Use of Force by Police Overview of National and Local Data

NT OF ME J T US U.S. Department of Justice R T A I P C E E D

O J C S Office of Justice Programs F A V M F O I N A C I J S R E BJ G O OJJ DP O F PR JUSTICE U.S. Department of Justice Office of Justice Programs 810 Seventh Street N.W. Washington, DC 20531

Janet Reno Attorney General

Raymond C. Fisher Associate Attorney General

Laurie Robinson Assistant Attorney General

Noël Brennan Deputy Assistant Attorney General

Jeremy Travis Jan M. Chaiken Director, National Institute of Justice Director, Bureau of Justice Statistics

Office of Justice Programs National Institute of Justice Bureau of Justice Statistics World Wide Web Site World Wide Web Site World Wide Web Site http://www.ojp.usdoj.gov http://www.ojp.usdoj.gov/nij http://www.ojp.usdoj.gov/bjs Use of Force by Police Overview of National and Local Data

Contributors:

Kenneth Adams Joel H. Garner Patrick A. Langan Geoffrey P. Alpert Lawrence A. Greenfeld Christopher D. Maxwell Roger G. Dunham Mark A. Henriquez Steven K. Smith

October 1999 NCJ 176330 Jeremy Travis Jan M. Chaiken Director, National Institute of Justice Director, Bureau of Justice Statistics

Robert J. Kaminski Program Manager, National Institute of Justice

Acknowledgments: This report is indebted to many individuals and organizations for their valuable assistance and insights. Special thanks are extended to the law enforcement agencies that cooperated with the researchers whose findings appear in this report. In so doing, the following agencies demonstrated the type of leadership so critical to the advancement of policing practice and policy: Charlotte-Mecklenburg (North Carolina) Police Department, Colorado Springs (Colorado) Police Department, Dallas (Texas) Police Department, Eugene (Oregon) Police Department, Miami-Dade (Florida) Police Department, St. Petersburg (Florida) Police Department, San Diego (California) Police Department, San Diego County (California) Sheriff’s Department, Springfield (Oregon) Police Department, and the many departments that have participated in the use-of-force database project of the International Association of Chiefs of Police.

Points of view expressed by contributors to this report do not necessarily represent the official positions or policies of the U.S. Department of Justice.

The National Institute of Justice and the Bureau of Justice Statistics are components of the Office of Justice Programs, which also includes the Bureau of Justice Assistance, the Office of Juvenile Justice and Delinquency Prevention, and the Office for Victims of Crime. Foreword

aw enforcement officers are authorized jurisdictions, and offers a researcher’s sug- Lto use force in specified circumstances, gestions for a future research agenda on are trained in the use of force, and typically police use of force, with special attention face numerous circumstances during their given to issues of excessive force. careers when use of force is appropriate—for Research consistently demonstrates that a example, in making some arrests, restrain- small percentage of police-public interac- ing unruly combatants, or controlling a dis- tions involve use of force. Various data ruptive demonstration. When the level of sources, including police use-of-force reports, force exceeds the level considered justifiable civilian complaints, victim surveys, and ob- under the circumstances, however, the activi- servational methods, confirm this basic find- ties of the police come under public scrutiny. ing. For example, the 1996 pilot test of the Incidents involving the use of excessive force PPCS found that about 1 percent of people by the police frequently receive attention reporting contacts with police said that offic- from the media, legislators, and, in some ers used or threatened force. Beginning in instances, civil and even criminal courts. July 1999, the PPCS is being fielded to a Whether the excessive force is aberrant much larger sample than responded to the behavior of individual officers or is a pattern 1996 test, and the results will be presented and practice of an entire law enforcement in a report next year. In the years ahead, it agency, both the law and public opinion is expected that the PPCS will provide the condemn such incidents. basis for a legislatively mandated annual report by the Attorney General documenting This report is one in a series of publications the prevalence of the use of excessive force. by the National Institute of Justice (NIJ) and the Bureau of Justice Statistics (BJS) NIJ-sponsored research at the local level that seek to inform public discussion by found that, in the context of the subset examining police use of force from many per- of police-public contacts involving adult spectives. The report provides an overview custody arrests, police used physical force of the state of research knowledge about (handcuffing excluded) in less than 20 per- police use of force, updates progress on the cent of 7,512 arrests studied (chapter 4). national BJS Police-Public Contact Survey Even in those instances, police primarily (PPCS) and the database project of the Inter- used weaponless tactics, such as grabbing or national Association of Chiefs of Police, pro- holding, which is consistent with the view vides the latest findings from NIJ-supported that relatively minor types of force dominate use-of-force research projects in several local statistics on police use of force. That view

iii Use of Force by Police

is further supported by research indicating public, enhancing the safety of the commu- that in incidents involving resistance by sus- nity and officers, and building widespread pects, their injuries resulting from police use support among those they serve. of force were typically minor (chapter 5). Ongoing research by NIJ and BJS seeks to Jeremy Travis provide the perspective, insight, and factual Director data needed by police and others to address National Institute of Justice use-of-force issues constructively. Through this and other policing research, we seek Jan M. Chaiken, Ph.D. to advance our goal of assisting law Director enforcement agencies in protecting the Bureau of Justice Statistics

iv Contents

Foreword ...... iii

Executive Summary ...... vii

1. What We Know About Police Use of Force ...... 1 by Kenneth Adams

2. Revising and Fielding the Police-Public Contact Survey ...... 15 by Lawrence A. Greenfeld, Patrick A. Langan, and Steven K. Smith

3. IACP National Database Project on Police Use of Force ...... 19 by Mark A. Henriquez

4. Measuring the Amount of Force Used By and Against the Police in Six Jurisdictions ...... 25 by Joel H. Garner and Christopher D. Maxwell

5. The Force Factor: Measuring and Assessing Police Use of Force and Suspect Resistance ...... 45 by Geoffrey P. Alpert and Roger G. Dunham

6. A Research Agenda on Police Use of Force ...... 61 by Kenneth Adams

Bibliography...... 75

v Executive Summary

ecent developments have heightened For example, about 1 percent of people who R concern about police use of force. They had face-to-face contacts with police said range from well-publicized incidents involv- that officers used or threatened force, ac- ing allegations of excessive force to the onset cording to preliminary estimates based on of “aggressive” policing, whose frequent the Bureau of Justice Statistics’ 1996 pretest emphasis on zero-tolerance enforcement is of its Police-Public Contact Survey (chapter sometimes regarded as encouraging use-of- 2). In 7,512 adult custody arrests, another force abuses. No matter what specific event study (chapter 4) notes that fewer than one triggers concern about police use of force, out of five arrests involved police use of how is the public to assess whether such physical force (defined as use of any weapon, force is, in the aggregate, a major problem? use of any weaponless tactic, or use of severe One way is to examine what research has restraints). That can be considered a low unearthed. rate in view of the study’s broad definition The organization of the of force. executive summary par- Overview: What Do We Know About Also known with substantial confidence is allels that of the report Police Use of Force? that police use of force typically occurs at the as a whole; that is, the lower end of the force spectrum, involving order of topics highlighted As discussed in chapter 1, research-based grabbing, pushing, or shoving. In the study in this summary tracks knowledge about police use of force can be focusing on 7,512 adult custody arrests, the chapter sequence. Oc- placed into three categories. The first perfor instance, about 80 percent of arrests in casional cross-references tains to knowledge that can be accepted with which police used force involved use of weap- to specific chapters are substantial confidence as “fact.” The second onless tactics. Grabbing was the tactic used intended to assist readers relates to use-of-force knowledge that can be about half the time. About 2.1 percent of all in locating more detailed accepted only with modest confidence be- arrests involved use of weapons by police. information. cause, for example, additional research is Chemical agents, such as pepper spray, were warranted. The third category consists of the weapons most frequently used (1.2 per- knowledge yet to be developed through cent of all arrests), with firearms least often research—that is, what is not yet known. used (0.2 percent). Known with substantial confidence From a police administrator’s point of view, these findings are predictable. Officers are Known with substantial confidence is that trained to use force progressively along a police use force infrequently. The data indi- continuum, and policy requires that officers cate that a small percentage of police-public use the least amount of force necessary to encounters involve force.

vii Use of Force by Police

accomplish their goals. The kinds of police areas where it is difficult to decide whether actions that most arouse the public’s con- an officer acted properly, given credible evi- cerns—such as fatal shootings, severe dence that use of force was necessary. beatings with fists or batons that lead to hospitalization, and choke holds that cause un- Known with modest confidence consciousness or even death—are not typical Regarding what is known with modest confi- of situations in which police use force. dence about police use of force, chapter 1 When injuries occur as a result of the use identifies three conclusions suggested by of force, they are likely to be minor. In one research data: study (chapter 5), researchers found that ● Use of force appears to be unrelated to an the most common injury to a suspect was a officer’s personal characteristics, such as bruise or abrasion (48 percent). age, gender, and ethnicity. This conclusion Another research finding that can be ac- should be accepted with caution, however. cepted with substantial confidence is that Additional verification is needed. use of force typically occurs when police are ● Use of force is more likely to occur when trying to make an arrest and the suspect is police are dealing with persons under the resisting. This conclusion is based on four influence of alcohol or drugs or with men- types of data: arrest statistics, surveys of po- tally ill individuals. Research findings in lice officers, observations of police behavior, this area are inconsistent, however. Fur- and reports by the public about their encoun- ther investigation, with an emphasis on ters with police. implications for training, could lead to a The foregoing findings leave open the issue reduction in the risk of force and injury of excessive force because issues of propor- for both police officers and civilians. tionality are not clearly addressed. Research ● A small proportion of officers are dispro- findings suggest, however, that many de- portionately involved in use-of-force bates over excessive force will fall into gray incidents. More research is needed.

About this report This report is one of a series of use-of-force sons age 12 or older. The other seeks to publications (see Bibliography, page 75) encourage as many local law enforcement generated by research supported by the agencies as possible to submit voluntarily National Institute of Justice or Bureau of and anonymously use-of-force data to a Justice Statistics. The data and findings central database for analysis. herein contribute to a better understanding Chapters 4 and 5 focus on the local level. of the extent and nature of police use of They present use-of-force findings based force and of the circumstances under which on data acquired from nine police agencies. such force is applied. The final chapter looks ahead by proposing A major objective of chapter 1 is to provide a research agenda on police use of force, an overview of what is known (and not with special attention given to issues of known) about police use of force and excessive force. thereby help readers put the issue in perspective. The next two chapters are updates Thus, this report begins with an overview of two national projects. One is designed to of what is known about police use of force, collect data on police-public contacts, includ- proceeds to outline what is being learned, ing those involving police use of force, from and concludes with a proposed plan for a nationally representative sample of per- future research.

viii Executive Summary

Among what is not known lead to excessive force by police seem obvious, or appear to be a matter of common As stated in chapter 1: “The incidence of sense, a great need for systematic re- wrongful use of force by police is unknown. search in this area exists. Research is critically needed to determine reliably, validly, and precisely how often ● Influences of situational characteristics transgressions of use-of-force powers occur.” on police use of force and the transactional nature of these events are largely Researchers and practitioners both tend to unknown. For example, little is known presuppose that the incidence of excessive beyond research indicating that situations force by police is very low. If use of force is most likely to involve police use of force uncommon, and civilian complaints are in- are interpersonal disturbance and violent frequent, and civilian injuries are few, then personal crime, and situations when sus- excessive force by police must be rare. That pects attempt to flee or physically resist conclusion may indeed be correct, but to the arrest. Those findings, however, do not extent that it hinges on official police statis- address the transactional, or step-by-step tics, it is open to serious challenge. unfolding, of police-public encounters. Was Current indicators of excessive force, such as suspect resistance the result of police use civilian complaints and civil lawsuits, are all of force, or did police use force after expe- critically flawed. The difficulties in measur- riencing suspect resistance? ing excessive force with complaint and lawsuit records have led academics and Updates on Two National Projects practitioners to redirect their attention to In 1996, the Bureau of Justice Statistics all use-of-force incidents. Theoretically, un- (BJS) and the International Association of derstanding all use-of-force incidents helps Chiefs of Police (IACP) initiated projects input wrongful use of force in perspective. volving collection of data encompassing po- As one example of how understanding all lice use of force. Both are currently ongoing. use-of-force incidents can help put excessive force in perspective, the study of 7,512 adult The BJS survey custody arrests (chapter 4) makes this To learn more about police use of force re- observation: quires an understanding of the reasons for “ . . . most arrests involve no force, excessive and the results of police-public encounters. or otherwise. When force is used, it typically As a step toward developing that under- involves less severe forms of tactics and standing, BJS supplemented the National weapon use. These findings provide a context Crime Victimization Survey with a pilot test “NCVS is based on inter- for understanding excessive force, which we of its Police-Public Contact Survey (PPCS) in views conducted with a know can involve low-level acts of force . . . 1996 (chapter 2). nationally representative as well as the acts of force that result in sample of U.S. house- Among the findings was a preliminary physical injury or death of civilians. Arrests holds and has become a estimate that about 1 percent of people re- that involve no force, however, cannot in- highly useful platform for porting contacts with police indicated that volve excessive force and arrests that involve testing new question- officers used or threatened force. In the low levels of force are less likely to involve naires and periodically majority of those instances, respondents said excessive force.” implementing supple- that their own actions, such as threatening ments.” —Greenfeld, et Additional gaps in use-of-force knowledge police, may have provoked officers. include the following: al., page 15 of this report. In July 1999, a second test of PPCS was ● The impact of differences in police organi- fielded to a much larger sample than that zations, including administrative policies, used in the 1996 pilot test. In addition, BJS hiring, training, discipline, and use of anticipates adding items to its periodic sur- technology, on excessive force is unknown. veys conducted among nationally representa- Although many conditions that arguably tive samples of those confined in local jails

ix Use of Force by Police

and prisoners held by State and Federal the force factor, a measure of the level of authorities. The new survey items would force used by officers relative to the level of provide, for the first time, information about resistance by suspects. respondents’ interactions, including use of force, with police during the arrest preceding Study on the amount of force used in incarceration. six jurisdictions

The IACP database project The six-jurisdiction study (chapter 4) gathered data about officers’ and suspects’ behav- Initiated in 1996, the IACP database project iors in connection with 7,512 adult custody is designed to collect use-of-force information arrests (arrests in which suspects are trans- from law enforcement agencies across the ported to a detention facility, in contrast to “...IACP designed the Nation (chapter 3). To promote accurate re- being issued a summons to appear before a project from the outset to porting and overcome potential reluctance judicial officer). The researchers focused on reflect operational reali- of agencies to participate, IACP decided that the amount of force used by and against po- ties of modern, street-level provision of data would be both voluntary lice, with the expectation that this informa- law enforcement, includ- and anonymous. tion would inform issues surrounding the ing the very meaning of use of excessive force. For instance, excessive Collected data pertain to reported use of ‘police use of force,’ de- force is typically but not necessarily associ- force stemming from police responses to calls fined as the amount of ated with more severe forms of force that for service, whether or not those responses force required by police to could or do result in injury or death. resulted in arrests. About 150 agencies are compel compliance by an expected to contribute data for the 1998–99 Emerging from the research is a more complete unwilling subject.” data year. Among preliminary findings: understanding of the frequency with which cer- —Henriquez, page 20 of tain types of tactics are used and what types of this report. ● Based on 1995 data reported by 110 agen- weapons are displayed, threatened, or actually cies, the police use-of-force rate was 4.19 used. The consistent findings across all six per 10,000 responded-to calls for service, jurisdictions are that most arrests (more than or 0.0419 percent. 80 percent) did not involve force by police (ex- ● Based on data reported for 1996–97, 87 cluding handcuffing) or by suspects. In 98 per- percent of 62,411 use-of-force incidents cent of arrests where force was used, no weapon involved officers using physical force. Of- was used, threatened, or even displayed. When ficers used chemical force in 7 percent of police used some form of weaponless tactic (hit- the incidents, firearms in about 5 percent. ting, kicking, wrestling, etc.), the most frequent tactic involved only grabbing (about half the ● Based on available data for 1996–97, about time). 10 percent of 2,479 officers using force sustained injuries, less than 1 percent serious. In addition to providing data on the use of About 38 percent of subjects were injured weapons and weaponless tactics, the study due to police use of force, with 1.5 percent identified three other elements, sometimes sustaining major injuries. included in the concept of use of force:

Because the data are not yet nationally ● Restraints. In about 82 percent of all representative, conclusions about national 7,512 arrests, officers reported use of use-of-force trends should not be attempted. handcuffs. Leg cuffs were used in 0.9 percent of arrests. Officers used more severe Two Local-Level Studies restraints in 0.4 percent of arrests. Re- straints were not used in approximately One of the local-level studies measured the 16 percent of arrests. amount of force used by and against police in six jurisdictions. The other measured and ● Motion. Pursuit on foot and by car oc- assessed police use of force and suspect re- curred in 3 percent and 2.4 percent, resistance in three jurisdictions and developed spectively, of all arrests. Pursuit did not

x Executive Summary

occur in 94.4 percent of arrests. Suspect The two-city Oregon site of Eugene/ flight most frequently occurred by foot (in Springfield. Researchers analyzed 562 po- 4.7 percent of arrests). In 93.5 percent of lice actions, 57 percent of which were taken arrests, suspects did not flee. by officers responding to calls for service and 33 percent by officers reacting to situations ● Voice. In 61.2 percent of arrests, police they had observed. The most common type of reported they used a conversational tone incident (25 percent) confronting officers was with suspects. street violence. Most police action (76 per- Among the measures of force used by police cent) was taken to apprehend or control a officers that were developed by the research- person. ers are physical force and physical force plus Officers often used more than one verbal or threats. The study found that 17.1 percent physical control tactic per incident. For in- of arrests involved physical force (use of a stance, 93 percent of 546 incidents involved weapon, weaponless tactic, or severe re- at least two tactics; 87 percent, at least three; straint) and that 18.9 percent entailed and 41 percent, at least four. The pattern of physical force or the display or threatened tactic use that emerged corresponds to the use of any weapon. traditional use-of-force continuum. The first To better distinguish between different types tactic used in an incident is nearly always of force—such as between grabbing and kick- the least severe use of force on the con- ing—the researchers developed a maximum tinuum; the second is almost always the sec- force measure, which involved officers’ rank- ond-most lenient; and so on, with very few ing 60 hypothetical types of force in terms of exceptions. their severity on a scale from 1 (least force- Of 504 reported incidents in which force was ful) to 100 (most forceful). When the mea- used, 1.8 percent resulted in injury to officers. sure was applied to the types of force officers They were most at risk for injury when wres- reported using, the study found that the tling, striking, or taking a suspect to the ranking score for commanding a suspect to ground. do something (1.3 percent of all arrests) was 22; for using handcuffs (57.3 percent of The level of force used by the department’s arrests), 28.2; and for displaying a handgun officers relative to the amount of the sus- (2.2 percent of arrests), 55.4. The average pects’ resistance—the force factor—averaged ranking score for the types of force used in slightly higher than the amount of resistance all arrests was 30. encountered. On average, more force than resistance was used. This does not necessar- The researchers state that their findings ily imply that the level of police force was are beginning to provide a stable picture of excessive. For example, an officer may justifi- police behavior and the amount of force that ably use more force than does a suspect to police use in arrest situations, but they note gain control of a situation. the findings remain tentative given the small number of jurisdictions involved in The Florida site. The study focused on data the research, among other reasons. in 882 official Control-of-Persons Reports prepared by officers’ supervisors in the Study on police use of force and Miami-Dade Police Department. Ninety- suspect resistance seven percent of suspects resisted. The study collected use-of-force data from The type of resistance most often reported three law enforcement agencies— police was actively resisting arrest (36 percent), fol- departments in two Oregon cities and one lowed by assaulting the officer (25 percent). county department in Florida (chapter 5). Twenty one percent of suspects attempted to The areas served by the Oregon departments escape or flee the scene. The most common were considered one site. type of force used by suspects was striking or

xi Use of Force by Police

hitting the officer (44 percent). Initially calm A Proposed Research Agenda suspects were least likely to resist officers but were the most likely to flee and the most The development of a research agenda on likely to resist with a gun or assault officers police use of force, with special attention with a vehicle. given to issues of excessive force, should be guided by these general considerations. The most common type of suspect injury was a bruise or abrasion (48 percent of those in- ● Research should provide new knowledge jured), followed by lacerations (24 percent), that significantly increases our under- and gunshot injuries (4 percent). The chance standing of the problem. of suspect injury was significant no matter ● Research should be policy relevant. what type of force was used by police. For example, officer use of fists entailed an 81 ● Research activities, taken as a whole, percent chance of suspect injury; use of a should be comprehensive and systematic. PR-24 baton, a 67 percent chance; and use of Within that general framework, more work a handgun, a 48 percent chance. is required on what various people—general Suspects who were reportedly impaired by public, minorities, police administrators, pa- alcohol or drugs were no more likely to resist trol officers, judges, offenders, etc.— have in officers than sober suspects. When they did mind when they refer to excessive force and resist, however, they were more likely than how they adjudge specific instances of police nonimpaired suspects to directly assault the behavior when questions of excessive force officer and more than twice as likely to use arise. This research is important because so- a gun. cial problems often require shared solutions, and shared solutions require a common basis The most common type of force used by officers of understanding and mutual respect for was use of hands and arms (77 percent of use- differences in views. of-force incidents). In 64 percent of incidents, officers grabbed or held suspects. There were Also needed is more and better data on no statistically significant differences in the police use of force. Most discussions occur in level of force used by male and female officers. an empirical vacuum where arguments are The ethnicity of an officer did not affect the made without the benefit of solid, useful general level of force used or whether force was information. used. Research is required on how use of force by Data suggest that officers are significantly police varies across time, cities, and indi- at risk for injury when they use force, par- vidual police departments. Research also is ticularly when they strike a suspect with needed on individual, situational, and orga- their fists (48 percent chance) or use their nizational factors related to variations in hands and arms to control a suspect (43 per- use-of-force levels, along with excessive force cent chance). Because most use-of-force inci- levels and should focus on the relation be- dents involved use of hands, arms, or fists tween excessive use of force, meaning the by officers, they are most at risk for injury frequency with which police use force, and when using precisely the types of force that excessive force, meaning instances in which they report using most frequently. police use more force than is necessary. Police officers’ use of force in relation to sus- Finally, interventions, changes, and reforms pect resistance—the force factor—averaged that may mitigate police use-of-force prob- slightly less force than the resistance en- lems should be identified, documented, and countered. Data indicate that officers are evaluated. more likely to be injured when using less force than that used by resisting suspects.

xii 1 What We Know About Police Use of Force by Kenneth Adams

mbrose Bierce, a social critic known for operate as members of the community they A his sarcasm and wit, once described the serve. The community, in turn, enters into a police as “an armed force for protection and solemn and consequential relationship with participation.”1 In this pithy statement, the police, ceding to them the power to de- Bierce identifies three critical elements of prive persons of “life, liberty, and the pursuit the police role. First, by describing the police of happiness” at a moment’s notice and de- as “armed,” their ability to coerce recalci- pending on them for public safety. Without trant persons to comply with the law is em- police, the safety of the community is jeopar- phasized. Because police carry weapons, dized. Without community support, police it follows that the force they use may have are dispossessed of their legitimacy and lethal consequences. The capacity to use robbed of their effectiveness. coercive, deadly force is so central to under- This three-element definition of police standing police functions, one could say that makes it easy to understand why abuse of Kenneth Adams, Ph.D., it characterizes a key element of the police force by police is of such great concern. First, is Associate Professor role. there is the humanitarian concern that po- and Chair of the Crimi- Second, the primary purpose of police is lice are capable of inflicting serious, even le- nal Justice Faculty, protection, and so force can be used only to thal, harm on the public. Second, there is the School of Public and promote the safety of the community. Police philosophical dilemma that in “protecting” Environmental Affairs, have a responsibility for safeguarding the the whole of society, some of its constituent Indiana University– domestic well-being of the public, and this parts, meaning its citizens, may be injured. Indianapolis. obligation even extends in qualified ways to Third, there is the political irony that police, protecting those who violate the law, who are who stand apart from society in terms of au- antagonistic or violent toward the police, or thority, law, and responsibility, also are part who are intent on hurting themselves. In of society and act on its behalf. Thus, rogue dealing with such individuals, police may actions by a few police, if condoned by the use force in reasonable and prudent ways to public, may become perceived as actions of protect themselves and others. However, the the citizenry. amount of force used should be proportional Recent developments in policing have el- to the threat and limited to the least amount evated concerns about police use of force required to accomplish legitimate police beyond ordinarily high levels. In particular, action. community policing, which is becoming wide- Third, the concept of participation empha- spread as a result of financial incentives by sizes that police and community are closely the Federal Government, and “aggressive” interrelated. Police are drawn from the policing, which is becoming widely adopted community, and as police they continue to as a solution to serious crime problems, have

