boatright@seattle.gov Carlton W.M. Seu carlton.seu@seattlegov Gary T. Smith gary.smith@seattle.gov 6 9 12 15 Via Messenger Via Facsimile Via U.S. Mail Via Electronic Mail Via ECF Notification 16 17 Dated this 6th day of April, 2017. 18 19 s/ Matthew Barge 20 21 22 23 24 MEMORANDUM SUBMITTING NINTH SYSTEMIC ASSESSMENT REGARDING USE OF FORCE - 2 12-CV-01282-JLR MERRICK J. BOBB, MONITOR Police Assessment Resource Center P.O. Box 27445 Los Angeles, CA 90027 (213) 623-5757