1 Use of Force by Police

come to the fore as perspectives of choice Use-of-force concerns also are reflected in by policing experts. Community policing the attention the media give to possible emphasizes the role of the community as instances of police abuse. An accumulation “coproducers” of law and order in conjunction of alleged abuse-of-force incidents, widely with the police. Communities naturally vary reported in the media, encourages over- in attributes, and they vary in how they are generalization by giving the impression that defined for the purposes of community polic- police brutality is rampant and that police ing. Consequently, some communities look to departments across the Nation are out of add restrictions on police use of force, while control. For example, Human Rights Watch others are satisfied with the status quo, and states, “Allegations of police abuse are rife in still others seek to ease current restrictions. cities throughout the country and take many Regardless of the community’s orientation forms.”3 on this issue, community policing means in- Before considering the details of recent creased levels of accountability and respon- research efforts on police use of force, it is siveness in key areas, such as use of force. useful to summarize the state of our knowl- Increased accountability hinges on new in- edge.4 We know some details about police formation, and new information stimulates use of force with a high degree of certainty. debate. These items represent “facts” that should The other emerging perspective is “aggres- frame our understanding of the issues. Other sive” policing, which often falls under the details about police use of force we know in rubric of broken windows theory, and, as a sketchy ways, or the research is contradic- strategic matter, is concerned with intensify- tory. These items should be subject to addi- ing enforcement against quality-of-life and tional research using more refined methods order maintenance offenses. The influence of of inquiry. Finally, there are some aspects of aggressive policing can be seen in the prolif- police use of force about which we know very eration of “zero tolerance” enforcement strat- little or next to nothing. These items repre- egies across the Nation. The concern is that sent critical directions for new inquiry. the threat posed by petty offenders may be As is often the case with important policy exaggerated to the point that use of force questions, the information that we are most becomes more commonplace and abuses of confident of is of limited value. In many force more frequent. cases, it does not tell us what we really need The Violent Crime Control and Law Enforce- to know, because it does not focus squarely ment Act of 1994 mirrored congressional on the important issues or is subject to concern about excessive force by authorizing competing interpretations. Conversely, the the Civil Rights Division of the U.S. Depart- information that is most critical for policy ment of Justice (DOJ) to initiate civil actions decisions often is not available or is very dif- against police agencies when, among other ficult to obtain. Such is the case with police conduct, their use of force reaches a level con- use of force. The issues that most concern stituting a pattern or practice depriving indi- the public and policymakers lack the kinds viduals of their rights. DOJ exercised that of reliable and solid information that ad- authority when, for example, it determined vance debate from the realm of ideological that an urban police department engaged in posturing to objective analysis. Nonetheless, such conduct and negotiated a consent decree it is important to take stock of our knowl- that put in place a broad set of reforms, in- edge so that it is clear which issues can be cluding an agreement by the department to set aside and which should be the target of document its use of force and to implement efforts at obtaining new knowledge. an early warning system to detect possible abuses.2

2 Chapter 1: What We Know About Police Use of Force

What, then, is the state of knowledge regard- police-public “interactions,” such as calls for ing police use of force? We begin with issues service, which capture variegated requests about which we have considerable informa- for assistance, lead to low rates of use of tion and a high degree of confidence in our force. Conversely, narrow definitions of knowledge. Discussed next are issues where police-public interactions, such as arrests, knowledge is modest and considerably more which concentrate squarely on suspects, lead research is merited. Finally, we conclude to higher rates of use of force. with issues that are critical to debates over The International Association of Chiefs of police use of force and about which little Police (IACP) is in the process of compiling knowledge exists. statistics on use-of-force data being submitted by cooperating agencies (see chapter 3). What We Know With Substantial These data indicate that force is used in Confidence About Police less than one-half of 1 percent of dispatched Use of Force calls for service. From this point of view, one might well consider police use of force a rare Police use force infrequently. event. This figure is roughly consistent with Whether measured by use-of-force reports, the preliminary estimate reported by BJS, citizen complaints, victim surveys, or obser- although the IACP figure is subject to the vational methods, the data consistently reporting biases that may exist in police indicate that only a small percentage of agency data. Furthermore, IACP data are police-public interactions involve the use not yet representative of the national picture of force. As Bayley and Garofalo observed, because of selection bias; the estimate is “Thus, the Commission police-citizen encounters that involve use of based on a small percentage of police depart- concludes that factors force and injury are “quite rare.”5 ments that voluntarily report information on substantially contributing use of force. Because there is no standard methodology to misperceptions about for measuring use of force, estimates can Garner and Maxwell found that physical use of physical and vary considerably on strictly computational force (excluding handcuffing) is used in deadly force by law grounds. Different definitions of force and fewer than one of five adult custody arrests enforcement officers different definitions of police-public interac- (see chapter 4). While this figure hardly include...[f]ailure to ap- tions will yield different rates6 (see sidebar qualifies as a rare event, it can be considered preciate the relative infre- “Working definitions”). In particular, broad low, especially in light of the broad definition quent use of physical and definitions of use of force, such as those that of force that was used. deadly force by law en- include grabbing or handcuffing a suspect, In characterizing police use of force as infre- forcement personnel....” will produce higher rates than more conser- quent or rare, the intention is neither to —New York State Com- vative definitions. The Bureau of Justice minimize the problem nor to suggest that mission on Criminal Jus- Statistics’ (BJS) 1996 pretest of its Police- the issue can be dismissed as unworthy of tice and the Use of Force, Public Contact Survey resulted in prelimi- serious attention. Society’s ends are best Report to the Governor, nary estimates that nearly 45 million people achieved peaceably, and we should strive to Vol. 1, New York: New had face-to-face contact with police over a minimize the use of force by police as much York State Commission 12-month period and that approximately as possible. However, it is important to put on Criminal Justice 1 percent, or about 500,000 of these persons, police use of force in context in order to un- and the Use of Force, were subjected to use of force or threat of derstand the potential magnitude of use-of- May 1987: 6. 7 force (see chapter 2). When handcuffing is force problems. Although estimates may not included in the BJS definition of force, the completely reassure everyone that police are number of persons increases to 1.2 million. doing everything they can to minimize the Expanding and contracting definitions of use of force, the data do not support the “police-public” interactions also work to af- notion that we have a national epidemic of fect use-of-force rates but in an opposite way police violence. from definitions of force. Broad definitions of

3 Use of Force by Police

Another purpose for emphasizing the infre- derived from such a small number of respon- quent nature of police use of force is to high- dents are subject to a wide margin of error. light the methodological challenges of trying This issue is particularly important if one to count or study infrequent events. In this is interested in tracking changes over time, regard, methodological approaches can because a very small change in reporting can vary considerably in terms of cost efficiency, have a very large impact on estimates. In the reliability, and precision of information ob- survey’s continuing development, the next tained. In BJS’s 1996 pilot household survey pilot test will use a sample about 10 times of 6,421 persons, 14 respondents, or roughly the size of the 1996 pilot test as well as 1 in 450, said that they were subjected to involve a redesigned questionnaire. use of force or threat of force by police over a Police use of force typically occurs at the year’s time. The household survey approach lower end of the force spectrum, involv- has the benefit of providing national-level ing grabbing, pushing, or shoving. estimates based on data that are free of police agency reporting biases. However, as Relatively minor types of force dominate noted by BJS, the preliminary estimates statistics on police use of force. Garner and

Working definitions Police use of force is characterized in a va- Reference also is made to “excessive use riety of ways. Sometimes, these character- of force,” a similar, but distinctly different, izations are functionally interchangeable term. Excessive use of force refers to high so that one can be substituted for another rates of force, which suggest that police are without doing injustice to the factual inter- using force too freely when viewed in the pretation of a statement. At other times, aggregate. The term deals with relative however, differences in terminology can be comparisons among police agencies, very consequential to a statement’s mean- and there are no established criteria for ing. For example, “deadly force” refers to judgment. situations in which force is likely to have “Illegal” use of force refers to situations in lethal consequences for the victim. This which use of force by police violated a law type of force is clearly defined and should or statute, generally as determined by a not be confused with other types of force judge or magistrate. The criteria for judg- that police use. ing illegal use of force are fairly well In contrast, “police brutality” is a phrase established. used to describe instances of serious “Improper,” “abusive,” “illegitimate,” and physical or psychological harm to civilians, “unnecessary” use of force are terms that with an emphasis on cruelty or savage- describe situations in which an officer’s ness. The term does not have a standard- authority to use force has been mishandled ized meaning; some commentators prefer in some general way, the suggestion being to use a less emotionally charged term. that administrative procedure, societal ex- In this report, the term “excessive force” is pectations, ordinary concepts of lawfulness, used to describe situations in which more and the principle of last resort have been force is used than is allowable when judged violated, respectively. Criteria for judging in terms of administrative or professional these violations are not well established. guidelines or legal standards. Criteria for To varying degrees, all of the above terms judging excessive force are fairly well es- can be described as transgressions of tablished. The term may also include within police authority to use force. its meaning the concept of illegal force.

4 Chapter 1: What We Know About Police Use of Force

Maxwell (see chapter 4) observed that police most frequently used (1.2 percent of arrests), use weaponless tactics in roughly 80 percent while firearms were the weapons least often of use-of-force incidents and that half the used (0.2 percent of arrests). Most police time the tactic involved grabbing the sus- departments collect statistics on all firearm pect. Alpert and Dunham (see chapter 5) discharges by officers. These data consis- found that in Miami 64 percent of use-of- tently show that the majority of discharges force incidents involved grabbing or holding are accidental or are directed at animals. the suspect. In the BJS pilot national survey, Only on infrequent occasions do police use it was estimated, preliminarily, that about their firearms against the public. One impli- 500,000 people were “hit, held, pushed, cation of these findings is that increased choked, threatened with a flashlight, retraining in how to use standard police weap- strained by a police dog, threatened with or ons will be of little value in dealing with day- actually sprayed with chemical or pepper to-day situations that involve use of force. spray, threatened with a gun, or experienced Training, if it is to be effective in reducing some other form of force.”8 Three-fifths of the use of force, needs to focus on how to these situations, however, involved only gain compliance without resorting to physi- holding. Finally, Pate and Fridell’s survey of cal coercion. law enforcement agencies regarding use of Use of force typically occurs when police force and civilian complaints also confirms are trying to make an arrest and the that minor types of force occur more fre- suspect is resisting. quently than serious types.9 “The first tactic used in Research indicates that police are most an incident is nearly al- As a corollary finding, when injuries occur as likely to use force when pursuing a suspect ways the least severe use a result of use of force, they are likely to be and attempting to exercise their arrest pow- of force on the continuum, relatively minor. Alpert and Dunham (see ers. Furthermore, resistance by the public and the second…is nearly chapter 5) observed that the most common increases the likelihood that police will always the second-most injury to a suspect was a bruise or abrasion use force. These findings appear intuitively lenient.” —Alpert and (48 percent), followed by laceration (24 per- sound given the mandate that police have Dunham, page 48 of this cent). The kinds of police actions that most regarding use of force. Police may use force report. captivate the public’s concerns, such as fatal when it is necessary to enforce the law or to shootings, severe beatings with fists or ba- protect themselves or others from harm. The tons that lead to hospitalization, and choke findings also seem logical in view of police holds that cause unconsciousness or even training curriculums and departmental death, are not typical of situations in which regulations. Alpert and Dunham (see chap- police use force. These findings reassure us ter 5) find that police almost always follow that most police exercise restraint in the use the prescribed sequence of control proce- of force, even if one has concerns over the dures they are taught, except when suspect number of times that police resort to serious resistance is high, in which case they tend to violence. skip the intermediate procedure. From a police administrator’s point of view, The conclusion that police are most likely to these findings are predictable. Officers are use force when dealing with criminal sus- trained to use force progressively along a pects, especially those who are resisting continuum, and policy requires that officers arrest, is based on four types of data: arrest use the least amount of force necessary to statistics, surveys of police officers, observa- accomplish their goals. tions of police behavior, and reports by the Another affiliated finding is that police public about their encounters with police. rarely use weapons. According to Garner and Arrest statistics show that resisting-arrest Maxwell (see chapter 4), 2.1 percent of adult charges often are involved in situations in custody arrests involved use of weapons by which officers use force. The interpretation police. Chemical agents were the weapons

5 Use of Force by Police

of this finding is ambiguous, however, be- arrests involved use of force by police, 14 cause officers may bring such charges in an percent of arrests involved use of force by attempt to justify their actions against a suspects. Police officers in Phoenix com- suspect. Some commentators even would ar- pleted a use-of-force survey after each arrest gue that resisting-arrest charges are a good to generate these data. indication that police officers acted inappro- Finally, Bayley and Garofalo tallied 36 in- priately or illegally. Because we are relying stances of force used by police or suspects on official reports by officers who are in- out of 467 police-public encounters observed volved in use-of-force incidents, and because firsthand by researchers.12 They found that they have self-interest in presenting the in 31 incidents police used force against sus- situation in the most favorable light possible, pects and in 11 incidents suspects used force we cannot rely on arrest records alone in against police. determining what happened. One implication of the research is that the Fortunately, other research is available to decision to use some level of force probably help clarify the situation. The pilot national has legal justification in most cases. Force is household survey by BJS included a series of likely to be used when suspects resist arrest questions about the respondent’s behavior and attempt to flee. Also, in a significant during contact with police.10 The preliminary number of instances, suspects use force analysis revealed that of the 14 respondents against the police. These findings leave open in the sample who reported that police used the issue of excessive force, since issues of or threatened force against them, 10 sug- proportionality are not clearly addressed. gested that they might have provoked the However, the findings do suggest that many officer to use force. The provocative behav- debates over excessive force will fall into iors reported by suspects include threaten- gray areas where it is difficult to decide ing the officer, assaulting the officer, arguing whether an officer acted properly, because with the officer, interfering with the arrest of there is credible evidence that the use of someone else, blocking or interfering with an force was necessary. officer’s movement, trying to escape, resisting being handcuffed, and resisting being placed in a police vehicle. What We Know With Modest Confidence About Police Use Research by Alpert and Dunham (see chap- Regarding suspect force ter 5) confirms that criminal suspects are of Force as a consistent predictor not always cooperative when it comes to Use of force appears to be unrelated to of police use of force: “This arrest. In almost all (97 percent) cases in an officer’s personal characteristics, remained true when con- which police officers used force in a Florida such as age, gender, and ethnicity. trolling for the possibility jurisdiction, the suspect offered some degree that some suspect use of of resistance. In 36 percent of use-of-force A small number of studies suggest that use force could be a incidents, the suspect actively resisted ar- of force by police is not associated with per- reaction to police use of rest, and in one-quarter of the incidents the sonal characteristics, such as age, gender, force.” —Garner, et al. (see suspect assaulted the officer. The research- and ethnicity. Bayley and Garofalo concluded note 11). ers observed that the most common type of that use of force is not related to age, al- 13 suspect force was hitting or striking a police though it may be related to experience. officer (44 percent). Worden, in an analysis of observational data on 24 police departments in 3 metropolitan Garner and colleagues, after using statistical areas, concluded that the personal character- controls for more than 50 characteristics of istics of police officers do not have a substan- the arrest situation, the suspect, and the po- tively significant effect on use of force.14 lice officer, found that forceful action by suspects was the strongest and most consistent Likewise, Garner and colleagues reported predictor of use of force by police.11 Further- that the race of suspect and officer is not 15 more, they found that while 22 percent of predictive of use of force. However, they

6 Chapter 1: What We Know About Police Use of Force found that incidents involving male police against suspects of their own race. The lack of officers and male suspects are more likely to relationship between race and use of force, as involve force. Alpert and Dunham (see chap- well as between gender and use of force, is ter 5) found that officer characteristics are of probably disheartening to those who argue little utility in distinguishing between force that integration of police agencies along ra- and nonforce incidents. cial and gender lines will do much to reduce the incidence of police violence. Again, more Hence, gender and ethnicity appear unre- research is needed to understand the situa- lated to use of force. Given the limited re- tion of minority and female police officers search in this area, these conclusions should with regard to their use of force. be accepted with caution and additional verification of these findings is needed. Use of force is more likely to occur when police are dealing with persons under It is widely accepted in criminology that the influence of alcohol or drugs or with violence, along with a wide variety of other mentally ill individuals. More research risk-taking and norm-violating behaviors, is is needed. a young man’s game. Thus, we should expect that young, male police officers should use Police come across a wide variety of situa- force more than their female colleagues or tions in their work. They encounter problems older officers. The fact that this is not clearly that range from relatively minor to serious the case seems surprising. to potentially deadly. They also interact with people exhibiting various mental states, in- A lack of relationship between age and gen- cluding persons who are hysterical, highly der, on the one hand, and use of force, on the agitated, angry, disoriented, upset, worried, other, may be a function of police hiring and irritated, or calm. deployment practices. Retirement plans keep the age of police officers lower than that of Two situations that often give police officers most other occupations, and seniority, which cause for concern are when suspects appear is derivative of work experience, often brings to be under the influence of alcohol or drugs more choice in work assignments, including and when civilians appear to suffer from duties that limit one’s contact with criminal serious mental or emotional impairments. suspects on the street. Both these tendencies The concern stems from the fact that in such serve to constrain variation in the age of situations a person’s rational faculties ap- “Suspects reported as police officers who are exposed to potentially pear impaired. In dealing with problem situ- impaired were more than violent situations. This may attenuate the ations, officers most often talk their way, twice as likely than sober relationship between age and use of force. rather than force their way, into solutions. suspects to use a gun to However, it is equally plausible that young For this reason, when a civilian is in a highly resist the police.” —Alpert male officers are assigned to high-crime irrational state of mind, the chances of the and Dunham, page 51 of areas where frequent use of force is neces- police officer having to use force presumably this report. sary to gain compliance. Finally, it is possible increase and the possibility of injury to both that exposure to the police culture works to officer and civilian increases as well. encourage the use of force, thus counterbal- Research carried out for the President’s ancing the decline in aggressivity that comes Commission on Law Enforcement and Ad- with age as demonstrated in criminological ministration of Justice observed that alcohol studies. More research is needed to disen- use by either a suspect or an officer in- tangle these relationships. creased the chances that force will be used.16 The finding that an officer’s race is unrelated Garner and colleagues found that alcohol to the propensity to use force runs counter to impairment by suspects was a consistent the argument that racial animosity lies at predictor of police use of force, while drug the heart of police abuse. Indeed, Alpert and impairment predicted increased use of force Dunham’s research (see chapter 5) indicates for some but not all measures of use of that officers are more likely to use force force.17 In contrast, Alpert and Dunham (see

7 Use of Force by Police

chapter 5) observed that alcohol or drug im- We often are told that a small number of pairment of suspects was unrelated to police people are responsible for most of the pro- use of force or subsequent injury. That find- ductive or counterproductive work in an or- ing is interesting because, although im- ganization. For example, we hear about the paired civilians did not demonstrate an 80/20 rule in organizational management. increased propensity to resist an officer’s That is, 20 percent of the workers account actions, when they did resist they were more for 80 percent of the work. Policing has its inclined to do so by actively resisting or as- counterpart explanation for deviant or ille- saulting the officer. gal behavior. It is called the rotten apple or Part of the disparity in findings between the rogue officer theory, and it is often used to President’s Commission’s research and more explain police corruption. Recently, a varia- recent studies may be attributed to the fact tion of this theory has become the principal that police officers today are better trained explanation for use-of-force problems in po- in how to deal with impaired civilians. Most lice departments. In this context, we speak of police officers now receive training in a vari- “violence prone” police officers and we point ety of violence reduction techniques, and this to these individuals as the reason why a development is partly attributable to con- department has problems with the use of 18 cerns over the President’s Commission’s force. findings and over the frequency with which People with extraordinary work perfor- police now are called to respond to large- mance, either good or bad, are noticeable scale violence, such as riots. when compared with their colleagues, and Questions about how police deal with civil- their salience leads us to think that their ians who appear to have impaired mental work is highly consequential to the good for- states are important from administrative tunes or misfortunes of an organization. The and practical points of view. Police officers utility of this perspective for police manag- are expected to exercise restraint in dealing ers attempting to deal with illegitimate use with impaired civilians, while at the same of force lies in the presumed concentration of time they need to be cautious about protect- problem behaviors in the work force. If only ing their safety as well as the safety of other a handful of police officers accounts for most civilians. This puts them in a precarious of the abuses, then effective solutions tar- situation, one in which mistakes of judgment geted at those individuals should deal with or tactics can have grave consequences. the problem. The nature of the solution, be it employee selection, training, oversight, or From a practical standpoint, police regularly discipline, is less important than its degree encounter civilians with impaired mental of effectiveness and its ability to be directed states, which makes the problem more than at the problem group of employees. academic. Alpert and Dunham (see chapter 5) found that in 42 percent of use-of-force The Christopher Commission, which investi- situations, suspects appeared to be under gated the Los Angeles Police Department the influence of alcohol or drugs. Overall, the subsequent to the Rodney King incident, 19 research on whether police use force more highlighted the “violence prone” officer theory. frequently in relation to civilians with im- The Commission, using the department’s paired mental states is inconsistent. Further database, identified 44 officers with 6 or more investigation, with an emphasis on implica- civilian allegations of excessive force or im- tions for training, could reduce the risk of proper tactics in the period 1986 through force and injury for both police officers and 1990. For the 44, the per-officer average for civilians. force-related complaints was 7.6 compared with 0.6 for all officers identified as having A small proportion of officers are dis- been involved in a use-of-force incident for the proportionately involved in use-of-force period January 1987 through March 1991. The incidents. More research is needed. 44 officers were involved in an average of 13

8 Chapter 1: What We Know About Police Use of Force use-of-force incidents compared with 4.2 for For example, an officer’s work assignment all officers reported to be using force. may involve a high-crime area that contains a high proportion of rebellious offenders. Put another way, less than one-half of Also, divisive, dehumanizing views of the 1 percent of the department’s sworn officers world, such as “us-them” and “good guy-bad accounted for more than 15 percent of alle- guy,” that facilitate violent behavior may be gations of excessive force or improper tac- supported by the organizational culture. tics. The degree of disproportion (30:1) is Further, administrative views of work roles striking and suggests that focusing efforts and products, communicated formally or on a handful of officers can eliminate informally, that emphasize crime control roughly 1 out of 7 excessive force incidents. through aggressive police behavior may This finding has led many police depart- encourage confrontational tactics that in- ments to implement early warning systems crease the chances of violent behavior by designed to identify high-risk officers before either civilian or police officer. Unless the they become major problems. Most of these reasons for violence propensity are accu- systems use administrative records, such as rately identified, the effectiveness of inter- disciplinary records and citizen complaints, ventions targeted at violent police officers is to monitor officer performance for possible a hit-or-miss proposition. problems. Of the 44 officers identified by the Christo- The concept of an early warning system for pher Commission in 1991, 14 subsequently “...a significant number of risk management of problem police officers is left the department as of October 1997. Of officers...repetitively mis- not new. In the early 1980s, a report on police the 30 remaining officers, two had a use-of- use force and persistently practices by the United States Commission force complaint that was sustained after ignore the written policies on Civil Rights found that “‘(e)arly warning’ review between 1991 and 1997.23 This low and guidelines of the De- information systems may assist the depart- number may be due to a variety of reasons, partment regarding force. ment in identifying violence-prone officers.”20 such as difficulties in sustaining citizen By their misconduct, this Consequently, it was recommended that “(a) complaints, reassignment of work duties, group of officers tarnishes system should be devised in each department negative publicity leading to a change in the reputations of the vast to assist officials in early identification of behavior, or greater circumspection when majority of LAPD officers violence-prone officers.”21 engaging in misconduct. However, the find- who do their increasingly Until recently, these systems received limited ing also may reflect regression to the mean. difficult job of policing acceptance, owing in part to concerns over This is a statistical phenomenon postulat- the City with courage, possible abuses. The abuses include use of ing that extreme scores gravitate toward skill, and judgment.” — inaccurate information, improper labeling the mean or average score, thereby becom- Independent Commission of officers, misuse of confidential records ing less extreme over time. on the Los Angeles Police regarding discipline and other personnel Department, Report of the For example, groups of police officers who matters, and social ostracism by peers and Independent Commission receive many citizen complaints, or who are community for officers identified as problem- on the Los Angeles Police disproportionately involved in the use of atic. There also were concerns about limited Department, Los Angeles, force, or who frequently are given poor resources and about increased legal liability CA: Independent Com- performance ratings, will tend to become for the organization and individual officers. mission on the Los Ange- “better” over time, in the sense of statisti- les Police Department, As Toch observes, the violence-prone officer cally looking more like the “average” 1991: 31. paradigm often is based on a variety of officers, even if nothing is done about these loosely articulated theories of violent behav- problems. Statistical regression represents ior.22 The theories include concepts such as a serious threat to the validity of early racial prejudice, poor self-control, and ego warning systems based on the assumption involvement. Furthermore, these theories that extreme patterns of behavior persist often overlook the possibility that greater- over extended periods of time. than-average use of force may be a product of situational or organizational characteristics.

9 Use of Force by Police

What We Do Not Know About involving use of force, and that only a hand- Police Use of Force ful of these complaints are sustained. The argument has appeal. We believe that The incidence of wrongful use of force the vast majority of police officers are profes- by police is unknown. Research is criti- sionals who respect the law and the public. If cally needed to determine reliably, use of force is uncommon, civilian complaints validly, and precisely how often trans- are infrequent, and civilian injuries are few, gressions of use-of-force powers occur. then excessive force by police must be rare. We do not know how often police use force in That conclusion may indeed be correct, but ways that can be adjudged as wrongful. For to the extent that it hinges on official police example, we do not know the incidence of statistics, it is open to serious challenge. excessive force, even though this is a very Current indicators of excessive force are all serious violation of public trust. We could critically flawed. The most widely available pull together data on excessive force using indicators are civilian complaints of exces- police disciplinary records and court docu- sive force and civil lawsuits alleging illegal ments, for example, but the picture would be use of force. Civilian complaints of excessive sketchy, piecemeal, and potentially deceiving. force are infrequent, and the number of sub- When it comes to less grave or less precise stantiated complaints is very low. These fig- transgressions, such as “improper,” “abusive,” ures are consistent with the argument that “illegitimate,” and “unnecessary” use of excessive force is sporadic. However, com- force, the state of knowledge is even more plaint mechanisms are subject to selection precarious. and reporting biases, and the operation of In discussing this issue, we will concen- complaint systems, which typically is man- trate on excessive force, because these aged by police, wields considerable influence transgressions are of utmost concern to the on whether people will come forward to public and because well-established profes- complain. sional and legal criteria are available to Civil lawsuits against police are exceedingly help us evaluate police behavior. Notwith- rare relative to the number of times that standing a generally agreed-upon terminol- police use force. Because the legal process is ogy, we should recognize that developing a highly selective in terms of which claims get count of excessive force that is beyond all litigated, lawsuits are a very unreliable mea- dispute is an unworkable task. This is so sure of illegal use of force. With both civilian because difficult judgments are involved in complaints and lawsuits, small changes in deciding whether use of force fits the crite- administrative practices can have a large ria for these categories in a given situation, impact on the magnitude of the problem and reasonable people will disagree in such measured in these ways. judgments. We clearly need more accurate, reliable, and valid measures of excessive The difficulties in measuring excessive and force if we are to advance our understand- illegal force with complaint and lawsuit ing of these problems. records have led academics and practitioners to redirect their attention to all use-of-force Academics and practitioners both tend to incidents. The focus then becomes one of presuppose that the incidence of excessive minimizing all instances of police use of force by police is very low. They argue that, force, without undue concern as to whether despite their shortcomings, agency statistics force was excessive. From this perspective, provide a useful picture of the use-of-force other records, such as use-of-force reports, problem. These statistics show that most arrest records, injury reports, and medical officers do not engage in force on a regular records, become relevant to measuring the basis, that few people are injured by police incidence of the problem. use of force, that only a small number of people complain about police misconduct

10 Chapter 1: What We Know About Police Use of Force

From a theoretical perspective, understand- organization is on the road toward disaster. ing all use-of-force incidents helps us to put Yet, we lack research that systematically wrongful use of force in perspective. How- addresses these questions. ever, because political, legal, and ethical Less formal aspects of police organizations— issues are very serious when we are dealing officer morale, administrative leadership, with excessive force, pressures to know the peer culture and influence, police-community incidence and prevalence of these events relations, relations with other government with precision will always be present. agencies, and neighborhood environments— As a corollary of our current inability to also plausibly have a part in levels of officer measure excessive force, we cannot discern misconduct. Alienated officers who do not with precision changes in the incidence of have a clear vision of their role and responsi- these events over time and across places. bilities and who are working in disorganized This means that we can neither determine agencies and interacting with the public whether excessive force problems are under stressful circumstances probably are getting better or worse nor determine the more likely to abuse their authority, includ- circumstances under which those problems ing their authority to use force. Research are more or less severe. that systematically addresses these questions is lacking. The impact of differences in police organizations, including administrative Methodological investigation of relations policies, hiring, training, discipline, between organizational elements and use-of- and use of technology, on excessive and force transgressions will help explain police illegal force is unknown. Research is misconduct at a theoretical level. More im- critically needed in this area. portantly, research on these questions will allow us to deal effectively with police misbe- A major gap in our knowledge about exces- havior. Faced with serious misconduct prob- sive force by police concerns characteristics lems in a police agency, we need to focus of police agencies that facilitate or impede scarce resources on those aspects of police this conduct. Although many of the condi- organizations that are most clearly related to tions that arguably lead to excessive or ille- ensuring proper conduct of officers with re- gal force by police seem obvious, or appear to gard to use of force. Generalized efforts to be a matter of common sense, we still greatly reform police organizations that are expected need systematic research in this area. We to reduce misconduct problems tend to be need to know, for example, which organiza- inefficiently focused and thus appear clumsy, tional characteristics are most consequen- inadequate, and misinformed. tial, which characteristics take on added significance in various environments, and Research must focus on establishing the which characteristics are redundant or relative cost-effectiveness of various strate- derivative of other characteristics. gies to reduce or eliminate police misconduct. Furthermore, only strategies that are solidly Many formal aspects of the organization— grounded in theory, practice, and empirical such as hiring criteria, recruit training, in- research will provide reliable solutions with service programs, supervision of field officers, predictable costs and benefits. disciplinary mechanisms, operations of internal affairs, specialized units dealing with eth- Influences of situational characteristics ics and integrity, labor unions, and civilian on police use of force and the transac- oversight mechanisms—plausibly are related tional nature of these events are largely to levels of officer misconduct. It makes sense unknown. More research is necessary. that poorly educated, badly trained, loosely Research on police-citizen encounters reveals supervised, and inadequately disciplined of- that use of force by police is situational and ficers are likely to be problematic, and that transactional. That is, police respond to when such officers are in the majority, the

11 Use of Force by Police

circumstances as they first encounter them or when there is a hostile audience to the and as they unfold over time. For example, encounter. At this point, however, knowledge Bayley and Garofalo observed that the situa- about the types of police-citizen encounters tions most likely to involve police use of force in which police are likely to use force is are interpersonal disturbance and violent rudimentary. personal crime.24 Beyond this, however, we do not know much about the types of events Police-public encounters are transactional in that enhance the likelihood that police will the sense that all the actors in a situation use force. contribute in some way to its development and outcome. Understanding the transac- Similarly, we have noted that when suspects tional nature of police use of force is impor- attempt to flee or physically resist arrest tant because it emphasizes the role of police police are more likely to use force. We also actions in increasing the chances that force noted that in many cases both police and will be used. suspects use force against each other. However, these findings do not address the From this perspective, it is possible to transactional nature of police-public encoun- minimize the use of force by modifying the ters in that they do not describe the step-by- behavior and tactics of police officers. By step unfolding of events and interactions. understanding the sequences of events that Knowing that police use force if suspects lead police to use force, we can gain a greater physically resist arrest, it matters if police degree of control over those situations and use force without provocation and the sus- possibly redirect the outcome. But we have pect responds by resisting or vice versa. only a basic understanding of the transactional nature of use-of-force situations, de- A variety of situational elements plausibly spite the fact that sequences of actions and are related to police use of force. If police are interactions are highly germane to determin- called to a scene where there is fighting, they ing whether use of force was excessive or may have to or believe they have to use force illegal. to subdue the suspects. If they are called to a domestic dispute where emotions are run- Organization of the Report ning high, they may have to or believe they have to use force to gain control of the situa- The next four chapters of this report focus tion. If they are called to intercede with a on major research studies dealing with po- civilian who is recklessly brandishing a lice use of force. They represent significant weapon, they may have to or believe they projects currently under way to understand have to use force to protect themselves and police use of force. others. Use of force in such circumstances Two of the projects are attempts at measur- may be justifiable, but to the extent that it is ing the incidence of police use of force na- predictable, we can prepare officers for these tionwide. BJS has developed a national-level encounters and devise alternative strategies data collection effort using a household sur- that minimize or eliminate the use of force. vey methodology to investigate police-public Some situational factors may increase the interactions, with a component on use-of- chances that force of questionable legitimacy force issues (see chapter 2). IACP is collect- will be used. For example, officers sometimes ing data on police use of force through a use force on the slightest provocation follow- voluntary reporting system (see chapter 3). ing a high-speed car chase, when adrenaline Two other projects on police use of force in- levels are high. They may use force more fre- volve citywide investigations across several quently when they are alone, because they locations. Chapter 4 reports on research in feel more vulnerable or believe that they can six jurisdictions; the research is important get away with it. They may use force more because it identifies factors associated with frequently as a way of emphasizing their use of force and because it addresses difficult authority when suspects are disrespectful

12 Chapter 1: What We Know About Police Use of Force measurement issues. Focusing on three po- 6. Adams, Kenneth, “Measuring the Preva- lice agencies, chapter 5 discusses research lence of Police Abuse of Force,” in And that centers on the use of force by both police Justice For All: A National Agenda for Un- and suspects; the research is important derstanding and Controlling Police Abuse of because it contributes significantly to under- Force, ed. William A. Geller and Hans Toch, standing the transactional nature of police- Washington, DC: Police Executive Research citizen encounters. Forum, 1995: 61–97. The final chapter outlines suggested direc- 7. Greenfeld, Lawrence A., Patrick A. tions for future research. A selected bibliog- Langan, and Steven K. Smith, Police Use of raphy concludes this report. Force: Collection of National Data, Washing- ton, DC: U.S. Department of Justice, Bureau Notes of Justice Statistics and National Institute of Justice, November 1997, NCJ 165040. 1. Bierce, Ambrose, The Devil’s Dictionary, New York: Dover, 1958: 101. 8. Ibid. 2. “Justice Department Consent Decree 9. Pate, Anthony M., and Lorie A. Fridell, Pushes Police to Overhaul Operations,” with Edwin E. Hamilton, Police Use of Force: Pittsburgh Post-Gazette, March 1, 1998, C–1. Official Reports, Citizen Complaints, and Le- gal Consequences, Vols. I and II, Washington, 3. Based on an investigation in 14 cities, DC: The Police Foundation, 1993. Human Rights Watch described the brutality situation as follows: “(p)olice officers engage 10. Greenfeld, Lawrence A., Patrick A. in unjustified shootings, severe beatings, Langan, and Steven K. Smith, Police Use of fatal chokings, and unnecessarily rough Force: Collection of National Data. physical treatment in cities throughout the 11. Garner, Joel, John Buchanan, Tom United States, while their police superiors, Schade, and John Hepburn, Understanding city officials and the Justice Department fail Use of Force By and Against the Police, to act decisively to restrain or penalize such Research in Brief, Washington, DC: U.S. acts or even to record the full magnitude Department of Justice, National Institute of the problem.” Human Rights Watch, of Justice, November 1996, NCJ 158614. Shielded from Justice: Police Brutality and Accountability in the United States, New 12. Bayley, David H., and James Garofalo, York: Human Rights Watch, 1998: 1, 27. “The Management of Violence by Police Patrol Officers”; and Bayley, David H., and 4. A previous summary of research on police James Garofalo, “Patrol Officer Effectiveness use of force can be found in McEwen, Tom, in Managing Conflict During Police-Citizen National Data Collection on Police Use of Encounters.” Force, Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics and 13. Ibid. National Institute of Justice, April 1996, 14. Worden, Robert, “The ‘Causes’ of Police NCJ 160113. Brutality,” in And Justice For All: A National 5. Bayley, David H., and James Garofalo, Agenda for Understanding and Controlling “The Management of Violence by Police Police Abuse of Force, 31–60. Patrol Officers,” Criminology, 27(1)(February 15. Garner, Joel, John Buchanan, Tom 1989): 1–27; and Bayley, David H., and Schade, and John Hepburn, Understanding James Garofalo, “Patrol Officer Effectiveness Use of Force By and Against the Police. in Managing Conflict During Police-Citizen Encounters,” in Report to the Governor, 16. Reiss, Albert J., Jr., Studies on Crime and Vol. III, Albany: New York State Commission Law Enforcement in a Major Metropolitan on Criminal Justice and the Use of Force, Area, President’s Commission on Law 1987: B1–88.

13 Use of Force by Police

Enforcement and Administration of Justice, 20. United States Commission on Civil Field Survey No. 3, Washington, DC: U.S. Rights, Who’s Guarding the Guardians? A Government Printing Office, 1967. Report on Police Practices, Washington, DC: United States Commission on Civil Rights, 17. Garner, Joel, John Buchanan, Tom 1981: 159. Schade, and John Hepburn, Understanding Use of Force By and Against the Police. 21. Ibid. 18. Toch, Hans, “The ‘Violence-Prone’ Police 22. Toch, Hans, “The ‘Violence-Prone’ Police Officer,” in And Justice For All: A National Officer,” 112. Agenda for Understanding and Controlling 23. Office of the Inspector General, Los An- Police Abuse of Force, 99–112. geles Police Commission, “Status Update: 19. Independent Commission on the Los An- Management of LAPD High-Risk Officers,” geles Police Department, Report of the Inde- Los Angeles: Los Angeles Police Commission, pendent Commission on the Los Angeles 1997. Police Department, Los Angeles, CA: Inde- 24. Bayley, David, H., and James Garofalo, pendent Commission on the Los Angeles “Patrol Officer Effectiveness in Managing Police Department, 1991. Conflict During Police-Citizen Encounters.”

14 2 Revising and Fielding the Police-Public Contact Survey by Lawrence A. Greenfeld, Patrick A. Langan, and Steven K. Smith

o learn more about police use of force supplements. The NCVS sample consists of T requires an understanding of the rea- all household members age 12 or older resid- sons for and the results of police-public ing in more than 40,000 U.S. households; encounters. As a step toward developing that each household member is interviewed twice understanding, the Bureau of Justice Statis- during a calendar year, resulting in about tics (BJS) fielded a pilot test in 1996 of the 200,000 interviews annually.) Police-Public Contact Survey (PPCS). Its pri- The PPCS pilot test involved interviews with mary purpose was to obtain information to 6,421 persons during the 1996 trial period. help guide future development of a final The respondents were asked about their con- Lawrence A. Greenfeld, questionnaire on the topic. This chapter tacts with police during the 12 months prior Patrick A. Langan, Ph.D., briefly reviews the 1996 survey, discusses to the interviews. Respondents interviewed and Steven K. Smith, improvements incorporated into a second in May 1996 were asked about contacts that Ph.D., are statisticians pilot survey, and describes other planned occurred anytime during the period June with the Bureau of Justice BJS efforts to learn more about police-public 1995 to May 1996; those interviewed in June Statistics, a component of encounters. 1996 were asked about contacts between the Office of Justice Pro- July 1995 and June 1996; and interviews in grams within the U.S. First Pilot Test of the Survey July 1996 covered the period from August Department of Justice. Questionnaire 1995 to July 1996. On average, the 12-month reference period included 6 months in 1995 BJS supplemented the National Crime Vic- and 6 in 1996. timization Survey (NCVS) with a pilot test of PPCS during May, June, and July 1996. The Administration of PPCS went smoothly. objective was to collect answers from respon- Among persons who had no contact with po- dents to a series of questions about the na- lice, the interview took 1 minute, on average, ture and consequences of their face-to-face to complete. Among those who had police interactions with police. contact, the interview averaged 10 minutes. (NCVS is based on interviews conducted Findings from the first PPCS were reported with a nationally representative sample of in 1997 in the BJS–National Institute of Jus- U.S. households and has become a highly tice publication Police Use of Force: Collec- useful platform for testing new question- tion of National Data.1 (See sidebar “Selected naires and periodically implementing special findings from the Police-Public Contact

15 Use of Force by Police

Survey.”) The 1996 PPCS was intended as a appropriateness or inappropriateness of pretest of the questionnaire; the survey and police conduct during the contact. its findings were not meant to be viewed as a source of indepth or precise statistics on Contact screen questions police use of force. Rather, survey findings The pretest revealed a number of broad provided empirical information to help categories of types of face-to-face contacts guide future development of an improved that needed encoding on the questionnaire. questionnaire. A number of respondents reporting contacts indicated that their employment brought Second Pilot Test of the Survey them into regular contact with police. This Questionnaire was so for hospital workers, tow-truck drivers, and lawyers, among others. In addition, During the first half of 1999, BJS focused on respondents indicated that they had contacts improving the content and administration of with police as the result of appearing in the PPCS questionnaire. Three areas of the court as jurors or witnesses. instrument were improved: items relating to screen questions used to identify whether a A major area for further development of the contact occurred, the type of contact, and screening elements of PPCS relates to traffic Screen questions are circumstances surrounding the contact; the stops. One of the major findings from the designed to determine determination of whether force was used or initial PPCS field test was that many of the quickly at the beginning threatened, the type of force used or threat- face-to-face contacts the public has with po- of an interview whether ened, the circumstances surrounding its use, lice are in connection with traffic stops. They a respondent needs to be and provocative actions by the respondent; may represent an area of potentially argu- asked subsequent ques- and the respondent’s perceptions of the mentative interactions between police and tions. For example, if a respondent had no police contact during the speci- Selected findings from the Police-Public Contact Survey fied period, subsequent questions about the na- In 1996 the Bureau of Justice Statistics • For nearly half of those with contacts, the ture, content, and conse- conducted a pilot test of the Police-Public encounters were initiated by the public. quences of police-public Contact Survey. Although not intended as For just under one-third of persons with encounters would be a source of detailed or precise statistics on contacts, police initiated them. police use of force, the pilot survey did pro- omitted. • Age is an important factor in both the vide preliminary estimates of the preva- frequency and type of police contacts lence of the public’s contact with police, experienced. Young people were the including contacts during which police least likely to initiate contact with police used force. Among the survey’s findings (their contacts most often were police are the following: initiated), while persons age 60 or older • An estimated 44.6 million persons (one in were the most likely to initiate contacts five U.S. residents age 12 or older) had with officers. face-to-face contacts with police officers • About 1 percent of people reporting con- during the prior 12 months. Men, whites, tacts with police indicated that officers and people in their twenties were the used force or threatened force. In the ma- most likely to have those contacts. jority of those instances, respondents said • An estimated 33 percent of residents who that their own actions, such as threatening had contact with police had either asked police or resisting arrest, may have pro- for assistance from officers or provided it voked officers. to them. About 32 percent of those who had contact with police had reported a crime, either as a victim or witness.

16 Chapter 2: Revising and Fielding the Police-Public Contact Survey the public and result in use-of-force inci- Determining whether force was used dents. This is likely to be true especially if The 1996 PPCS asked respondents to indi- such stops are thought to be motivated by cate whether force was threatened, used, or factors not strictly related to law enforce- not used during contacts. BJS replaced this ment purposes. Recent studies in New and substituted a list of police behaviors that Jersey and Maryland suggest that black might have occurred, including threatening motorists may be more likely than others to or restraining respondents during encoun- be stopped by police.2 ters. Handcuffing, considered a standard po- The PPCS questionnaire has been revised lice practice, is listed within the enumerated to capture more detailed information about group of police behaviors and will not be traffic stops, including more about the na- treated in the questionnaire separately from ture and frequency of traffic stops involving other behaviors that might have occurred persons of different races. To gather more during contacts. information on the nature and conse- Paralleling these queries about police behav- quences of traffic stops, BJS has added a ior during contacts are questions about the group of questions to the PPCS. Of particu- respondents’ behavior during encounters to lar interest will be an effort to determine learn about provocative actions that may from respondents what they believe to be have occurred. Such questions will apply to the basis for stops, the reason given by the all respondents, not merely to those indicat- officers for stops, and the content of the ing force had been used. interaction during stops, such as tickets issued, warnings, verbal questioning, Experience with a pat-down or frisking was searches, and arrests. Information gathered asked only of those respondents who indi- from the survey can be used to build on cated that in contacts with the police, they prior research regarding the treatment of believed that they were suspects in a crime. different categories of the populace by the This was too limiting, and BJS has expanded criminal justice system. the use of this item to all respondents. Those traffic stops resulting in handcuffing, Similarly, questions relating to subsequent threats, or use of force will also be identified. criminal charges resulting from contacts For such incidents, new items have been were limited to respondents who believed added to the survey to ascertain potentially that the contacts occurred because police provocative behaviors or drug or alcohol use considered them suspects in a crime. BJS by respondents. will ask such questions of all respondents. Several other modifications to the question- Respondents’ characterizations of incidents naire were needed. For example, existing questions about respondents’ traffic-related The 1996 PPCS asked respondents to contacts with police were modified to deter- characterize whether they believed police mine whether such respondents had been behavior was proper or improper during passengers or drivers. In addition, screen use-of-force incidents. Such a question will questions were added to determine explicitly be asked of all respondents who have police who initiated contact between respondents contact. and police officers. The 1996 questionnaire In addition, the 1996 PPCS did not ask those provided information only on the type of who experienced use of force whether they contact; who initiated it had to be inferred perceived police behavior to have been exces- (e.g., “Received a traffic or parking violation” sive for the circumstances; BJS has added implies the officer initiated the contact). such an item to the next version of the survey.

17 Use of Force by Police

Survey Administration for Second samples of those confined in local jails and Pilot Test prisoners held by State and Federal authorities. BJS anticipates introducing new survey BJS and the U.S. Bureau of the Census items that will gather information about began the 1999 pilot test using the revised respondents’ interactions with police during PPCS in July. The fieldwork will continue the arrest preceding their incarceration. through December 1999. This will result in The new items will provide, for the first time, a national sample of respondents about 10 information on the content of those contacts times the size of the first pilot test. Prior to between police and criminals, including the upcoming test, BJS redesigned the ques- the methods by which they were appre- tions as noted above, circulated them for hended, use of force during such events, and comment, and secured Office of Manage- provocative behaviors by known offenders. ment and Budget clearance for the use of the supplement for the full NCVS Notes national sample. 1. Greenfeld, Lawrence A., Patrick A. In addition, BJS and the Bureau of the Langan, and Steven K. Smith, Police Use of Census conducted extensive testing of the Force: Collection of National Data, Washing- revised instrument with volunteer respon- ton, DC: U.S. Department of Justice, Bureau dents in the Census Bureau’s Cognitive of Justice Statistics and National Institute of Research Laboratory. Justice, November 1997, NCJ 165040. Additional Efforts to Learn About 2. See Lamberth, John, “Driving While Black: A Statistician Proves That Prejudice Police-Public Encounters Still Rules the Road,” Washington Post, During the next 3 to 5 years, BJS will be August 16, 1998, C1. undertaking the regular periodic surveys conducted among nationally representative

18 3 IACP National Database Project on Police Use of Force by Mark A. Henriquez

ndividual police agencies, law enforcement Basic Concepts Underlying the Project Iorganizations, and some States have long considered collection of use-of-force data as a Essential to creating IACP’s police use-of-force means to better serve their communities and database was the belief that data contribu- to better distinguish between misperception tions should be voluntary and anonymous. and reality of police use-of-force issues. Re- Also of basic importance was that the defini- flecting this, Congress enacted legislation in tion of “police use of force” should reflect opera- 1994 that, among other things, directed the tional realities of modern, street-level law U.S. Attorney General to collect data on police enforcement. use of excessive force. Voluntary and anonymous reports Mark A. Henriquez is In 1995 the Bureau of Justice Statistics Project Coordinator for (BJS) and the National Institute of Justice Police use of force is an extremely sensitive the IACP National Police (NIJ)—components of the Office of Justice issue, in part because agencies and the pub- Use-of-Force Database Programs (OJP) within the U.S. Department lic alike harbor preconceptions and because Project. of Justice (DOJ)—proposed to cofund devel- data and resulting reports could be used or opment of a national use-of-force database interpreted either accurately or inaccurately. by the International Association of Chiefs of For example, the legislation directing the U.S. Police (IACP). One purpose of such a data- Attorney General to “acquire data about the base would be quantification of the extent use of excessive force by law enforcement of- and types of force used by police. ficers” also makes deprivation of civil rights Following approval of first-year funding, initial unlawful as evidenced by “pattern or practice” project activities began in September 1996. Early and allows the Attorney General, through in 1997, NIJ and BJS cofunded the project for civil action, to “obtain appropriate equitable another year. Since early 1998, IACP has been and declaratory relief to eliminate pattern or responsible for all project funding. practice.” Thus, if use-of-force data provided to IACP from the field—whether standing The balance of this chapter discusses basic alone or accompanied by analysis—were concepts underlying IACP’s use-of-force data- associated with individual departments, base project, highlights selected preliminary litigation could result. findings derived from the database, and discusses the future of the project. The chapter’s Under those circumstances, the law enforce- addendum provides details about key element community would have faced the ments involved in database development. dilemma of being very reluctant to provide

19 Use of Force by Police

use-of-force data, yet realizing that funda- through the software, the team identified a mental elements of modern police service are “street continuum” of force, which consists of pursuit of truth and subordination to the the types of force used by officers on a day- will of the communities that agencies are to-day basis to bring subjects under control. sworn to serve. Therefore, IACP decided that The IACP database consists of data on re- provision of such data would be both volun- ported use of force stemming from police tary and anonymous to promote accurate responses to calls for service, whether or not reporting and overcome potential reluctance those responses resulted in arrests.1 Catego- of agencies to participate. ries of force include the following:

Reflecting operational realities ● Physical force (use of fists, hands, feet, etc.). ● Chemical force (the discharge of Mace, When examined from the perspective of pepper spray, and similar agents). day-to-day law enforcement activities, many previous use-of-force definitions were consid- ● Electronic force (the discharge of Tasers, ered not sufficiently workable or functional stun guns, or other electronic weapons). to be applied nationwide to all jurisdictions ● Impact force (use of batons and the like). and department types for the purposes of the IACP project. Thus, IACP designed the ● Lethal force (firearm discharge of any project from the outset to reflect operational kind). realities of modern, street-level law enforce- Additionally, IACP software enables department, including the very meaning of “police ments to track various subcategories of force, use of force,” defined as the amount of force such as dog bites, edged weapons, vehicles, required by police to compel compliance by and nail guns. However, the project team an unwilling subject. excluded certain measures of force because The project team developed a unique soft- they were considered to fly in the face of re- ware package to track the basic types of ality and practicality. (See sidebar “Measures force used by officers and suspects in typical of force excluded from the database.”) encounters. On the basis of data submitted

Measures of force excluded from the database IACP’s project team decided to exclude • Routine or voluntary handcuffing of pris- from the database certain measures of oners for transport or during field ques- force that the team considered were too tioning or investigation. broad to allow agency reporting in an ac- • Display or presentation of an officer’s curate and timely fashion or beyond what weapon. police typically perceive or record as appli- cations of force: Although any of the above measures could be and are included in some academic stud- • Presence of a police officer at the ies of police use of force, IACP excluded scene. them from its database to allow creation of a • Presence of a K–9 at the scene. concise, universally accepted, and practi- cally achievable information base on police • Presence of chemical or electronic use of force in the United States. Inclusion of less-than-lethal devices at the scene. the elements listed above would have overly • Verbal commands by an officer. complicated the project and substantially reduced local agency participation.

20 Chapter 3: IACP National Database Project on Police Use of Force

Agencies contributing to the use-of-force database

IACP has obtained a substantial level of Although not nationally representative at use-of-force data from participating agen- this early stage of project development, the cies. Hundreds of others are establishing received data provide indicators regarding deadlines for submitting data either to their police use of force against subjects and State chiefs’ organizations or to IACP subject use of force against officers. As of directly: this writing, agencies representing populations ranging from 1,000 to more than 1 • About 4,000 agencies have requested million have contributed data on thousands the requisite software. of incidents. Reported use-of-force inci- • An estimated 1,000 agencies are using dents applicable to 1996 and 1997 total the software to capture use-of-force data. 24,383 and 24,033, respectively (based on data submitted through September 1998). • Some 150 agencies are expected to contribute data for the 1998–99 data year.

Selected Preliminary Use-of-Force and 27 for 1997 on 2,310 police use-of-force Findings incidents indicate that the vast majority of such incidents occurred in arrest-related Although the database is in the early stages situations. For the 1996–97 period, those of development, the level of interest and agencies also reported the following: support from local police agencies is encour- ● Of 2,264 use-of-force confrontations for aging. (See sidebar “Agencies contributing to which the race of officers and subjects was the use-of-force database.”) Nonetheless, the known, 909 were intraracial (officer and data are not yet nationally representative. subject of the same race) and 1,335 were Conclusions about national use-of-force interracial. trends, therefore, should not be attempted For a more detailed at this time. ● About 10 percent of 2,479 officers using account of use-of-force force sustained injuries. Less than 1 per- But preliminary findings based on the data project findings, see Inter- cent of the injuries were major; none re- may provide useful insights, however tenta- national Association of sulted in death. About 38 percent of the tive, into current police and subject use-of- Chiefs of Police, Police subjects were injured as the result of po- force issues. The IACP calculation of the Use of Force in America: lice use of force, including approximately use-of-force rate is based on dispatched calls Research in Progress 1.5 percent with major injuries. (Data for service. For example, based on 1995 data Report on the IACP Na- spanning the 1995–97 period indicate reported by 110 agencies, the police use-of- tional Database Project, that of 75,082 use-of-force incidents, force rate was 4.19 per 10,000 responded-to Alexandria, Virginia: 3,274, or about 4 percent, resulted in calls for service, or 0.0419 percent. Jurisdic- International Association officer injuries, all but 39 minor.) tion size for 78 of those agencies was 35,000 of Chiefs of Police, April population or less; for 4 departments, ● Of 3,972 reported incidents involving use 1998. 500,000 population or more. of force, 20 resulted in complaints by subjects. Data reported for 1995–97 indicate that of 62,411 use-of-force incidents during the period, about 87 percent involved officers using Future of the Database physical force.2 Officers used chemical force From the inception of the database project in 7 percent of the incidents, firearms in in 1996, it has had—and is expected to about 5 percent. continue to have—a three-tiered impact on Data received by IACP through September police use-of-force policies and practices in 1998 from 26 agencies reporting for 1996 the United States. At the national level, the

21 Use of Force by Police

IACP annual report Police Use of Force in 7 pilot States, database software is now in America provides summary and incident in- place in more than 1,000 agencies nationwide formation to police leaders, the public, and and in the last quarter of 1998, Illinois and the media. At the State level, State associa- Rhode Island joined the project, followed by tions of chiefs of police increasingly use the Maryland and Missouri in early 1999. Newly project to provide leadership to their con- recruited States may contribute data from as stituent law enforcement agencies. Finally, far back as 1991. IACP believes that the ben- local police agencies use IACP-provided soft- efits of its use-of-force database project will ware to capture use-of-force information per- continue to increase as the number of contrib- taining to their own departments and use it uting departments approaches national for such purposes as the following: representation.

● To promote improved policies, training, and procedures governing departmentwide Notes use of force. 1. Data submitted to IACP by any given ● To reduce potential department liability agency may also reflect use-of-force inci- arising from frivolous or unfounded legal dents resulting from responses to other actions related to use of force. than calls for service—for example, from officer-initiated contacts with the public. ● To contribute data to statewide use-of- force data repositories, where applicable, 2. The data for 1996 and 1997 reflect agency and to the IACP database. use-of-force reports received through Sep- tember 1998 and, therefore, are not yet com- The IACP database is a dynamic one. The plete. Although the IACP database project number of departments providing new or up- started in 1996, some agencies supplied data dated data increases almost on a daily basis. for 1995. For example, from a modest beginning of

Addendum: Key Elements in Building der Patrol (Immigration and Naturalization the IACP Use-of-Force Database Service). This group, from the outset, has provided advice on local concerns, State as- To ensure the success of the database project, sociation perspectives, and logistical issues. IACP created two advisory panels to support This group also provides midyear and end- initial project activities. The first was an ad year input on project direction and selected hoc committee of police and justice leaders milestones. brought together at IACP headquarters to react to the project’s scope and give general The SACOP role policy input and advice. This group consisted The State Associations of Chiefs of Police of DOJ representatives, State police, county (SACOP) Division of IACP laid the founda- sheriffs, local law enforcement officers and tion for this program from its inception. In other criminal justice professionals. Consensus particular, the Virginia Association of Chiefs on key issues, such as definitions and data of Police had already instituted a statewide elements, was achieved during the project’s use-of-force reporting program that served start-up phase. as the model for IACP’s national effort. The A second and continuing advisory committee strength and reach of SACOP organizations consists of representatives from each of within most of the States provided a flexible seven pilot State associations, discussed be- framework for developing and expanding the low, and a representative from the U.S. Bor- project.

22 Chapter 3: IACP National Database Project on Police Use of Force

In collaboration with SACOP leadership, content”), it soon became clear that an auto- IACP initially identified seven pilot State mated data collection system was required. associations that would help coordinate the The ideal system would need to: contribution of information for the national ● Be compatible across the widest possible database effort: Arkansas, New Jersey, New spectrum of computers used in depart- York, Vermont, Virginia, Washington, and ments nationwide. West Virginia. DOJ also arranged early Fed- eral support by designating the U.S. Border ● Employ a graphical user interface that Patrol to contribute data. would facilitate self-instruction by the user. Each pilot State association identified five local police departments to contribute data to ● Handle the relationship between inci- the project. To accomplish the data collection dents involving a single officer and mul- and transfer, IACP developed a new software tiple subjects, multiple officers and a application specifically for this purpose. single subject, and any other possible combination. A three-tiered approach to data collection Staff decided on a system that could meet Data flow was designed from the outset to these needs and was scalable in that it could emanate from local departments, through support the inevitable evolution of revisions, the respective SACOP offices, and then to customizations, additions, and expansions IACP. The identities of all participating inherent in software projects of this type. departments are masked by use of a self- Because data flowed from local agencies to installing agency reporting code, which is SACOP regional data repositories and from automatically appended to all local agency there to IACP, two versions of the software data records. This code is the only means by were produced. The first, the local agency ver- which individual data records can be distin- sion, was intended to automate data capture guished from one contributing source to an- at the local department level. This software other. The SACOP organizations have the was equipped with a simplified data export capability to generate regional use-of-force function by which the user could write the data based on the anonymous records captured use-of-force data to a floppy disk by reported to them by departments in their simply clicking an on-screen button. States. The second version of the software was in- Larger departments, which maintain elec- tended for use by the SACOP data reposito- tronic repositories of their own use-of-force ries. The SACOP version of the software data, have the option of reporting their data remains identical to the local agency version (in any data format) directly to IACP in cases in terms of the number and type of data ele- where the appropriate SACOP office is not ments captured. However, the SACOP ver- participating in the project. Even in these sion is provided with an import capability by instances, the identity of the contributing which data on disks originating from local agency is protected by an agency reporting agencies can easily be incorporated into the code in the same manner that the identity of regional database. Data from each of the re- a department is protected when contributing gional databases are periodically exported to data through its State organization. floppy disk and sent to IACP for inclusion in Software design and training its database. Several software improvements have been made. Based on the large number of data elements relevant to this study (see sidebar “Database

23 Use of Force by Police

Training for IACP database administrators derlying the project. At the conclusion of the was provided in a formal class setting by training, students, all of whom were sworn IACP in August 1997. The 8-hour class pro- law enforcement personnel from participat- vided hands-on technical training in setting ing local police departments, were presented up, operating, and troubleshooting the use- with certificates identifying them as IACP of-force software and in the philosophy un- use-of-force database administrators.

Database content

Guided by the broad framework of use-of- • Number of third parties involved. force issues and by IACP’s experience with • Age of officer(s)/subject(s). other sensitive police policy issues, the project team decided upon the following • Type of assignment. elements for inclusion in the use-of-force • Duty status. database: • Education of officer(s)/subject(s). Department characteristics: • Race/ethnicity of officer(s)/subject(s). • Report year. • Type of force used by officer(s)/subject(s). • Jurisdiction size. • Officer/subject injury. • Department type. Complaint characteristics: • Calls for service. • Complaint time. • Ethnicity demographics. • Complaint date. • Number of use-of-force incidents by type. • Complaint year. • Number of force-related complaints. • Number of officers involved. • Complaint resolution. • Number of subjects involved. • Types of less-than-lethal weapons authorized. • Number of third parties involved. • Use-of-force training and policies in place. • Age of officer(s)/subject(s). • Administrative policies for use-of-force • Type of assignment, duty status. complaint. • Education of officer(s)/subject(s). Incident characteristics: • Race/ethnicity of officer(s)/subject(s). • Incident time. • Type of force used by officer(s)/subject(s). • Incident date. • Officer/subject injury. • Incident year. • Previous complaints against officer. • Number of officers involved. • Complaint disposition. • Number of subjects involved.

24 4 Measuring the Amount of Force Used By and Against the Police in Six Jurisdictions by Joel H. Garner and Christopher D. Maxwell

his study examines the amount of force forms, and constructing the measures of T used by and against law enforcement force reported here. The departments en- officers and more than 50 characteristics couraged officers to complete police survey of officers, civilians, and arrest situations forms and assisted the research team in associated with the use of more or less force. gaining access to suspects in local detention Data were gathered about suspects’ and centers. police officers’ behaviors from more than Throughout the project, these law enforce- 7,500 adult custody arrests1 in 6 urban law ment managers worked together as a group enforcement agencies. Joel H. Garner, Ph.D., is to promote the use of consistent measures Research Director, Joint The participating agencies were the across the six jurisdictions and to enhance Centers for Justice Charlotte-Mecklenburg (North Carolina) the use of the findings by the participating Studies, Inc. Christopher Police Department, Colorado Springs departments. This collaborative design was D. Maxwell, Ph.D., is (Colorado) Police Department, Dallas intended; in fact, it is not clear whether Assistant Professor of (Texas) Police Department, St. Petersburg there was any other way to produce the Criminology, Michigan (Florida) Police Department, San Diego information included in this chapter. State University. (California) Police Department, and San Emerging from this research is a more com- Diego County (California) Sheriff’s plete understanding of the frequency with Department. which certain types of tactics are used and To organize, present, and understand the what types of weapons are displayed, threat- nature and characteristics of the force used ened, or actually used. The consistent find- in representative samples of arrests, this ings across all six jurisdictions are that most study developed four measures of the arrests do not involve any force by the police amount of force used by police officers and or by suspects. In those situations where four parallel measures of the force used by some type of force is used, typically no arrested suspects. weapon is used, threatened, or even displayed. When police use some form of weap- The participating agencies also played a ma- onless tactic (hitting, kicking, wrestling, etc.), jor role in designing and implementing this the most frequent tactic involves grabbing research. Senior police managers in each de- only. partment actively contributed to delineating research goals, designing data collection

25 Use of Force by Police

The Importance of Systematic tive of each department’s annual arrests. We Samples estimated that the number of arrests needed to obtain reliable estimates of the amount of Prior research on the use of force has, in force varied between 900 and 1,200 in the many instances, been limited to instances six jurisdictions. However, we did not draw where some type of force, usually deadly a random sample of arrests throughout the force, was used.2 This use of samples that do year because that would have entailed com- not represent all police behavior limits our plicated procedures for starting and stopping ability to describe when force is used and data collection by police officers. We chose to when it is not used. sample arrests continuously over a 2- to 7-week period, depending on the size of the One senior researcher, Albert J. Reiss, Jr., department and the rate at which officers has argued that the absence of systematic made arrests. samples of police behavior means that the entire approach to the study of deadly force Data collection began at different times in was flawed “because analysts of police use of different departments, so the total sample deadly force focus on situations in which the included arrests during the summer, fall, decision was made to use it, such as firing a and the winter of 1996–97. We began data weapon . . . they ignore all decisions where collection in the Colorado Springs Police De- force gave way to alternative ways of coping partment in mid-August 1996 and completed with situations.”3 data collection in Charlotte-Mecklenburg in the second week of February 1997. It took The flaws that Dr. Reiss identified can have 50 days to obtain data on 1,249 arrests in real-world, and sometimes disastrous, conse- Colorado Springs but only 2 weeks to obtain quences. For instance, the congressional data on 1,192 arrests in Dallas. investigation into the Challenger shuttle disaster identified the failure of the engineers In all 6 jurisdictions, we obtained 7,512 us- at NASA to examine evidence from a sys- able surveys. The proportion of adult custody tematic sample of all of the previous 21 arrests for which we obtained completed, shuttle flights.4 The engineers examined usable surveys varied from 85 percent to 93 only the eight Shuttle flights with identified percent, and we determined that there were 0-ring problems, and among those flights, no systematic differences between arrests there was no correlation between 0-ring with or without a completed survey. The failures and temperature at the launch pad. large size and representative nature of our However, in the complete set of all 21 flights, sample provide a solid basis for describing the ultimately deadly correlation was clear. the nature of use of force in the six participating departments. This research employed systematic samples The authors’ research was of adult custody arrests in order to provide This research also sought to avoid other supported under grant a comparison of circumstances when force is methodological flaws. We collected informa- number 95–IJ–CX–0066 used with the circumstances when force is tion both from police officers and from by the National Institute not used. This is an important characteristic. arrested suspects and did so in ways that of Justice. By using a complete set of police behavior, protected the confidentiality of officers and even when no force is involved, we have suspects (see sidebar “Measuring force”). In sought to avoid types of errors that occurred assessing the factors associated with the use in the Challenger study as well as most of of force, this study used multivariate statisti- the other police studies. cal tests appropriate for the measurements available and the hypotheses being tested. Our approach to data collection compro- Simple descriptive information alone can be mised ideal research procedures with the misleading, especially when research seeks to real-life necessities of working within an associate the use of force with characteristics operational agency. We wanted to obtain a of officers, suspects, and arrest situations. sample of adult custody arrests representa-

26 Chapter 4: Measuring the Amount of Force Used

Our design has some limitations. There are weapon, or threatened to use a weapon. some circumstances in which the police use However, it is not always clear when a force but do not make an arrest, and our ap- weapon is actually used. Does a firearm have proach will miss those incidents. In addition, to be discharged to be “used”? In addition, is the primary source of our data is self- a firearm or a baton displayed just by being reported marks by police officers on a two- carried by an officer, or does display mean page form. These types of measures provide only removed from a holster? If a weapon is some uniformity but often miss important displayed, is an officer or a suspect threaten- distinctions that can be obtained by more ing to use it? We think the distinctions be- indepth interviews or firsthand observations tween use, threatened use, and display are by independent observers. important but acknowledge that the survey form might be an imprecise method to mea- For instance, we asked officers to record sure these differences. whether they used a weapon, displayed a

Measuring force The primary means by which the research age procedures of the Joint Centers for project collected information on the use of Justice Studies. force was a one-page, front and back, form Under the legislation authorizing the re- completed by law enforcement officers on a search program at the National Institute of systematic sample of adult custody arrests. Justice, Congress made confidential re- This form, illustrated on the two following search data “immune from legal process” pages, was used to record characteristics of and specified that data identifiable to an the arrest situation, the suspect, the officer, individual shall not be “used for any pur- and the specific behavioral acts of officers, pose in any judicial, legislative, or admin- suspects, and bystanders in a particular istrative proceeding.”† This confidentiality arrest. The form was derived from a similar protection was communicated to officers study* conducted in Phoenix, Arizona, dur- by their departments and included on the ing 1994 but was modified to conform to the survey form; interviewed suspects were local characteristics, police terminology, and told of this protection prior to their agree- departmental policies of the participating ing to participate in the research. agencies. Similar items were asked of a smaller sample of suspects interviewed in These procedures increased the likeli- local jails shortly after arrest. hood that officers and suspects would provide more accurate information be- The forms were completed by arresting cause officers would be less constrained officers but were not reviewed or controlled by fears that their individual answers by police managers; thus, the forms were might be communicated to others within not departmental records but research the department and might even possibly data. Both the police officers completing be used against them. The confidentiality the forms and the suspects who were inter- provided to research subjects by the Con- viewed were research subjects, and the gress makes the findings of this research confidentiality of their responses was pro- more reliable and, therefore, more useful tected by the data management and stor- to congressional and other policymakers.

* Garner, Joel, John Buchanan, Tom Schade, and John Hepburn, Understanding the Use of Force By and Against the Police, Research in Brief, Washington DC: U.S. Department of Justice, National Institute of Justice, November 1996, NCJ 158614. † 42 U.S.C. 3789(g).

27 Joint Centers for Justice Studies, Inc. Multisite Arrest Tactics Study INSTRUCTIONS: Complete one form for each adult arrest. This form is to be completed by the arresting officer. This officer will respond as first officer. When additional officers are involved, please record any actions on their part as well. PLEASE MARK EACH ITEM WITH A CHECK ( ✓) OR A NUMBER AS APPROPRIATE ARREST/OFFENSE # ______8. Number of Persons, Including Yourself, Present at Arrest Scene 1. Division/Assignments (✓ All That Apply) Number Present Initial Contact Completion of Arrest C NE NW NC SE SW INV Unit SO ICP # of Officers Investigative Administrative # of Suspects 2. Suspect’s Custody Status Upon Your Arrival (✓) Not Already in Custody # of Bystanders ✓ Already in Custody: Police, Courts 9. Suspect’s Relationships ( One for Each) Already in Custody: Private Security/Citizen Relationship to Victim Relationship to Bystanders 3. Officer’s Prior Knowledge of Location (✓) Unknown No Prior Knowledge No Relationship If Prior Knowledge, What Known YN Acquaintance/Friend Location Believed to be Nonthreatening Family/Intimate Location Known for Criminal Activity 10. Characteristics of Officers Location Believed to be Hazardous to Police 1st Officer 2nd Officer 4. Officer’s Prior Knowledge of Suspect (✓) Age years years No Prior Knowledge Height ft. in. ft. in. If Some Prior Knowledge, What Known? YN Weight lbs. lbs. Affiliated Gang Member Race White Black Hispanic Oth. White Black Hispanic Oth. Confirmed Gang Member Sex Male Female Male Female Believed to Carry Weapons Believed to Have a Criminal Record 11. On-the-Job Medical Attention Before Today Y N Believed to be Cooperative First Aid at Scene Believed to be Assaultive Treated at Hospital 5. Suspect’s Impairment Yes No Unknown Admitted to Hospital ✓ Drugs Y N 12. Type of Approach 13. Part of Shift ( ) Alcohol Routine Approach Time Shift am Began pm Other Backup Requested Y N 6. Location of Completed Arrest Y N Priority Call Other Duty Inside Outside Used Lights and Sirens Off Duty Suspect’s Residence Street Y N 14. Initial Contact with Suspect Y N Other Residence Parking Lot Dispatched On-View Club/Bar Suspect’s Yard Priority Code Initiated by Citizen Restaurant Other Yard Hazard Code Initiated by Officer Retail Store Other Outside 15. General Demeanor (✓) Type of Demeanor Civil Antagonistic 7. Visibility at Arrest Completion (Circle Number) Suspect’s Demeanor Toward Police Excellent Good Moderate Poor Bystander Demeanor Toward Police 10 9 8 7 6 5 4 3 2 1 Police Demeanor Toward Suspect Specific Actions by Officers and by Suspect 16. Suspect’s General Response to Police Y N Suspect Officers Immediate Compliance with Officer’s Requests 19. Type of Flight or Pursuit Disrespectful or Obscene Gesture No Pursuit/Flight Threatening Stance YN If Flight or Pursuit, What Type YN Verbal Resistance On Foot/Bicycle Passive Resistance (go limp, etc.) In Motor Vehicle Evade, Hide or Flee From Police In Helicopter Impede Officer’s Movements Suspect Officers Resist Cuffing Y N 20. Weaponless Tactics Y N Resist Placement in Police Vehicle Compliant Gentle Hold Only Assaultive Toward Police Spit Used or Tried to Use Deadly Force Against Police Grab Arm Suspect Officers Twist Arm Y N 17. Words Between Officer & Suspect Y N Push, Shove Conversational Voice Wrestle, Scuffle Command Voice Hit or Punch Shouting/Cursing Kick Verbal Threats Bite, Scratch Pressure Point 18. Type of Restraints Used Y N Carotid Hold/Lat. Vascular Restraint Hand Cuffs Control Hold (Specify)______Leg Cuffs Other (Specify) ______Other More Severe Restraints

21. Weapon Possession, Threatened Use, Display or Actual Use For Questions About Weapons, a Blank means NO SUSPECT POLICE Verbal Display/ Weapons Verbal Display/ Possession Threat Brandish Use Possession Threat Brandish Use Stick/Blunt Object Baton Knife/Edged Weapon Flashlight Handgun Chemical Agent Rifle/Shotgun Motor Vehicle Canine Other Item (specify ______) If weapon used, describe how weapon used: Suspect Officers Suspect Officers Y N 22. Injuries During This Arrest Y N Y N 23. Medical Attention This Arrest Y N Complaint of Pain/Strained Muscle, etc. Offered and Refused Temporary Chemical Irritation First Aid at Scene Bruise, Abrasion, Scratch, Burn Transported to Hospital Puncture, Cut Other (specify ______) Gunshot, Knife Wound Thank you for your time and effort. All information on this form identifiable to an individual will be kept confidential by Internal Injuries the Joint Centers for Justice Studies in accordance with Concussion/Loss of Consciousness Federal law (42 U.S.C. ¤3789(g)) which states that these research data are "immune from legal process" and shall not Broken Bone or Teeth be "used for any purpose in any action, suit, or other judicial, Other Injury (specify ______) legislative, or administrative proceeding." Use of Force by Police

The design of this research—systematic what elements of the arrest or the police- samples, multiple sources of information, public encounter constituted force, abuse of and multivariate analysis—was guided by force, or excessive force. Our approach has an assessment that much of the prior re- been to explicitly define and measure force search had confounded the measurement of and, building on prior research in Phoenix, force with definitions of what is and is not Arizona,5 we identified five elements of force: excessive force. In this project, we deferred weapons, weaponless tactics, restraints, the difficult task of defining and measuring motion, and voice. excessive force. We focused on the measurement of the amount of force, with the expec- Weapons tation that this information would inform There is a strong consensus that the use of issues surrounding the use of excessive force. a weapon constitutes force and that the use For instance, excessive force is typically but of certain types of weapons—e.g., handguns not necessarily associated with more severe and rifles—involves more force than the use forms of force that could or do result in in- of such other weapons as batons and oleo- jury or death. resin capsicum (pepper spray). What is less Our findings are that most arrests involve no clear is the meaning of “use.” For instance, force, excessive or otherwise. When force is does a firearm have to be discharged to be used, it typically involves less severe forms of “used”? Also unclear is whether the posses- tactics and weapon use. These findings pro- sion, threatened use, or display of a weapon vide a context for understanding excessive constitutes force by law enforcement officers force, which we know can involve low-level or by suspects. Our approach to this uncer- acts of force (such as verbal threats or cursing tainty was to have officers mark on the data against compliant suspects) as well as the collection form whether they or the suspects acts of force that result in physical injury or possessed, displayed, threatened to use, or death of civilians. Arrests that involve no used any of seven different types of weapons. force, however, cannot involve excessive force Combining data from all sites, exhibits 4–1 and arrests that involve low levels of force are and 4–2 show the frequency with which less likely to involve excessive force. Although officers report that they or arrested suspects the exact relationship between the amount of used, threatened to use, or displayed certain force and excessive force remains to be clari- weapons. Use of weapons is infrequent; in fied, this research seeks to inform future law 97.9 percent of all adult custody arrests, enforcement policies, practices, and training police did not use a weapon (exhibit 4–1). In by identifying what kinds of force are and are 99.3 percent of all such arrests, suspects did not currently being used by and against law not use a weapon (exhibit 4–2). As noted in enforcement officers. exhibit 4–1, the most frequent weapon used by the police was some form of a chemical The Elements of Force agent, mostly oleoresin capsicum;6 it was This chapter emphasizes measuring the used in 88 or 1.2 percent of the arrests in amount of force used by law enforcement this study. officers and by suspects. The task of measur- The second most frequent weapon was the ing the amount of force required establishing flashlight, used in 41 (0.5 percent) arrests. an understanding of the specific behavioral Handguns were used by the police in 11 (0.1 acts that constitute “force” and how much percent) arrests; rifles or shotguns were used force is involved in each of those behaviors. by the police in 7 (0.1 percent) of the arrests Prior research had traditionally employed (no officer reported that such use involved simple dichotomies between the presence or discharge of a firearm). absence of physical force or abuse of force or excessive force without much attention to The most frequent weapon used by suspects was a knife; it was used in 18 (0.2 percent) of

30 Chapter 4: Measuring the Amount of Force Used all arrests (exhibit 4–2). Suspects used hand- differences in incidents in which weapons guns in 12 (0.2 percent) arrests and rifles or are used and those in which they are dis- shotguns in 5 (0.1 percent) arrests.7 played but not used. A somewhat different pattern emerges in Exhibit 4–1 also reveals a finding similar exhibit 4–1 where we examine instances in to one reported in the Phoenix study: police which weapons were either displayed or officers report that they use, display, and used. Handguns were displayed or used by threaten to use a flashlight more often than police in 202 (2.7 percent) arrests and rifles they use, display, or threaten to use a baton. or shotguns were used or displayed 31 (0.4 Batons were used in 15 (0.2 percent) arrests; percent) times. Thus, firearms are infre- flashlights in 41 (0.5 percent). In addition, quently used but are the most frequent officers report that they used motor vehicles weapon displayed. On the other hand, chemi- as weapons in 15 arrests and either used or cal agents were the most frequently used threatened to use them in 21 arrests. These weapon but ranked second (118 or 1.6 per- findings do not conform to conventional cent) when use and display are counted. Ex- thinking about the relative frequency of hibit 4–1 indicates that in 215 (2.8 percent) weapon use or even on the types of equip- of 7,512 arrests, officers went so far as to dis- ment used as weapons. play a firearm but did not, ultimately, use a Officers report that suspects use, display, firearm. These findings suggest important and threaten to use weapons less frequently

Exhibit 4–1: Weapons threatened, displayed, or used by police in 7,512 arrests

Used Displayed or Threatened, Used Displayed, or Used Arrests Percent Arrests Percent Arrests Percent of Arrests of Arrests of Arrests No Weapon Involved 7,354 97.9 7,151 95.2 7,130 94.9 Weapon Involved 158 2.1 361 4.8 382 5.1 All Arrests 7,512 100.0 7,512 100.0 7,512 100.0 Type of Weapon* None 7,354 97.9 7,151 95.2 7,130 94.9 Baton 15 0.2 39 0.5 43 0.6 Flashlight 41 0.5 72 1.0 73 1.0 Handgun 11 0.1 202 2.7 204 2.7 Chemical Agent 88 1.2 118 1.6 130 1.7 Rifle/Shotgun 7 0.1 31 0.4 32 0.4 Motor Vehicle 15 0.2 21 0.3 21 0.3 Canine 20 0.3 31 0.4 37 0.5 Other 19 0.3 19 0.3 19 0.3

* Since some arrests involved the use, display, or threatened use of more than one weapon, the percentages under type of weapon do not add to 100 percent.

31 Use of Force by Police

than officers do. Exhibit 4–2 notes that the arm. Each of these tactics involves direct knives were the weapons most frequently physical contact between the officer and the used (18 or 0.2 percent) by suspects, followed suspect and does not involve use of specific closely by sticks (17 or 0.2 percent) and objects in applying force. The police survey motor vehicles (14 or 0.2 percent). When form listed 12 tactics and exhibit 4–3 dis- handguns, rifles, and shotguns are combined, plays the frequency with which officers re- the weapon most frequently displayed or ported the use of these tactics. In 6,328 or used (43 or 0.6 percent) by suspects is a fire- 84.2 percent of the arrests in this study, the arm. Suspects threatened, displayed, or used police reported that they used no weaponless a firearm in 63 arrests. These findings about tactics. Among arrests involving a weapon- suspects confirm our earlier findings about less tactic, the most frequent “tactic” was the police: understanding the use of force is grabbing, used 954 times (12.7 percent of all advanced by considering not only the use but arrests). Other common tactics involved us- also the display and the threatened use of a ing a control hold (164 times or 2.2 percent weapon. of arrests), arm twisting (281 or 3.7 percent), wrestling (233 or 3.1 percent) and pushing Weaponless tactics or shoving (145 or 1.9 percent). Police officers use and are trained to use a Police reported that suspects used weapon- variety of weaponless tactics, from carotid less tactics in 412 or 5.5 percent of all ar- control holds to simply grabbing a suspect by rests (exhibit 4–4). Wrestling was the tactic

Exhibit 4–2: Weapons threatened, displayed, or used by suspects in 7,512 arrests

Used Displayed or Used Threatened, Displayed, or Used Arrests Percent Arrests Percent Arrests Percent of Arrests of Arrests of Arrests No Weapon Involved 7,460 99.3 7,411 98.7 7,367 98.1 Weapon Involved 52 0.7 101 1.3 145 1.9 All Arrests 7,512 100.0 7,512 100.0 7,512 100.0 Type of Weapon* None 7,460 99.3 7,411 98.7 7,367 98.1 Stick 17 0.2 31 0.4 56 0.7 Knife 18 0.2 34 0.5 42 0.6 Handgun 12 0.2 29 0.4 47 0.6 Chemical Agent 5 0.1 5 0.1 7 0.1 Rifle/Shotgun 5 0.1 14 0.2 16 0.2 Motor Vehicle 14 0.2 14 0.2 15 0.2 Canine 4 0.1 4 0.1 6 0.1 Other 9 0.1 14 0.2 17 0.2

* Since some arrests involved the use, display, or threatened use of more than one weapon, the percentages under type of weapon do not add to 100 percent.

32 Chapter 4: Measuring the Amount of Force Used most often used by suspects (262 times or handcuffs, leg cuffs, and more severe re- 3.5 percent of all arrests), followed closely by straints, such as hobbles or body cuffs. As pushing or shoving by the suspect (166 times with most other items, the survey provided or 2.2 percent). As with weapons, the use of for a specific “yes” or “no” response for each weaponless tactics by officers and by sus- item. In the 7,512 arrests in this study, pects was infrequent. officers reported that they used handcuffs in 6,182 (82.3 percent) (exhibit 4–5). In 67 (0.9 Although weapons and weaponless tactics percent) arrests, the police used leg cuffs and are typically included in most understanding in 29 (0.4 percent) they used more severe of what constitutes the use of force, the three restraints.9 other elements discussed next are sometimes, but not always, considered part of the The use of restraints appears to be frequent use of force. but not universal; handcuffs predominate, but in a small proportion of arrests (1.3 per- Restraints cent) restraints more severe than handcuffs were used. Handcuffing alone is not typically One element of force that officers alone em- perceived as involving force, but our under- ploy is use of restraints.8 The police survey standing of force might include the use of form lists three possible types of restraints— more severe restraints, some of which have

Exhibit 4–3: Officer use of weaponless tactics in 7,512 arrests

Arrests Percent of Arrests No Tactics Used 6,328 84.2 At Least One Tactic Used 1,184 15.8 All Arrests 7,512 100.0 Type of Tactic* Spit 32 0.4 Grab 954 12.7 Twist Arm 281 3.7 Wrestle 233 3.1 Push/Shove 145 1.9 Hit 30 0.4 Kick 14 0.2 Bite/Scratch 11 0.1 Pressure Hold 83 1.1 Carotid Hold 31 0.4 Control Hold 164 2.2 Other Tactic 70 0.9 Number of Tactics 2,048

* Since some arrests involved the use of more than one tactic, the percentages under type of tactic do not add to 100 percent.

33 Use of Force by Police

been associated with injury to suspects and, percent). In a small number of arrests, sus- in some instances, even death.10 pects fled on foot (354 or 4.7 percent) or in a motor vehicle (128 or 1.7 percent). When the Motion police did pursue a suspect, it was most often (224 or 3.0 percent) on foot; motorized pur- One aspect of police-public encounters in suits, including helicopter pursuits, occurred arrest situations is suspect flight and officer in 199 (2.7 percent) arrests in our sample. pursuit. Although most research and policy discussions on the use of force do not address Flight and pursuit do occur, but it is not issues around either flight or pursuit, we clear the extent to which these actions in- include both as potential elements of force. volve the application of what we typically Our police survey form recorded whether mean when we speak of the use of force by suspects attempted to flee and, if so, whether police or by suspects. Still, flight and pursuit they fled on foot, in a motor vehicle, or by can result in serious injury to officers, sus- other means. pects, or bystanders, especially if conducted in a motorized vehicle, and such actions are As displayed in exhibits 4–6 and 4–7, in included in some definitions of what consti- most arrests, there was no flight (7,027 or tutes use of force. 93.5 percent) and no pursuit (7,089 or 94.4

Exhibit 4–4: Suspect use of weaponless tactics in 7,512 arrests

Arrests Percent of Arrests No Tactics Used 7,100 94.5 At Least One Tactic Used 412 5.5 All Arrests 7,512 100.0 Type of Tactic* Spit 74 1.0 Grab 114 1.5 Twist Arm 128 1.7 Wrestle 262 3.5 Push/Shove 166 2.2 Hit 66 0.9 Kick 74 1.0 Bite/Scratch 39 0.5 Pressure Hold 16 0.2 Carotid Hold 12 0.2 Control Hold 21 0.3 Other Tactic 54 0.7 Number of Tactics 1,026

* Since some arrests involved the use of more than one tactic, the percentages under type of tactic do not add to 100 percent.

34 Chapter 4: Measuring the Amount of Force Used

Voice force in any given situation. We have identified and elaborated on these elements in or- We include as a potential element of force der to record a broad range of activities that, what police said to suspects and what sus- under different definitions, could be consid- pects said to police. Our police survey form ered use of force. This project purposefully listed four categories of speech—conversa- attempted to measure aspects of police- tional, commands, shouting or cursing, and public encounters, such as weapon use, that verbal threats. In more than half of all ar- clearly involved force and other aspects, such rests (4,599 or 61.2 percent), police reported as speaking in a conversational tone, that that they used a conversational tone with did not involve physical force. The design of the suspect; in 2,297 or 30.6 percent of ar- this research was to measure many specific rests, they reported that they commanded and concrete behaviors against which differ- the suspect to do something (exhibit 4–8). ent definitions of force could be applied. To Police reported shouting or cursing at suspects in 73 (1.0 percent) of the arrests and threatening them in another 58 (0.8 percent) arrests. Finally, the police reported that they Exhibit 4–5: Police use of restraints in 7,512 arrests said nothing to the suspects in 485 or 6.5 Restraint Type Arrests Percent of percent of all arrests in this study. All Arrests Suspects, according to the police survey, spoke to the police in a conversational tone No Restraints Reported 1,234 16.4 in 4,970 (66.2 percent) arrests (exhibit 4–9). Handcuffs 6,182 82.3 The police reported that suspects used a command voice in 240 (3.2 percent) arrests, Leg Cuffs 67 0.9 shouted or cursed at officers in 638 (8.5 per- More Severe 29 0.4 cent) arrests, and made verbal threats in 473 (6.3 percent) arrests. The survey form could not capture the details of what was said by officers or suspects, but officers reported the Exhibit 4–6: Police pursuit in 7,512 arrests use of shouting or profane language as well as the use of threats in a small percentage Pursuit Type Arrests Percent of of arrests; in a larger percentage, but still a All Arrests distinct minority, of arrests, suspects made No Pursuit Reported 7,089 94.4 threats or conversed using a raised voice or obscene language. Pursue on Foot 224 3.0 Although the core understanding of the use Pursue in Car 177 2.4 of force typically does not involve what is Pursue in Helicopter 22 0.3 said but what is done, the nature of verbal communication, especially if it involves threats, shouting, or cursing, can be an element of force and needs to be incorporated Exhibit 4–7: Suspect flight in 7,512 arrests into how we think about and measure the use of force. Flight Type Arrests Percent of All Arrests Summary: Elements of force No Flight Reported 7,027 93.5 The five elements of force—weapons, weap- Flee on Foot 354 4.7 onless tactics, restraints, motion, and voice— identify different dimensions of the use of Flee in Car 128 1.7 force and provide a framework to measure Other 3 0.0 the existence of force and the amount of

35 Use of Force by Police

not record all aspects of each element of measures of force used by police officers— force would preclude the possibility of using Physical Force, Physical Force Plus Threats, those aspects to determine if force was used Continuum of Force, and Maximum Force. and, if it was used, how much force was used. We also developed four comparable measures of force by suspects. Each of these The behaviors recorded capture the details measures is a summary of behaviors derived of specific behaviors by officers and suspects, by combining specific actions by law enforce- but they do not constitute complete mea- ment officers and by suspects in different sures of force. Taken singly, none of the five ways. We recognized that no single measure elements of force captures fully everything is likely to capture well all the different un- that is typically meant by the use of force. A derstandings of the use of force. Thus, this fully developed measure of force requires the research used multiple measures of force in use of definitions that determine precisely order to incorporate more precisely the vari- how combinations of these elements consti- ous ways in which force is conceptualized by tute the presence of force or increases in the the police, the public, and researchers. amount of force used by and against the police. The next section provides more detail on Physical force how we translated our five abstract sets of behaviors into measurements of force. The first measure is a traditional conceptual dichotomy of those arrests where physical Four Measures of Force force was or was not used. We defined the use of physical force for officers and for suspects Using the items included on the police sur- in a parallel but slightly different manner vey, the research team constructed four (see first definition in sidebar “Definitions of measures of force”). For both the police and for suspects, the definition of physical force Exhibit 4–8: Police voice in 7,512 arrests includes any arrest in which any weapon or Voice Tone Arrests Percent of weaponless tactic was used. In addition, we All Arrests include as examples of physical force by the police arrests in which officers used a more No Voice Reported 485 6.5 severe restraint—prone cuffing, hobble, body cuff, or leg cuff. Conversational 4,599 61.2 Command 2,297 30.6 Physical force plus threats Shout/Curse 73 1.0 Our second measure, Physical Force Plus Threats 58 0.8 Threats, includes all the elements of Physi- cal Force but adds use of threats and displays of weapons. This measure combines actual physical force with threatened force. Exhibit 4–9: Suspect voice in 7,512 arrests Although this combination may be inappro- priate for some purposes, threats of violence Voice Tone Arrests Percent of are typically reported as violence in other All Arrests measures, such as in the FBI’s Uniform Crime Reports. Our second measure incorpo- No Voice Reported 1,191 15.9 rates the threat component of the use of Conversational 4,970 66.2 force. Command 240 3.2 The use of dichotomies between physical force and no physical force is a traditional Shout/Curse 638 8.5 approach to understanding and measuring Threats 473 6.3 the use of force. The strengths of these dichotomous measures include their ability to

36 Chapter 4: Measuring the Amount of Force Used be applied consistently across all jurisdic- Continuum of force tions and types of law enforcement agencies The third measure of force, Continuum of and to capture those elements of force that Force, developed in this project captures the are frequently salient to the police and to the rankings of force commonly used by law en- public. Their weakness is that they group forcement agencies to indicate distinct levels together all uses of force, from a push or a of suspect resistance and levels of police shove to the discharge of a firearm, and response (see third definition in sidebar make no distinctions among activities, such “Definitions of measures of force.”).11 The as the use of handcuffs or pursuits, that are gradients of force used by the participating not typically included in definitions of physi- departments are similar to those used by cal force. To address these potential limita- many police departments in their arrest tions of these two dichotomous measures, we tactics training and in their policies on the developed two other measures with other appropriate use of force. Our measurement strengths and other weaknesses.

Definitions of measures of force Definition 1: Measure of Physical Force Police Suspect Use of Severe Restraints Use of Any Weaponless Tactic Use of Any Weaponless Tactic Use of Any Weapon Use of Any Weapon

Definition 2: Measure of Physical Force Plus Threats Police Suspect Use of Severe Restraints Use of Any Weaponless Tactic Use of Any Weaponless Tactic Use, Display, or Threatened Use Use, Display, or Threatened Use of Any Weapon of Any Weapon

Definition 3: Categories of Police Continuum of Force Charlotte-Mecklenburg Colorado Springs Dallas Officer Presence Officer Presence Officer Presence Verbal Direction Verbal Control Verbal Control Soft Empty Hand Soft Control Techniques Empty Hand Control Oleoresin Capsicum Control and Compliance Intermediate Weapons Hard Empty Hand Hard Control Techniques Lethal Force Intermediate Weapons Impact Weapons Lethal Force Lethal Force

St. Petersburg San Diego Police San Diego Sheriff Officer Presence Officer Presence Deputy Presence Verbal Direction Verbal Commands Verbal Direction Restraint Devices Control/Compliance Soft Hand Control Transporter Soft Impact Chemical Agents Takedown Lethal Force Hard Hand Control Pain Compliance Intermediate Weapons Countermoves Lethal Force Intermediate Weapons Lethal Force

37 Use of Force by Police

of this “continuum of force” is intended not item was “An officer uses a baton” and an- only to reflect the official policies of the par- other was “An officer threatens to use a hand- ticipating law enforcement agencies but also gun.” Officers were asked to rank these items to incorporate into our research the widely based not on departmental policy but on their held notion that the force/no force dichotomy personal experience. We asked all officers to is sometimes inadequate to capture all the indicate, in their opinion, how much force was important variations in the ways police involved in each type of force. handle encounters with the public and the This exercise resulted in a measure that nature of suspect resistance to the police. makes reasonable (but not necessarily per- Unlike the Physical Force dichotomies, the fect) distinctions between different types of Continuum of Force measures are purpose- force. Officer presence, conversation, and fully responsive to the specific use of force commands are ranked near the bottom and policy and training in each department. Be- the use of weapons, especially firearms, are cause these measures are not consistent, it ranked near the top. These rankings include is not possible to combine cases from the six some elements of force that are included nei- jurisdictions into one measure of police use ther in most discussions of force nor in our of force and one measure of suspect use of Physical Force or our Continuum of Force force. measures. For instance, officers rank the use of handcuffs at 28.2 and chasing a suspect The Continuum of Force measures capture in a car at 41.4. Experienced officers in our distinctions among types of force (like vari- survey ranked these behaviors as involving ous weaponless tactics and the use of weap- substantial amounts of force, but our mea- ons) that are not possible in the two Physical sures of Physical Force and Continuum of Force dichotomous measures. The Con- Force would count arrests that involved just tinuum of Force measures have a natural handcuffing or just a pursuit as involving no ranking of categories from less forceful to physical force or as mere officer presence. more forceful. This research quantifies that natural ranking and creates a scale in which The second step in developing the Maximum each category is considered more forceful Force measure is to determine if such than the previous category. One weakness of behaviors occurred in our sample of 7,512 using these policy categories as a numerical arrests and, if so, to weigh them according to scale is that this formulation, in Charlotte- the rankings made by police officers. When Mecklenburg for instance, considers the dif- police officers reported that they twisted a ference between officer presence and verbal suspect’s arm, the amount of force for that directions to be the same as the difference arrest was measured as 35.1; when they between intermediate weapons and lethal used a carotid hold, the amount of force was force. Few observers would accept that these measured as 56. When officers reported that differences are equivalent. they engaged in two or more forceful acts, we recorded the one with the highest ranking— Maximum force hence the name Maximum Force. This multisite research developed a fourth These rankings, depicted later in exhibit measure of force, which we call Maximum 4–13, are presented not as a perfect or uni- Force. This measure ranges from 1 to 100, versal scale but as an example of how the with a ranking score of 1 being the least amount of force can be quantified in a way forceful and 100 being the most forceful. We that approximates our understanding of created this measure in a two-step process. In variation in the use of force. For purposes of five of the six participating law enforcement this research, the important issue is that agencies, we asked 503 experienced officers to this type of measure captures important rank more than 60 hypothetical types of force aspects of the use of force that would be on a scale from 1 to 100. For instance, one missed if research were limited to simple dichotomies or the Continuum of Force

38 Chapter 4: Measuring the Amount of Force Used

Exhibit 4–10: Most of 7,512 adult custody arrests did not involve force (six jurisdictions)

100 82.9% 81.1% (6,229) (6,094) 80 No

Ye s 60

18.9% Percent of All Arrests Percent 17.1% (1,418) 20 (1,283)

0 Did Police Use Physical Force? Did Police Use or Threaten Force?

measures. There are differences between we believe, reasonable but not the only rea- grabbing and kicking and between threaten- sonable definitions that could be used. The ing to shoot someone and actually shooting Continuum of Force measures are derived them; the Maximum Force measure is an at- from departmental policies, but these poli- tempt to measure those real but imprecisely cies vary from department to department known differences. and within departments over time. The Maximum Force measure is the most inno- The development and use of a variety of vative effort and perhaps the least well de- detailed measures of force is intended to en- veloped, but unlike the other measures, it courage researchers and policymakers to ex- reflects the relative ranking of experienced plicitly include or exclude specific behaviors police officers. and to explicitly consider the severity of different types of force. It may be too early to We are not yet prepared to assert that one establish uniform measures, but it is not too form of measurement is to be preferred over early to start proposing uniform measures. other forms. Certainly, improvements can be made in the measures we have developed, Summary: Measures of force but future research needs to be explicit about how force is measured and to justify The use of force is not a simple concept that why the particular measures of force used is easily measured. This research has taken are appropriate. Until such measures are the issue of measurement seriously and developed and justifications provided, we developed a variety of measures that, as a recommend the four measures reported here. group, capture many if not all the crucial distinctions that are commonly made about the amount of force used by and against po- The Distribution of Force by Police lice officers. Our efforts at measurement are and Against the Police not definitive. The definitions of Physical Exhibit 4–10 displays the number and per- Force and Physical Force Plus Threats are, cent of the adult custody arrests in this

39 Use of Force by Police

Exhibit 4–11: Most severe force type used as percentage of the 1,418 arrests involving force 100

77% (1,092) 80

20 9.5% 11.1% (158) Percent of Arrests Involving Force of Arrests Involving Percent (135) 2.3% (33) 0 Threats Restraints Weaponless Tactics Weapons Used Types of Force Used

Exhibit 4–12: Type of tactics used as a percentage of all 1,184 tactics used

60 49.7% (589) 50

30 18.5% (219) 20 Percent of Tactics Used Tactics of Percent 8.3% 7.7% 6.8% 7.7% (98) 10 (91) (80) (91) 1.4% (16) 0 Twist Arm Wrestle Push/ Hit/Kick Control Other Grab Shove Hold Tactic Police Tactics Used

40 Chapter 4: Measuring the Amount of Force Used study that met our definition of Physical When the Maximum Force measure is ap- Force. Law enforcement officers reported plied, with its 1-to-100 ranking scores, most that they used physical force in 1,283 or 17.1 arrests still involve little or no force but a percent of our sample of adult custody ar- proportion of arrests involve substantial rests; they used physical force or threats of amounts of force. Exhibit 4–13 notes the force in 1,418 or 18.9 percent of the 7,512 large number of arrests (4,305 or 57.3 per- arrests in this study. Thus, whichever of cent of all 7,512 arrests) at a ranking of 28.2. these two definitions is used, our findings In these arrests, the most forceful behavior are that law enforcement officers used or by the police was handcuffing. The exhibit threatened to use physical force in fewer also reveals that there was variation among than one of every five adult custody arrests. arrests in which some form of forceful action was taken. Some actions, such as the display However, as we argued earlier, this simple or the use of a handgun, were very forceful dichotomous measure may not provide a suf- and generate a Maximum Force ranking of ficiently clear understanding on the nature 55.4 and 81.7. of force used by the police. To help better understand the nature of force, exhibit 4–11 The Maximum Force measure captures a displays the frequency with which different number of activities, such as police chasing elements of force (threats, restraints, weap- suspects in a car, that are not counted as onless tactics, or the use of weapons) were force in our definitions of Physical Force, the most severe form of force used among Physical Force Plus Threats, and in depart- all instances in which law enforcement mental policies on the Continuum of Force. officers used some force. The exhibit shows In addition, this measure takes into account that the predominant types of force used by differences, sometimes severe, between police officers do not involve firearms or types of force. Grabbing a suspect and using other weapons but some form of direct physi- a firearm are both examples of Physical cal contact, which we categorized as “weap- Force and are counted equally in that mea- onless tactics.” In almost 80 percent of all sure; in the Maximum Force measure, grab- incidents involving physical force or threats bing ranks at 33.0 and using a handgun at of force, the most severe form of force used 81.7. Thus, this measure captures items by law enforcement was a weaponless tactic; that officers think involve force and weigh less than 12 percent of the arrests which met the amount of force based on a scale that our definition of Physical Force Plus Threats can range from 1 to 100. involved the use of a weapon. Summary and Discussion Focusing more closely on weaponless tactics, exhibit 4–12 indicates those instances in This research collected information from a which officers use some form of weaponless systematic sample of adult custody arrests tactics. This exhibit reveals that the most and used that information to construct a frequent type of weaponless tactic was grab- variety of measures of force. We have used bing the suspect. In 7.8 percent of all 7,512 these data to describe the amount of force arrests and 49.7 percent of all 1,184 arrests used by the police in six urban jurisdictions. in which the police used at least one tactic, We have emphasized various definitions of the most severe tactic used was categorized force and demonstrated a variety of methods as a “grab.” Other less frequent types of tac- that explicitly and quantitatively describe tics involved the use of control holds, arm force. twisting, pushing or shoving, wrestling, or Our research suggests that no one measure hitting or kicking. These findings confirm captures well all the elements that go into the results of prior research that established our understanding of what comprises force by that most adult custody arrests do not in- police officers against civilians during an ar- volve force or threats of force and those arrest. At the present time, there is no single rests that do involve force are typically at conception of what constitutes the use of the low end of severity.

41 Use of Force by Police

Exhibit 4–13: Average ranking of police behaviors by 503 officers in 7,512 arrests

Ranking Number of Percent of Police Behaviors Score Arrests Arrests No Police Actions Reported 1.0 62 0.8 Police Speak in Conversational Voice 15.6 153 2.0 Police Gently Hold Suspect 15.9 83 1.1 Shaded rows identify Two Police Officers Present 20.6 668 8.9 police behaviors that Police Command Suspect to Do Something 22.0 99 1.3 occurred in 2 percent or Police Shout/Curse at Suspect 22.5 3 0.0 more of all arrests. Police Spit on Suspect 23.2 2 0.0 Police Chase Suspect in Helicopter 24.0 1 0.0 Police Verbally Threaten Suspect 25.4 5 0.1 Police Push Suspect 26.7 0 0.0 Police Use Handcuffs 28.2 4,305 57.3 Police Chase Suspect on Foot/Bicycle 29.3 95 1.3 Police Use Leg Restraints 30.0 14 0.2 Police Threaten to Use Flashlight 30.9 0 0.0 Police Threaten to Use Chemical Agent 31.7 1 0.0 Police Possess Canine 31.9 10 0.1 Police Threaten to Use Baton 32.0 1 0.0 Police Grab Suspect 33.0 461 6.1 Police Display Baton 34.6 4 0.1 Police Use Pressure Hold 34.7 10 0.1 Police Twist Suspect’s Arm 35.1 98 1.3 Police Use Other Tactic 35.2 32 0.4 Police Display Chemical Agent 37.0 7 0.1 Police Use Severe Restraints 37.1 17 0.2 Police Bite Suspect 37.7 0 0.0 Police Display Flashlight 37.8 7 0.1 Police Use Choke Hold 38.9 78 1.0 Police Possess Shotgun 40.2 640 8.5 Police Kick Suspect 40.6 1 0.0 Police Hit Suspect 40.8 2 0.0 Police Chase Suspect in Car 41.4 137 1.8 Police Use Chemical Agent 45.9 31 0.4 Police Threaten to Use Car as Weapon 46.0 0 0.0 Police Threaten Suspect With Canine 46.1 5 0.1 Police Wrestle With Suspect 48.2 184 2.4 Police Use Flashlight 49.9 23 0.3 Police Threaten to Use Shotgun/Rifle 51.8 1 0.0 Police Use Canine 52.1 12 0.2 Police Threaten Suspect With Handgun 52.4 2 0.0 Police Use Baton 53.0 6 0.1 Police Use Other Weapon 53.1 10 0.1 Police Display Handgun 55.4 165 2.2 Police Use Carotid Hold 56.0 31 0.4 Police Display Shotgun/Rifle 57.4 23 0.3 Police Use Car as Weapon 69.4 10 0.1 Police Use Shotgun/Rifle 79.2 2 0.0 Police Use Handgun 81.7 11 0.1 Average Ranking Score 30.0

42 Chapter 4: Measuring the Amount of Force Used force, and this constrains our ability to stitutes excessive force. Our understanding implement precise measures of the presence of these relationships might benefit from of force or the amount of force. The multiple more precise understandings of how force is measures we have developed and imple- measured and a comparison of the relation- mented here illustrate how different elements ship between the amount of force used by of force can be combined into meaningfully the police and the amount of force used distinct measures. In this research we found against the police. that the use of force is relatively infrequent, regardless of the measure used. When the use Notes of force does occur, the amount of force is usually at the low end of our measures of force. 1. Custody arrests involve transporting suspects to a detention facility; instances where Combined with the similar findings from our suspects are arrested and issued a summons initial study in Phoenix, these substantive to appear before a judicial officer are not in- findings are beginning to provide a stable cluded in this research. picture of police behavior and the amount of force the police use in arrest situations. 2. Geller, William A., and Michael Scott, These findings, however, remain tentative Deadly Force: What We Know, Washington, given the small number of jurisdictions DC: Police Executive Research Forum, 1992. involved in this research and the room for 3. Reiss, Albert J., Jr., “Police Brutality— improvements in methods of data collection Answers to Key Questions,” in Law and and in precision in measuring the amount of Order Police Encounters, ed. Michael Lipsky, force. Although this research demonstrates New Brunswick, NJ: Aldine, 1980: 127. that police agencies and researchers can work collaboratively to describe the amount 4. Vaughan, Diane, The Challenger Launch of force used by police officers, much remains Decision: Risky Technology, Culture, and to be done to improve our measurements Deviance at NASA, Chicago: University of and to use those measures to determine the Chicago Press, 1996. types of circumstances in which more force is 5. Garner, Joel, John Buchanan, Tom Schade, used. and John Hepburn, Understanding the Use This multisite research project collected of Force By and Against the Police, Research information about more than 50 potential in Brief, Washington, DC: U.S. Department predictors of force, and additional analysis of of Justice, National Institute of Justice, those data will be forthcoming in the near November 1996, NCJ 158614. future. This line of research holds great 6. Some arrests involved the use of more promise for identifying the actual nature of than one weapon. police use of force as well as identifying those characteristics of police recruitment, 7. After some discussion, the research team training, tactics, and philosophy that can as- (researchers and agency personnel) decided sist police departments in moving away from to include on our officer survey form the a reliance on the use of force and toward an possibility that a motor vehicle might be increased reliance on the use of information used as a weapon. and cooperation with the communities they 8. This research assumed that in success- serve. fully completed arrests, suspects would not Future research also needs to focus on the be using restraints on officers. relationship between the overall amount of 9. In 666 (8.8 percent) arrests, the officers force used in a law enforcement agency and explicitly marked on our forms that they did the nature, scope, and extent to which the not put handcuffs (or other restraints) on force that is used meets various social under- the adults that they arrested and took into standings and legal definitions of what con- custody. In another 568 (7.5 percent) of the

43 Use of Force by Police

arrests, the officers did not indicate whether ton, DC: U.S. Department of Justice, Na- they did or did not use any type of restraint. tional Institute of Justice, June 1995. 10. Granfield, John, and Jami Onnen, Execu- 11. For purposes of this research, some of tive Brief: Pepper Spray and In-Custody the participating law enforcement agencies Deaths, Alexandria, VA: International collapsed several formal departmental Association of Chiefs of Police, 1994; and rankings into a smaller number of measur- National Institute of Justice, Positional able categories that reflected distinct and Asphyxia—Sudden Death, National Law measurable differences in officer behavior. Enforcement Technology Center, Washing-

44 5 The Force Factor: Measuring and Assessing Police Use of Force and Suspect Resistance by Geoffrey P. Alpert and Roger G. Dunham

ollecting and interpreting information collected. The two police departments had C on police use of force is a persistent prob- 204 officers, 110 of whom were assigned to lem for police managers and researchers. patrol duties. The departments had 150,841 Although such data are critical to both the contacts with the public, 7 complaints of police and the public, they remain difficult to excessive force, 31 complaints of discourtesy, collect, measure, and interpret objectively. and 2 lawsuits during 1995. There has been an energetic effort to collect The Miami-Dade Police Department data on all police use-of-force incidents, in- (MDPD), located in Dade County, Florida, is cluding excessive force, by various groups responsible for all law enforcement activities Geoffrey P. Alpert, Ph.D., and by assorted methods.1 The problems in the unincorporated areas of the county. is Professor of Criminol- with data collection on such organizationally (It was formerly known as the Dade County ogy, College of Criminal sensitive and controversial acts suggest the Sheriff’s Department and Metro-Dade Police Justice, University of need for standardizing measurement and Department.) In addition, MDPD contracts South Carolina. Roger G. providing reliable and valid measures.2 with many municipal agencies in Dade Dunham, Ph.D., is Profes- sor, Department of Sociol- This chapter presents information collected County to perform specialized services ogy and Criminology, from the police departments of the sister cit- within those agencies’ jurisdictions. In 1995, University of Miami. ies of Eugene and Springfield, Oregon, and the unincorporated areas of Dade County the Miami-Dade Police Department. After a had a population of approximately 2 million, brief description of the sites and a presenta- and the county included 1,840 square miles. tion of their use-of-force information, the The Department had 2,725 sworn officers, concept of the “force factor,” which is a mea- 845 of whom were assigned to patrol. There sure of police force incorporating the officers’ were more than 1 million reported contacts actions relative to the suspects’ physical (arrests and nonarrests) during the 3-year resistance, is developed. study period (fourth quarter 1993 through 1995), of which 133 resulted in complaints of Description of the Sites excessive force, 243 in complaints of discourtesy, and 18 in lawsuits. Eugene and Springfield, Oregon, are located in the Willamette valley, cover 52 square These police departments differ on impor- miles, and had a combined population of tant characteristics. There are obvious geo- 178,000 in April 1995, when the data were graphic and size differences that make the

45 Use of Force by Police

findings relevant to a wide audience. Fur- Suspect characteristics and behavior ther, differences in the social and ethnic Suspects ranged in age from 12 years to 86 environments of the cities studied provide years. The average age of suspects was be- diverse contexts in which to analyze police tween 28 and 29. Eighty-four percent of the use of force. Unfortunately, different data suspects were males. Although no ethnic elements were collected from the sites, which information about suspects was collected on make some comparisons problematic. the agency form, there were very few minori- Eugene and Springfield Police ties. Fifty-two percent of suspects were calm, reasonable, and cooperative. However, 19 Departments percent were reported as under the influence The Eugene and Springfield dataset was cre- of drugs or alcohol and 17 percent as emo- ated from items in the Police Officers’ Essen- tionally upset or abusive. Eleven percent of tial Physical Work Report Form, which was suspects appeared mentally unstable and completed by department members during unpredictable, and 2 percent were violent. April 1995. These data are unique because Most suspects were perceived by officers to they include a broad range of police work but be average (55 percent) or below average were not collected to evaluate force used by (20 percent) in physical abilities. The major- the police. This dataset was part of a larger ity of suspects did not resist the officer (61 effort to identify physical abilities necessary percent); 18 percent put up only slight resis- for police work. Data included all police- tance. Four percent were characterized as public contacts, including those related to having a high level of resistance, 2 percent forceful encounters. As a result, an unobtru- as violent, and 1 percent as explosive. The sive measure of police use of force was most common type of resistance was to push available. or pull the officer to resist an arrest or to escape. The findings from the Eugene and Spring- field Police Departments are reported in the Officer characteristics and behavior following order: The ages of the officers ranged from 25 to 60, ● Circumstances surrounding the incidents. with a mean of 37 years. Length of service ● Suspects’ characteristics and actions. as an officer ranged from 9 months to nearly 34 years. Average length of service was 12 ● Officers’ characteristics and actions. years. Eighty-six percent of the officers were Most of the 562 police actions analyzed were males. Most officers were assigned to patrol initiated by dispatched calls (57 percent), (91 percent). although 33 percent of the incidents were Control tactics ranked by severity. Data initiated by the officer who observed a situa- presented in exhibit 5–1 show that at least tion and reacted to it. two verbal or physical control tactics were used per attempt to apprehend a suspect. Of Circumstances surrounding the contacts the 546 incidents covered by the exhibit, all The most common type of incident was involved at least one tactic (96 percent being street violence (25 percent). However, 14 per- verbal); 93 percent, two tactics; and 87 percent of all incidents involved domestic vio- cent, three. There is a fairly large percentage lence, and another 14 percent pertained to drop for incidents involving use of a fourth resisting an investigation. Thirty-two per- tactic (41 percent). Use of a fifth (8 percent cent of incidents did not fall into one of the of incidents) or sixth (5 percent) tactic was predefined categories. Most of the police relatively uncommon. action was taken to apprehend or control a Exhibit 5–1 shows use of multiple tactics person (76 percent). and how they fall on a use-of-force continuum. The types of force listed down the

46 Chapter 5: The Force Factor

Exhibit 5–1: Incidents involving officer use of control tactic types—listed as a continuum of force from least to most severe

Types of Control Tactics Incidents Involving at Least This Number of Different Tactic Types 1 Type 2 Types 3 Types 4 Types 5 Types 6 Types Verbal Command 525 — — — — — 96% Handcuff Suspect 16 480 — — — — 3% 94% Search Suspect 1 22 449 — — — 0.2% 4% 95% Use Wrist/Arm Lock 1 3 20 183 — — 0.2% 0.6% 4% 82% Use Takedown 1 1 1 24 18 — 0.2% 0.2% 0.2% 11% 40% Block/Punch/Kick — 1 — 2 9 — 0.2% 1% 20% Strike Suspect 1 — — 2 5 4 0.2% 1% 11% 13% Wrestle Suspect — 2 2 1 10 7 0.4% 0.4% 0.4% 22% 23% Pepper Spray — — 1 2 2 10 0.2% 1% 4% 33% Use Baton — — 1 — — 4 0.2% 13% Use Firearm — 1 — 9 1 1 0.2% 4% 2% 3% Other Tactic — — 1 — — 4 0.2% 13% Multiple Tactics 1 — — — — — 0.2% Incident Totals for Columns 546 510 475 223 45 30 Column Incident Total as 100% 93% 87% 41% 8% 5% Percent of All (546) Incidents

Total use-of-force incidents=546. Guide to exhibit 5–1: Regarding the data opposite the “Handcuff Suspect” tactic type, for example, the tactic was used as the first tactic in 16 use-of-force incidents in which, as noted at the top of the column, at least one type of control tactic was used. The 16 incidents are 3 percent of all (546) incidents in which at least one tactic was used. However, “Handcuff Suspect” was used as the second tactic in 480 use-of-force incidents in which, as noted at the top of the column, at least two types of control tactics were used. The 480 incidents are 94 percent of all (510) incidents in which at least two control tactic types were used.

47 Use of Force by Police

lefthand side of exhibit 5–1 are in a se- officers listed use of wrist or arm lock tech- quence that reflects a typical use-of-force niques (fourth level on the continuum) as the continuum, from the lowest level of force fourth tactic used. In 11 percent of incidents (verbal commands) to the highest (use of a with four tactics, officers listed takedowns, firearm). Across the top of the table is the which constitute the next degree of force on number of tactic types the officer used in a the continuum. The exception at the fourth- given incident. Although tactics are not nec- tactic stage is firearm use (4 percent of such essarily listed in the temporal sequence that incidents). Apparently, continuum adherence they occurred, it can be argued that officers breaks down more often in incidents when do follow a use-of-force continuum. four types of force are necessary. This becomes even more apparent in situations For example, when an incident involves at when five or six types of force are used by least one tactic type, the tactic listed as the the officer. In these situations, the use of first used is almost always the lowest level of force is more scattered and distributed along force: a verbal command (used in 96 percent the more severe end of the continuum. of incidents). Three percent of such incidents involved handcuffing as the first tactic used, Findings indicate that officers used mul- also at the low end of the continuum of force. tiple force tactics most of the time, usually Very few of such incidents involved higher two tactics, three tactics in 87 percent of the levels of force as the first tactic used. incidents, and four in about 40 percent. In these instances, officers also seem to follow When incidents involved at least two tactic a typical continuum of force, with minor de- types, the second did not include verbal com- viations that do not vary by more than one mands, in contrast to 94 percent of the inci- or two levels on the continuum. However, dents involving handcuffing the suspect. in the few instances when more than four Searching the suspect occurred in 4 percent force tactics per incident are used, they are of the incidents; few involved force more se- scattered over the more extreme end of the vere than a search as a second tactic. force continuum. For incidents entailing three tactic types, no Although we can infer a probable order of officers listed verbal commands or handcuff- tactics from the data, a temporal sequence ing as the third used. Most listed searching is not clear in all cases. Future research the suspect (95 percent of the incidents); should focus on this sequential ordering of about 4 percent of the incidents involved a officers’ use of multiple force tactics and the wrist or arm lock as the third tactic, which is concurrent level of suspect resistance. A the next level of force on the continuum; and well-conducted interaction model could con- very few incidents were recorded as involv- tribute to understanding police/suspect in- ing force levels further down the continuum. teractions in these dangerous situations. The pattern that emerges follows the tradi- Control tactics used with varying tional use-of-force continuum. The first tactic amounts of suspect resistance. The pur- used in an incident is nearly always the least pose of this analysis was to determine how severe use of force on the continuum, and the many officers followed the typical continuum second tactic used in an incident is nearly al- of force for a given level of suspect resis- ways the second-most lenient. Officers appar- tance. The continuum of force reflects an ently follow the continuum with very few escalation from verbal commands to deadly exceptions, and those exceptions seem to devi- force. Nearly 97 percent of the incidents in- ate by only a small degree in relation to the volved initial use of a verbal command when whole range. the suspect offered no resistance. Deviating This same pattern seems to pertain to those from this typical process of verbally direct- incidents involving use of a fourth tactic by ing the suspect increased when suspect re- officers. None listed the first three levels of sistance was moderate or high (13 percent force. In 82 percent of four-tactic incidents, of all suspects), and violent or explosive

48 Chapter 5: The Force Factor

(12 percent). Three percent of officers devi- suspect (327 incidents), most incidents in- ated from first handcuffing the suspect dur- volved officer verbal commands (8 percent), ing the encounter when there was no or handcuffing (65 percent), or wrist/arm slight resistance by the suspect. Deviating locks (27 percent). However, three incidents from the typical control process increased to (1 percent) did involve use of a firearm. almost 16 percent of officers when suspect The force used by some officers when facing resistance was moderate to high. Less than 1 slight resistance was more than the force percent (0.4 percent) deviated from searching used by officers who faced no resistance. the suspect when there was no resistance. When the suspects used slight resistance, Slightly more than 4 percent of the officers most incidents involved officer use of verbal deviated from this typical process when re- commands, handcuffing, or wrist/arm locks sistance was slight, 19 percent when resis- (altogether 90 percent). There were a few tance was moderate or high, and 13 percent cases of takedowns (3 percent), one incident when resistance was violent or explosive. in which an officer struck a suspect, and six situations in which an officer used a fire- Highest level of force used. The next arm (6 percent) as the most severe tactic analysis involved determining the highest used. When suspects resisted at a moderate level of officer force used in each incident and or high level, officers used verbal com- comparing that with the level of resistance mands, handcuffing, or wrist/arm locks as by the suspect. This analysis provides an- their highest level of force in 48 percent of other way to determine if the level of the incidents. force used by an officer was consistent with the level of the suspect’s resistance. Finally, when suspect resistance was violent or explosive, all incidents involved force According to the data presented in exhibit beyond verbal commands and handcuffing. 5–2, when there was no resistance by the

Exhibit 5–2: Most severe control tactics used by the officer by level of suspect’s resistance

Control Tactics No Resistance Slight Moderate/High Violent or by Suspect Resistance Resistance Explosive Behavior Verbal Commands 26 (8%)* 7 (7%) 3 (4%) — Handcuff Suspect 211 (65%) 41 (41%) 6 (8%) — Wrist/Arm Lock 87 (27%) 42 (42%) 29 (36%) 4 (24%) Pepper Spray — — 2 (3%) — Block/Punch/Kick — — 1(1%) — Takedown — 3 (3%) 14 (18%) 5 (29%) Strike Suspect — 1 (1%) 4 (5%) 3 (18%) Wrestle Suspect — — 16 (20%) 2 (12%) Use Baton — — 4 (5%) 2 (12%) Use Firearm 3 (1%) 6 (6%) 1 (1%) 1 (6%) Total Incidents 327 (62%)† 100 (19%) 80 (15%) 17 (3%) Total use-of-force incidents=524. * The 26 incidents are 8 percent of total incidents (327) in the column. † The 327 incidents are 62 percent of all (524) use-of-force incidents.

49 Use of Force by Police

When suspects acted violently, officers ter with the police officer. Officers reported reported four incidents in which a wrist or erratic behavior 24 percent of the time. How- arm lock was the highest level of force used ever, 23 percent of the time, suspects were against the suspect (24 percent). The most calm when interacting with the officer. frequently used type of force was a takedown Suspect resistance and injury. Because (29 percent), followed by striking the suspect these cases involved some degree of use of (18 percent), wrestling the suspect (12 per- force by the officer, it is not surprising that cent), and using a baton (12 percent). One almost all cases involved suspects who officer reported an incident involving use of showed some degree of resistance (97 per- a firearm (6 percent). The data seem to sug- cent). The category of resistance most often gest that, in terms of the incidents as a reported was actively resisting arrest (36 per- whole, officers’ use of force reflected a con- cent), followed by assaulting the officer (25 tinuum ranging from lower to higher levels. percent). Twenty-one percent of suspects at- Miami-Dade Police Department tempted to escape or flee the scene. The most common type of suspect injury was The Miami-Dade dataset included 882 offi- a bruise or abrasion (48 percent of those in- cial Control-of-Persons Reports from the last jured). The next most common injuries were quarter of 1993 and all of 1994 and 1995. lacerations (24 percent) and injuries from These data were reported by the officer’s gunshots (4 percent). Most suspects resisted supervisor after interviewing the officer, by using their hands and arms only (65 per- suspect, and available witnesses. The cent). An additional 14 percent used their department’s computerized information was fists against the officer, and 12 percent used used to create the dataset. their feet or legs. Less than 5 percent used a Findings from the Miami-Dade Police De- gun (handgun, rifle, or shotgun). One percent partment are reported in the following order: used a vehicle to assault the officer, and another 1 percent used a cutting instrument. ● Suspects’ characteristics and actions. The most common type of force used by the ● Arresting officers’ characteristics and suspect was striking or hitting the police of- actions. ficer (44 percent). In 27 percent of incidents, ● Analyses of interaction patterns between the suspect pushed or pulled the officer, and officers and suspects. in another 20 percent, the suspect grabbed or held the officer. Eight percent of incidents ● Analysis of officer and suspect ethnicity. involved verbal threats or threatening movements or behaviors by the suspect. Suspect characteristics and behavior Role of alcohol or drug impairment on Suspects ranged in age from 12 to 90. The the suspects’ behavior. Another important mean age was 28.6 years. Of the 882 sus- question addressed was whether the suspect pects, 46 percent were black and 54 percent appeared intoxicated by alcohol or impaired white. Thirty-five percent of the total num- by drugs and how that affected the confron- ber of suspects were Hispanic, most of whom tation. The 370 suspects who were reported were white. Eighty-nine percent of the sus- intoxicated by alcohol or impaired by drugs pects were male, and 11 percent were fe- were less likely to be calm and more likely to male. Of the 42 percent of suspects who appear visibly upset (23 percent and 18 per- appeared impaired by alcohol or drugs at cent, respectively), and more likely to be er- the time of the incident, 24 percent were re- ratic in their behavior (24 percent) or highly ported affected by alcohol and 18 percent agitated (33 percent). were affected by a variety of illegal drugs. A number of suspects were highly agitated or Suspects who were reported impaired were erratic in their behavior during the encoun- no more or less likely to resist the officer than sober suspects, but when they did

50 Chapter 5: The Force Factor resist, they resisted in different ways. Im- However, 12 percent were given first aid, and paired suspects were more likely to resist 6 percent were treated by emergency rescue actively or to directly assault the officer than personnel at the scene. Less than 1 percent nonimpaired suspects. In spite of this, sus- of officers were treated at a hospital or by pects who were reported impaired were no their personal physician. more likely to receive force by the officer or Role of officer characteristics. In no to be injured during the arrest than were department do all officers respond precisely sober suspects. Similarly, suspect impair- the same to situations, although rules, ment by drugs or alcohol was not related to regulations, and policies of the department whether the officer was injured during the should narrow the range of officers’ re- incident. sponses to within acceptable and appropri- Although the overall significance of the rela- ate limits. In the cases examined here, tionship between impairment and type of officer characteristics did not make much of “There do not appear to be suspect resistance was not statistically sig- a difference in whether force was used or in any empirically validated nificant, there was a fairly large difference the level of force used. There were no statis- research studies which in resistance with a gun. Suspects reported tically significant differences in the level of support the assertion that as impaired were more than twice as likely force used by male and female officers. Fur- race, ethnicity, gender, or than sober suspects to use a gun to resist ther, the ethnicity of the officer did not affect age of police officers are the police. the general level of force used or whether related to misuse of physi- force was used. Officer age differences were cal or deadly force.” Officer characteristics and behavior statistically significant, but the differences —New York State Com- may reflect the differences in assignments of mission on Criminal Jus- The officers ranged from 21 to 66 years of younger versus older officers, which was not tice and the Use of Force, age, with a mean age of 34. Most officers studied. As the average age of the officers Report to the Governor, were Anglos (54 percent), Hispanics (31 per- increased, the level of force they used Vol. I, New York: New cent), and blacks (14 percent). Eighty-nine decreased. York State Commission percent of the officers were male, and 11 on Criminal Justice and percent were female. Most officers were as- Interaction patterns between officer the Use of Force, May signed to patrol (92 percent), and 5 percent and suspect 1987: 301. were sergeants. This section focuses on the interaction pat- Officer force and injury. The most com- terns between officer and suspect. In other mon type of force used by officers was use of words, is there a relationship between the hands and arms (77 percent of use-of-force suspect’s initial behavior and the officer’s incidents). In 8 percent of use-of-force response? Ninety-two percent of suspects incidents, officers used (discharged) their offered some resistance. The categories weapons, and in another 7 percent they used included “attempted to flee” (31 percent), dogs (K–9s). In a majority of the incidents “actively resisted the officer” (23 percent), (64 percent), officers grabbed or held the sus- “passively resisted” (20 percent), “assaulted pects. The next most common use of force the officer” (17 percent), and “resisting to was to strike or hit the suspect (10 percent of incite others” (1 percent). Although calm sus- the incidents). pects were the least likely to actively resist The most common injury to officers was or assault the officer, they were the most bruises or abrasions (64 percent of those likely to attempt to flee, even more so than injured), followed by sprains or strains (15 suspects perceived to have mental deficien- percent), and lacerations (15 percent). Of cies or problems (as defined by Florida’s injured officers, 2 percent were bitten by the Baker Act). Further, it was the suspects who suspect, 2 percent suffered broken or frac- initially acted in a calm manner who were tured bones, and 1 percent were injured by the most likely to resist an officer with a gun gunshots. The vast majority of injured officers or to assault the officer with a vehicle. received no treatment (76 percent).

51 Use of Force by Police

The initial behavior of the suspect did not tions is important. Officers used higher influence whether the suspect was injured levels of force against suspects of their own during the arrest, but it did influence the ethnic group than against suspects of other level of force used by the officer. Suspects ethnic groups. For example, Anglo officers who were initially calm were the least likely used higher levels of force against Anglo sus- to have force used against them. They were pects than black or Hispanic officers used no more likely to have slight force used against Anglo suspects. Black officers used against them than other suspects. However, higher levels of force against black suspects they were among the top two groups to be than did Anglo or Hispanic officers, and forcibly subdued by the officer using some Hispanic officers used more force against method other than hands. Hispanic suspects than did Anglo or black An analysis of suspects’ initial behavior and officers. The differences were the least pro- officers’ injuries resulted in an interesting nounced for Anglo officers and the most finding relating to the dangerousness of pronounced for black officers. Baker Act suspects. Suspects who were de- Among other explanations, this could be scribed as visibly upset or highly agitated due to a tendency to deploy officers in areas inflicted more officer injuries than other with a preponderance of citizens of their own suspects (40 percent and 39 percent respec- ethnicity. However, with the greater diver- tively), and Baker Act suspects inflicted sity of neighborhood ethnicity in recent fewer injuries than other suspects years, this finding may reflect a proclivity on (20 percent). the part of officers to respond differently to There was a strong relationship between the members of various ethnic groups. If this level of officer force and the chance of officer were true, each ethnic group might feel more injury. Increasing levels of officer force, re- comfortable using force on suspects from gardless of the level of suspect resistance, cor- its own group. Another interpretation is responded with higher probabilities of officer an officer’s possible concern that race and injury. When no force was used, 2 percent of politics might be dragged into the situation officers were injured. Minimal force situa- when an officer uses force against a suspect tions resulted in 15 percent of officers being of another ethnic group. As a result, officers injured, and situations involving officers forc- may try to avoid such situations. ibly subduing suspects with their hands re- Data in exhibit 5–3 compare officer/offender sulted in 69 percent of officers being injured. ethnic matches with the degree of resistance However, when officers used force other by suspects. Although there does not seem to than their hands, injuries were reduced to be a relationship between ethnic matches 15 percent. and whether a suspect offers resistance, Clearly, increasing levels of suspect resis- there are differences in the levels of resistance increase the chance of an injury to the tance. Although based on a small number of attending officer. No resistance or passive cases, the ethnic match resulting in the resistance seldom resulted in an officer in- greatest likelihood of a suspect assaulting jury. However, when the suspect attempted the officer occurs when a black officer is to flee or actively resisted arrest, the chance arresting an Anglo suspect (46 percent). of an officer injury is increased dramatically. Contrast this to the likelihood of assault The chance of an officer injury increased when an Anglo officer is arresting an Anglo even further when the suspect incited others suspect (14 percent), or when a black officer or directly assaulted the officer. is arresting a black suspect (17 percent). In exhibit 5–4, officer ethnic matches are Ethnicity of officers and suspects compared with the level of force used by the The relationship between the ethnicity of the officer. Force was used most often when the officer and that of the suspect in force situa- officer was black and the suspect was Anglo

52 Chapter 5: The Force Factor 79 (10%) 165 (21%) 293 (37%) 41 (5%) 191 (24%) 794 (100%) † 5 (5%) 14 (13%) 24 (22%) 33 (30%) 2 (2%) 33 (30%) 111 (100%) e 3 percent of all (794) incidents. Resistance Resistance to Flee Resisted Arrest/Incite Officer Totals =0.006 The 25 incidents in the column ar Exhibit 5–3: Officer/suspect ethnic matches and resistance by suspect Ethnic Matches No Passive Attempted Actively Resisted Assaulted Row Anglo/AngloAnglo/BlackAnglo/HispanicBlack/AngloBlack/Black 2 (2%)*Black/Hispanic 7 (4%) 4 (3%)Hispanic/Anglo 7 (8%)Hispanic/Black 15 (8%) 20 (13%) —Hispanic/Hispanic 2 (3%) 13 (16%) — 2 28 (4%) 46 (18%) (23%) 48 3 (3%) (57%) 9 — (11%) 58 64 (37%) (32%) 6 (6%) 1 (13%) 2 (2%) 17 (21%) 7 (8%) 16 (8%) 2 (1%) 2 5 12 (18%) (11%) (14%) (35%) 6 30 (37%) 51 (26%) 43 24 (28%) (26%) 84 18 (100%) (40%) 4 199 (36%) 6 (100%) 9 155 (35%) (11%) (100%) 32 (35%) 1 14 (2%) (17%) — — 9 (10%) 13 81 (29%) (100%) 16 (18%) 5 (46%) 4 45 (24%) (100%) 91 (100%) 11 (100%) 17 (100%) Totals 25 (3%) p use-of-force incidents=794. Total total (84). 2 percent of the row * The 2 incidents are †

53 Use of Force by Police

(100 percent) or Hispanic (100 percent). in the same manner. Even though the force Force was used least often when the officer factor is a relative measure, in situations was Hispanic and the suspect black (93 per- where the level of police force is greater than cent). Force with hands was used most often the level of resistance, there is no necessary when the officer was black and the suspect implication that the level of police force was either Anglo (73 percent) or Hispanic was excessive or improper. For example, an (77 percent). Force, other than hands, was officer may justifiably use more force than used most often when the officer was Anglo does a suspect to gain control of a situation. and the suspect black (32 percent). Similarly, it is possible that a suspect’s resistance may exceed the level of force used by The Force Factor the officer. A force factor representing such a disparity does not necessarily mean that the Prior research on use of force by police has officer’s level of force was too weak or im- focused on the highest level of force used or proper. A weaker police use of force, relative the highest level reached in an encounter. to the suspect’s level of force, could represent This analysis differs from previous ones be- an incident in which a suspect shoots an cause its focus is on the level of force used by officer who was unable to respond. Similarly, the police relative to the suspect’s amount of it could represent a suspect who attacked an 3 resistance, which we call the force factor. officer but who was controlled with a mini- This section describes the force factor and mum of police force. In any case, the most concludes with comments on its implications interesting cases are those that reflect the as it applies to policy and training. greatest differences between force and To calculate the force factor, both the sus- resistance. pects’ level of resistance and the officers’ An important application of the force factor level of force must be measured and scaled is the analysis of police use of force within

Exhibit 5–4: Officer/suspect ethnic matches and level of force used by the officer

Ethnic Matches No Minimal Force With Other Row Force Force Hands Force Totals Anglo/Anglo 1 (1%)* 21 (25%) 45 (54%) 17 (20%) 84 (100%) Anglo/Black 7 (4%) 42 (21%) 88 (44%) 63 (32%) 200 (100%) Anglo/Hispanic 8 (5%) 38 (25%) 71 (46%) 38 (25%) 155 (100%) Black/Anglo — 2 (18%) 8 (73%) 1 (9%) 11 (100%) Black/Black 3 (4%) 15 (18%) 41 (50%) 23 (28%) 82 (100%) Black/Hispanic — 2 (12%) 13 (77%) 2 (12%) 17 (100%) Hispanic/Anglo 1 (2%) 13 (29%) 19 (42%) 12 (27%) 45 (100%) Hispanic/Black 6 (7%) 13 (14%) 48 (52%) 25 (27%) 92 (100%) Hispanic/Hispanic 7 (6%) 33 (30%) 57 (51%) 14 (13%) 111 (100%) Totals 33 (4%)† 179 (23%) 390 (49%) 195 (25%) 797 (100%) p=0.030 Total use-of-force incidents=797. * The 1 incident is 1 percent of the row total (84). † The 33 incidents are 4 percent of all (797) incidents.

54 Chapter 5: The Force Factor a police department. Comparisons can be To calculate the force factor, we subtracted made between units to understand the use of the level of resistance (1–4) from the level of force and the reasons for differences. Other police force (1–4), Force minus Resistance= comparisons can be made for various officer Force Factor. The range of the force factor is characteristics, such as tenure with the de- from –3 to +3. A zero is interpreted as force partment, training, and assignment, to gain commensurate with the level of resistance. insight into variations of use of force found For example, no resistance and no force within the department. Findings can help would be 1–1=0, or passive resistance and guide training and supervision. minimal police force would be 2–2=0. If the level of force is higher than the level of resis- Oregon use-of-force data tance, the force factor is positive, with one point for each level of discongruence up to a Using the Oregon data, we measured the maximum of +3. If the level of force is lower level of suspect resistance in four ordinal cat- than the level of resistance, then the force egories: (1) no resistance, (2) slight resistance, factor is negative, one point for each level of (3) moderate or high resistance, and (4) vio- discongruence up to a maximum of –3. lent or explosive resistance. The corresponding categories for officer levels of force are Exhibit 5–5 depicts the Oregon police officers’ (1) no force, (2) slight force, (3) forcibly sub- use of force in relation to suspects’ resistance. dued suspect with hands, and (4) forcibly The distribution of scores resembles a normal subdued suspect using methods other than (bell-shaped) curve. This distribution of hands. (See sidebar “Force-related cases indicates that most incidents fall in terminology.”) the middle, with fewer cases at the extremes.

Force-related terminology

Suspect resistance: control. In some cases in this resistance category, the officer decided that he or she No resistance. Suspect was cooperative needed to use weapons or other special and followed all verbal instructions given by tactics to gain control instead of engaging the officer. the suspect directly. Slight resistance. Suspect resisted the officer’s actions and the officer had to use Officer force: strong directive language and/or minimal No force. Officer used typical verbal force (skills) to encourage suspect to coop- commands. erate and follow directions. Slight force. Officer had to use strong direc- Moderate or high resistance. Suspect im- tive language and/or minimal physical force peded officer’s movement or resisted cuffing to encourage the suspect to cooperate and or placement in a car. This level of resistance follow directions. required the officer to use arm/wrist locks and/or distraction techniques or fighting skills Forcibly subdued suspect with hands. to gain compliance and control. Officer used an arm/wrist lock, takedown, block, punch, or kick, and/or struck or Violent or explosive resistance. In this, the wrestled the suspect. most extreme, level of resistance, the suspect struggled or fought violently and re- Forcibly subdued suspect using methods quired the officer to (1) use fighting skills to other than hands. Officer used chemical disengage, (2) use a chemical agent, baton, agent, baton, gun, or other special tactics or or firearm, or (3) continue fighting to gain weapons.

55 Use of Force by Police

Exhibit 5–5: Force factors for 538 use-of-force incidents (Eugene/Springfield, Oregon)

300 290

250

200

150 131

100 92 Use-of-Force Incidents Use-of-Force 50 19 1 2 3 0 Ð3 Ð2 Ð1 021 3

Force Factors

In the Oregon data, the distribution is is compelling, interpreting any differences slightly skewed to the positive side, meaning could be problematic because each dataset that, on average, more force than resistance represents a different selection of incidents was used. as discussed above.

Miami-Dade use-of-force data Conclusions and Implications for In this dataset, the level of civilian resis- Policy and Training tance from the Control-of-Persons Reports Police use-of-force policies set the tone for was recoded into four ordinal categories how legitimate force can be used against similar to those used to analyze the Oregon civilians in a particular jurisdiction. data: (1) no resistance, (2) passive resistance, Whether departmental policies have an im- (3) active resistance, and (4) assaulted of- pact in the area of nonlethal force is an em- ficer. The corresponding categories for levels pirical question that has yet to be answered. of police force are (1) no force, (2) minimal However, research on policies regarding the force, (3) forcibly subdued suspect with discharge of firearms and pursuit driving hands, and (4) forcibly subdued suspect us- indicates that policies, training, and account- ing methods other than hands. The force fac- ability systems make a significant difference tor was calculated using the same method in the number of firearm discharges and explained above. The distribution of scores pursuits in which officers and agencies are for the Miami-Dade data is close to a normal involved.4 Assuming that use-of-force inci- curve, but slightly skewed to the negative dents follow the same trend, a relationship side, indicating, on average, the use of less should exist between the use of force by po- force than resistance (exhibit 5–6). lice and the policies that govern such behav- Although a comparison between the two ior. Policies that govern use of force should sites of Miami-Dade and Eugene/Springfield focus on four main objectives: maximizing

56 Chapter 5: The Force Factor

Exhibit 5–6: Force factors for 838 use-of-force incidents (Miami-Dade, Florida)

500

409 400

300

224 200 143 Use-of-Force Incidents Use-of-Force 100 40 17 3 2 0 Ð3 Ð2 Ð1 0 1 2 3 Force Factors

the safety of officers, minimizing injuries cantly at risk for injury any time they use to civilians, protecting the rights of those force, particularly when they strike a sus- against whom force is used, and providing pect with their fists or use their hands and officers with the tools needed to make ar- arms to control a suspect. Because most rests effectively and restore order. The major use-of-force incidents (80 percent) involve objective of the policies and training is to the use of hands, arms, or fists, Miami-Dade reduce or minimize injuries. officers are most at risk for injury when using precisely the type of force that they The Oregon and Miami-Dade data paint report using most frequently. somewhat different pictures of the injuries suffered by officers during use-of-force inci- Overall, Oregon data show far fewer injuries dents. Of 803 incidents analyzed for injuries to officers during incidents involving the use from Miami-Dade, 308 (38 percent) resulted of force. Of 504 reported incidents where in a reported officer injury. The vast majority force was used, 9 (1.8 percent) resulted in an of reported injuries (79 percent) were minor injury to an officer. Officers in Springfield and consisted of bruises, strains, or soreness. and Eugene are most at risk for injury when Nevertheless, 45 officers were lacerated, wrestling (21.1 percent), striking (12.5 per- 6 were bitten, 5 suffered a broken bone or cent), or taking a suspect to the ground (3.8 fracture, 1 received a puncture wound, percent). In none of the eight incidents when 1 received internal injuries, and 3 were shot. a police baton or pepper spray was used did an officer suffer an injury. The series of blue bars in exhibit 5–7 depict the chances of officer injury (not including The chances of suspect injury are significant simple soreness) according to some of the no matter what type of force is used by the more common ways in which Miami-Dade police. The series of brown bars in exhibit 5– officers reported using force. These figures 7 summarizes the chances of suspect injury suggest that Miami-Dade officers are signifi- when various types of force are used by

57 Use of Force by Police

Exhibit 5–7: Chance of officer/suspect injury by type of police force used in Miami-Dade

100 Officer Suspect 90 81% (17) 73% 80 (454) 67% 64% 70 (6) (7) 60 48% 48% 50 (10) 43% (29) (264) 40 27% 30 (2) 22% (2) 18% 20 (11) Chance of Officer/Suspect Injury

0 Fist Hands/Arms PRÐ24 Baton Foot/Leg Handgun Type of Police Force Used

Miami-Dade officers. A suspect is more likely increased likelihood of injury is small. Fur- to suffer injury if struck with a fist than thermore, even in cases where a suspect was with a PR–24 police baton. This may be due injured, the force factor mean was still nega- to the training that police receive in how to tive (–0.114), indicating that, overall, officers use the baton in a manner that minimizes use force that is less than the resistance of- the risk of injury. In any event, the chances fered by suspects. These findings can be an of a suspect being injured are greatest when important source of information for formu- the officer uses his fists, hands, arms, feet, or lating policies and training that help reduce legs during the encounter. the possibility of injuries. The force factor analysis of the Miami-Dade In addition to the policy implications above, data yields two important findings with re- the findings from this research point to spect to injuries. First, the data indicate that several training issues that need to be ad- officers are more likely to be injured when dressed by police agencies. These issues using less force relative to the resistance of include: the suspect. In other words, if an officer does ● Better training is needed in the use of not escalate the amount of force used in re- weaponless (empty hand) control tactics. sponse to an increasingly violent suspect, the Because the vast majority of use-of-force officer is more likely to be injured. Second, incidents are low level in nature, police the data show that injuries to suspects in- officers will continue to rely on their crease only minimally as the amount of force hands, arms, and feet to control most re- used by the police increases relative to the sistive suspects. Currently, these common amount of resistance. Although suspects are types of encounters result in a dispropor- more likely to receive injuries when police tionate number of injuries to officers and use more force relative to resistance, this

58 Chapter 5: The Force Factor

suspects. If officers were better trained ● Officers need more and better training in and prepared to deal with these types of how to avoid or defuse violent encounters encounters, it seems likely that the num- before they arise. If future policies require ber and severity of injuries arising from officers to take reasonable measures to them would decrease. avoid the use of force, then officers must be properly trained in conflict avoidance ● A use-of-force continuum that matches and crisis management techniques. How suspect resistance with officer response successful an officer is at avoiding vio- levels, combined with a robust training lence is a function, at least in part, of how program that reinforces what level of well trained the officer is in defusing emo- force is appropriate in a given encounter, tionally charged situations. should help reduce officer and suspect injuries. Directions for Future Research ● Significantly more training is needed in the proper use of chemical agents. The In examining the use-of-force landscape and Oregon data indicate that pepper spray in discussing the findings of this research, was used in 2 of 547 use-of-force encoun- at least four important areas remain unex- ters. Similarly, pepper spray was used 4 plored. First, we know very little about the times in 803 encounters by the Miami- effectiveness of various types of nonlethal Dade police. The Miami-Dade figures are force used by police. What is needed is a undoubtedly low because the police de- comprehensive evaluation of the effective- partment does not issue chemical agents ness of all types of police force commonly to patrol officers but does permit use of used in street-level encounters. chemical agents in specialized tactical Second, research is needed that identifies operations. The Oregon officers appear to in detail the sequential order of how violent be using pepper spray infrequently. More encounters unfold. As noted above, there is a training on the use and potential abuse great need to develop an interactive model of chemical agents may help reduce the that can better explain the active and reac- number of officer and suspect injuries. tive aspects of these encounters. Although ● If the PR–24 baton is to be retained, anecdotal evidence is abundant, there is officers need regular retraining and prac- little empirical research on what factors tice in how to use it effectively. In Dade immediately trigger the use of force by and County, every reported instance when against police, how force is actually used by the PR–24 side-handled baton was used suspects against police, and how officers involved a strike. To those who advocate respond. its use, the advantage of the PR–24 is its Third, little reliable research exists that ability to be used as a defensive and con- identifies the extent of police use of excessive trol-type weapon. When employed prop- force. Although we can say with relative con- erly, the PR–24 can be used to trap and viction that police use of force occurs on an hold the hands and arms of suspects to infrequent basis, we cannot conclude with bring them under control. Apparently, the nearly the same certainty how many of those PR–24 is not being used to its full capac- incidents involve excessive force. ity. This is not surprising because the use of a PR–24 is a diminishing skill that Finally, there is a need to explore measure- takes a great deal of practice to retain ment issues and uses of the force factor. one’s ability to use it to full advantage. Studying police use of force without taking If officers cannot remain proficient in its into account levels of suspect resistance proper use, then police agencies should should be avoided. Research results that do reevaluate whether to continue to issue not include the relative measure of force fail the PR–24 or whether another impact to impart a thorough understanding of the weapon may be more appropriate. police-public encounter. Creating force factor

59 Use of Force by Police

scores for individual officers, assignments, ton, DC: U.S. Department of Justice, Bureau units, and departments can be an important of Justice Statistics and National Institute of step in understanding and controlling police Justice, April 1996, NCJ 160113. use of force. 2. Alpert, Geoffrey, and William Smith, “How Of course, measuring excessive force is Reasonable Is the Reasonable Man?: Police highly problematic; indeed, even defining ex- and Excessive Force,” Journal of Criminal cessive force is difficult and definitions may Law and Criminology, 85(1994): 481–501. vary considerably depending on the situa- 3. Alpert, Geoffrey, and Roger Dunham, The tion.5 In spite of this difficulty, if we consider Force Factor: Measuring Police Use of Force the importance to the Nation of knowing how Relative to Suspect Resistance, Washington, often its police officers abuse their authority, DC: Police Executive Research Forum, 1997. comprehensive research on excessive force must continue to receive a high priority. 4. Alpert, Geoffrey, Police Pursuit: Policies and Training, Research in Brief, Washington, Notes DC: U.S. Department of Justice, National Institute of Justice, May 1997, NCJ 164831; 1. Greenfeld, Lawrence, A., Patrick A. Langan, and Alpert, Geoffrey, and Lorie Fridell, Police and Steven K. Smith, Police Use of Force: Vehicles and Firearms, Prospect Heights, IL: Collection of National Data, Washington, DC: Waveland Press, 1992. U.S. Department of Justice, Bureau of Justice Statistics and National Institute of Justice, 5. Alpert, Geoffrey, and William Smith, “How November 1997, NCJ 165040; and McEwen, Reasonable Is the Reasonable Man?: Police Tom, National Data Collection on Police Use of and Excessive Force.” Force, Bureau of Justice Statistics, Washing-

60 6 A Research Agenda on Police Use of Force by Kenneth Adams

his chapter puts forth a research agenda large. This situation places the researcher on T on police use of force, with special atten- the horns of a dilemma. If one decides to live tion given to issues of excessive force. In with the problem of small sample size, there some respects the task is easy. We have a will be serious limitations on one’s ability to limited empirical understanding of the conduct statistical analyses. As a result, problem; hence, any new research will be a conclusions will be subject to heavy qualifi- welcome addition to the current state of cation, and the value of new information will knowledge. In other respects, however, the be diminished. If one opts for larger samples, task is difficult. If left to unbounded develop- the cost of research increases dramatically. ment of inquiry, a research agenda quickly This means that fewer investigations will be loses focus and takes on a scattershot qual- carried out for a given sum of money, bring- ity that diminishes the return on one’s scien- ing the cost-benefit issue to the foreground tific investments. Research budgets are not as a major consideration. There are several Kenneth Adams, Ph.D., is unlimited, so it is important to establish strategies to be discussed later, in particular Associate Professor and strategic priorities. Also, elements of uncer- triangulation and use of targeted samples, Chair of the Criminal tainty and unpredictability in the scientific that help the researcher walk a middle road Justice Faculty, School of enterprise make it difficult to say which av- through this dilemma. Public and Environmen- enues of investigation will be most fruitful. From a pragmatic point of view, use of force tal Affairs, Indiana Three general considerations guided the by police is difficult to study because it is a University–Indianapolis. development of this research agenda: sensitive and politically charged topic, espe- ● Research should provide new knowledge cially when issues of excessive force are that significantly increases our under- involved. In many instances, research will standing of the problem while operating depend on the cooperation of police agencies, within real-world constraints. and for police administrators cooperation with researchers can be a mixed bag, carry- ● Research should be policy relevant. It ing the possibility of harm as well as benefit. should lead to improvements in our Notions of professionalism and commitments efforts to deal with problems. to excellence push police executives to deal ● Research activities, taken as a whole, with use-of-force problems. Yet, they also are should be comprehensive and systematic. keenly aware of political realities. Pressures from within the organization, such as that Development of a research agenda on use of from unions, for example, can offer resis- force by police entails several considerations. tance to reform, and groups outside the Use of force is a relatively infrequent event. organization, such as elected government Consequently, sample sizes will tend to be officials and community organizations, often small and measurement error will tend to be

61 Use of Force by Police

stand poised to gripe at the first hint of whether force used by police in a specific failing. situation was excessive. The project of the International Association If we are to advance our understanding of of Chiefs of Police (IACP) is viable because excessive force, we need not only a common it collects data under two important condi- definition but also common perceptions of tions: voluntariness and strict confidentiality which instances fit the definition. Likewise, (see chapter 3). These conditions serve to the relation of use of force to other issues of insulate police departments from potentially police behavior and misconduct is easy to negative repercussions. They also work to delineate conceptually. As a practical matter, distort the scientific picture by presenting however, we do not know the size and data only from departments willing to coop- strength of these hypothesized relations. erate and by limiting access to important This information will allow us to determine information regarding the characteristics of when interrelated phenomena should be those police departments. What in the politi- treated separately or as one, or when one cal arena is considered prudent is in the phenomenon can be substituted for the scientific arena a handicap, operating as a other, as convenient or necessary. source of error and bias. Judging what force is excessive Within the context of these challenges and limitations, a multifaceted research agenda Research on excessive force by police must that moves forward on several fronts is grapple with the problem of determining needed. Four priority areas for future re- what is excessive. The complication is that search on police use of force are proposed: labeling force as excessive involves a judgment, the outcome of which depends on the ● Establishing the conceptual boundaries of information that is available, the criteria “excessive force.” that are used, and the manner in which the ● Improving measurement. criteria are applied. An obvious problem is that the criteria used in making judgments ● Identifying important variation and about excessive force are not always the correlates. same, leading to differences of opinion. ● Evaluating efforts to manage use-of-force Judges apply legal standards; police admin- problems. istrators apply professional standards; and citizens apply “common sense” standards. Establishing the Conceptual The information available on which to make Boundaries of Excessive Force these judgments also varies considerably. Judges typically render judgments in the In studying police use-of-force problems, we context of an adversarial trial process that, eventually come to ask, “What is excessive in principle, is designed to lay bare all rel- force?” and “How does excessive force relate evant information. Administrative review of to other forms of violence and misconduct by police conduct generally takes place behind police?” At an abstract level, the concept of closed doors by trained professionals who excessive force is not hard to delineate. A have to temper objective reasoning (e.g., go- variety of definitions already exist, some of ing by the book) with subjective understand- which are widely used and accepted. How- ing (e.g., “putting themselves in another ever, when it comes to applying a definition officer’s shoes”). Regarding the thoroughness of excessive force to individual situations, of internal police investigations, there is a thereby rendering a judgment that a par- professional “push” to be exhaustive and ticular police officer acted wrongly in a spe- demanding and a collegial “pull” to be politic cific situation, linguistic neatness breaks and practical. down into untidy legal complexity. Judg- ments will differ, sometimes widely, about

62 Chapter 6: A Research Agenda on Police Use of Force

The public tends to make judgments about education, and experience with the criminal police use of force with more emotion and justice system? Answers regarding racial dif- less information than do judges or police ad- ferences, for example, are relevant to percep- ministrators. They react, somewhat reflex- tions of fairness and building of community ively, to newspaper and television reports of solidarity. excessive force or stories from friends and Judgments of excessive force are complex neighbors claiming that the police victimized because they involve many considerations. “Excessive force is almost them or someone they know. Which factors carry the most weight in mak- always a matter of Finally, there are differences in how often ing these judgments? What role does the ac- degree...and circum- decisions are made and in the consequences tions and characteristics of the suspect play? stance. Excessive force of decisions. Lawsuits alleging excessive How are the actions and characteristics of complaints frequently force by police are relatively uncommon, al- the officer weighed? How do situational arise in situations where though the consequences of a single decision factors, such as time of day, location, and the officer, armed and can immediately and acutely transform en- number of bystanders, influence judgments? obliged to confront crimi- tire organizations and communities. Police What role does the police organization’s nality, was doing his job; administrative decisions regarding excessive rules and procedures play in judgments? where he was duty-bound force occur more frequently than judicial de- What does the public, as well as lawyers and to intervene; where dan- cisions, usually in the context of disciplinary police, think about legal and professional cri- ger was present and some proceedings or policy reviews. The short- teria for determining excessive force? These force was necessary; or, term consequences of these decisions focus issues are important if we are to know what where witnesses are lim- on the individuals involved in the situation. different kinds of people think about the ex- ited to the victim and The public’s judgments on excessive force cessive force problem. Likewise, these issues the officer and there is occur habitually, being made by innumerable are important if we are to understand and a marked divergence in people on a continual basis. The immediate interpret data on excessive force generated their views about what consequences of these decisions are seem- by courts and police organizations as well as happened.” —Cheh, Mary ingly trivial, but the cumulative long-term by survey research. M., “Are Lawsuits an An- effect can be devastating in terms of police- swer to Police Brutality?” The problem of excessive force relates to sev- community relations. And Justice for All: Un- eral other areas of police behavior that raise derstanding and Control- We know something about judicial and similar issues. In particular, excessive force ling Police Abuse of Force, administrative judgments of excessive force overlaps with use of force generally (exces- ed. William A. Geller and because decision criteria are spelled out, sive and otherwise), civilian injuries by po- Hans Toch, Washington, decisionmaking processes are in place, and lice, police misconduct and illegal behavior, DC: Police Executive Re- decision outcomes are visible. However, and work performance. search Forum, 1995: 235. much more can be known about these issues, particularly with regard to variations across Focus on the broader picture localities and organizations. Perhaps be- For several reasons, it is useful for research cause the public’s judgments are so silently to focus on the broader picture of all use-of- ubiquitous, we know almost nothing about force events. First, the process of identifying how these decisions are made. use-of-force incidents is more factual and There is a need for research on judgments less judgmental than that of identifying of excessive force. How do police differ in excessive force incidents. Hence, there are their judgments from judges and lawyers? fewer problems, both normative and scien- The answer holds implications for the out- tific, in counting and studying events. Also, come of lawsuits. How does the public differ recordkeeping systems already exist in many from police administrators in their judg- police departments that document use-of- ments? In an era of community policing, the force events in considerable detail. This answer holds serious implications for police- means that data are readily available for community relations. How do people vary analyses within and across departments. in their judgments with regard to personal Second, a focus on all use-of-force events characteristics, such as race, age, gender, leads us to consider the problem of excessive

63 Use of Force by Police

use of force, as distinguished from the prob- Finally, another perspective is to view the lem of excessive force. This turn of phrase excessive force problem as part of a much describes situations in which use of force, broader set of police misbehavior. It may be though arguably legitimate and justifiable that officers who engage in excessive force on a case-by-case basis, can be seen in the are “problem” officers in a much more gen- aggregate as disquietingly frequent and eral sense. On the job, they may disregard potentially problematic. A focus on excessive procedures, disobey rules, show poor judg- use of force shifts the question from “Was ment, and have bad attitudes. Off the job, force used legitimately in this situation?” to they may drink too much, abuse their “Why is force being used so often?” By con- spouses, and get into fights and traffic acci- centrating on use-of-force levels that are dents. An officer’s use-of-force difficulties potentially counterproductive and hence may be part of a larger constellation of prob- inadvisable, we indirectly address the exces- lem behaviors that become manifest in citi- sive force problem because in scrutinizing zen complaints, disciplinary actions, and use-of-force incidents for necessity, excessive poor performance evaluations. These rela- force events are likely to stand out in stark tions, which should be investigated, are relief. theoretically important because they may require that we broaden our explanations of Third, a broader perspective, one that em- excessive force to include other problem be- phasizes questions of how often force is used haviors. Such theoretical explanations are and whether force is used too often, leads us policy relevant to the extent that efforts to to consider a wider array of options for deal- deal with use-of-force problems require a full ing with problems of police-public violence. understanding of police misbehavior in order By not sorting use-of-force events into cat- to be successful. Finally, as a practical mat- egories of “good” and “bad” and then focusing ter, these relations can be used to identify on the “bad,” discussions of solutions can problem officers early on, before their prob- more easily move away from issues of pun- lems swell and get out of hand. ishment and discipline. By downplaying the moral element, we more quickly come to Overlap and convergence consider the variety of remedial mechanisms that exist in police organizations, such as Research should focus on relations among rules and procedures, training, and other various use-of-force problems, emphasizing programs. Thus, our attention is drawn to areas of overlap and convergence and con- aspects of the formal organization that can centrating on issues of measurement and be manipulated in ways that address use-of- prediction. Some questions that need to be force problems. answered are as follows:

Another approach to the excessive force ● What is the extent of overlap among rates problem is to concentrate on instances in of use of force, excessive force, and civilian which police injure civilians. Not every ex- injury? cessive force incident involves civilian inju- ● How stable are these relations? ries, and not all injuries to civilians involve excessive force. A focus on injuries is none- ● To what extent can one measure substi- theless useful because it emphasizes out- tute for the other? comes of police use of force that are serious ● Can overall rates of force be used to iden- and consequential. Also, police injuries to tify police departments or police officers civilians are easy to count because they are that are likely to have problems with ex- relatively unambiguous and because medical cessive force? Can rates of civilian injury records, generated independently of police be used in the same manner? records, can be used to investigate these events. ● Can problem behaviors in the workplace be used to identify officers who are likely

64 Chapter 6: A Research Agenda on Police Use of Force

to have problems with excessive force? a variety of measures into an index that is Can behaviors outside the workplace be robust because the combined errors of the used in a similar manner? individual measures tend to cancel each other out. Improving Measurement The various data sources that are available Measurement problems are common in the to investigate issues of police use of force social sciences. The quality of data typically already have been discussed elsewhere in can stand major improvement. This too is detail.1 The focus below is on steps needed to the case with police use of force. Although it improve data sources so that we can move is desirable to have a sound, unambiguous, toward a multiple-indicator approach. accurate measure of excessive force, for example, in police departments throughout Official records the country, it could take a decade or more For several reasons, official records are and of work before we begin to approximate this will continue to be the major source of infor- goal. In the near term, we will have to make mation on police use of force generally, as do with what we have and work for small well as on excessive force. Although official incremental improvements in data availabil- records provide less-than-perfect data, ity and quality. from a practical viewpoint the advantages strongly outweigh the disadvantages. Among Multiple indicators of use of force the benefits of a good system of official In situations in which data contain a fair recordkeeping are wide geographic coverage, amount of “noise” (meaning error) that collection of detailed information, the possi- masks the true picture of what is being mea- bility of linking various record systems for a sured, and in which significantly improved more comprehensive picture, and up-to-date measurement is not feasible, perhaps due to statistics. Also, because data collection takes the expense, time, or effort that is required, place at the local agency level, there is the use of multiple measures and triangulation possibility of distributing the burden of data of data is an advisable strategy. Multiple collection across many organizations. measurement involves use of several mea- The drawbacks of official record systems sures that capture different perspectives on typically are lack of standardization across the object being measured. The idea is that agencies and poor data quality. These prob- several measures, when compared with and lems are exacerbated in the context of na- contrasted against each other, will offer a tional coverage, which involves coordination more complete picture than any one mea- of many different agencies that are geo- sure, which, standing alone, is known to be graphically dispersed. For this reason, incomplete. simple, easy-to-use data collection proce- Triangulation is an analytic approach based dures are to be preferred over complex, diffi- on multiple measures. The approach capital- cult-to-master procedures, particularly in izes on the fact that each measure is imper- the early stages of establishing a reporting fect in different ways and that each offers a system. different perspective on the problem. If all measures point in the same direction, one’s Use-of-force reports confidence in the results of statistical analy- Many police departments require a use-of- ses is increased. The underlying phenom- force report to be filed any time an officer enon is robust enough that a consistent uses force against a civilian in the line of “picture” emerges through the noise. If, how- duty. The report describes the circumstances ever, measures point in different directions, of the event and the nature of the force that then one must consider the specific limita- was used. These reports can be extremely tions of each measure in making an assess- valuable on a number of counts. Use-of-force ment. Sometimes it is possible to combine

65 Use of Force by Police

reports allow police administrators to moni- Civilian injury and hospitalization records tor the number of incidents by individual Another source of information on police use officers, geographic areas, or organizational of force is medical records on injuries that units. They provide a basis for periodic ad- civilians receive at the hands of police. These ministrative review to determine whether records focus on injury and treatment, and proper procedure is being followed when they are maintained by health-related agen- force is used. They offer the possibility of cies rather than police agencies. Medical supplementing raw counts of incidents with records capture the most serious use-of-force detailed information on officers, suspects, incidents, including the most serious in- and the circumstances of encounters. Finally, stances of excessive force. documentation contained in use-of-force reports may prove useful in defending the All police departments should be required to department’s actions in litigation. maintain and publish statistics on civilian injuries caused by police officers. Police Although police officers now spend a lot of agencies should work in cooperation with time on paperwork, use-of-force reports are hospitals and emergency rooms to develop not unduly burdensome to officers because these statistics. they use force infrequently in their work. Garner and his colleagues surveyed all ar- Compiling and publishing statistics on civil- rests in the Phoenix Police Department for ian injuries should not place a heavy burden 2 weeks regarding use of force.2 During this on police agencies, again because these time, 1,585 adult arrests were surveyed, of events do not happen very often. Indeed, one The Phoenix study was which about 1 in 5 involved any type of might think along the lines of a supplemen- supported by the National physical force by police. Thus, a requirement tal Uniform Crime Reports reporting system Institute of Justice under that every use-of-force incident be docu- parallel to that on law enforcement officers grant 92–IJ–CX–K028. mented translates into roughly 150 reports killed in the line of duty but concentrating per week for a police department with more on civilian injury. The data can involve than 3,000 employees. Since most of the simple counts of events, broken down so that force that police use is at a relatively low criminal suspects can be distinguished from level (i.e., holding, grabbing), the number of other civilians and the method of injury can reports can be reduced by raising the thresh- be identified. old for reporting. Court records All police departments should be required to maintain use-of-force reports. If the Lawsuits involving police use of force tend officer’s use of force is serious, it is to the to contain allegations of excessive force, organization’s benefit to document its occur- although court records are poor measures of rence. Furthermore, these reports should be how often such incidents occur, because the subject to periodic review and scrutiny, both allegations probably constitute a highly in terms of general trends and on a case- biased and selective subset of police-citizen by-case basis. Finally, police departments encounters. Use-of-force incidents that lead should be required to report periodically to lawsuits typically reflect some combina- their statistics on use of force to a State or tion of a highly motivated plaintiff, egregious Federal agency. Although some effort will action by police, and a tenacious lawyer. The have to be made to standardize reporting occurrence of police misconduct is only one procedures, this suggestion could be imple- element in the litigation equation and it may mented fairly quickly. As computer use not be the most important. However, law- becomes more widespread among front-line suits are directly relevant to the issue of how police officers, reporting requirements much use-of-force transgressions by police should become less burdensome. Detailed cost society. Knowing the costs incurred un- suggestions for a national reporting require- der current policies and practices, as mea- ment can be found elsewhere.3 sured by monetary awards to plaintiffs, can

66 Chapter 6: A Research Agenda on Police Use of Force help put the price of reform into broader fi- entire city, only high-crime or low-income nancial perspective. These cases, which merit areas might be surveyed. In lieu of sampling systematic study, show that there are some all adults, only young adults or males might costs to having police use force, although it is be sampled. Instead of questioning all civil- not clear how much of these costs can be at- ians, only persons who have been arrested tributed to excessive force. might be queried. Presently, information about lawsuits against If survey efforts concentrated on high-risk police is distributed across several govern- persons or situations, the number of use-of- ment agencies, and there is little coordina- force incidents captured in a study could be tion and feedback. City attorneys may not increased, which would mean that statistical routinely notify police administrators about analyses would be more reliable and poten- the outcome of litigation, so they are not al- tially more sophisticated. The data can be ways in a position to follow up with appropri- used for a variety of purposes. They can be ate administrative action. Financial liability used in epidemiological fashion to generate is an important aspect of the use-of-force more reliable point estimates of police use of problem, and citizens are entitled to know force across various conditions. They can be what police misconduct costs them. Thus, used to gain a more complete theoretical every police department should make avail- understanding of use-of-force incidents using able on a periodic basis data on the number, forms of elaboration analysis that examine types, and outcomes of lawsuits filed against subgroups within the sample. They can be the department, separately identifying those used to study important subpopulations that involve allegations of excessive force as within society, both in terms of point esti- well as other use-of-force issues. mates and theoretical analyses. Depending on the purpose of the investigation, the char- Survey methods acteristics of the sample being targeted and Survey methods tend to be relatively ineffi- the types of questions being asked will vary. cient at capturing police use-of-force incidents. If the scope of persons included in a survey The yield depends to some extent on the is restricted, some questions will go unan- definition of force and on the timeframe being swered. For example, if a survey concen- referenced. Definitions that include verbal trates on males, information on females will threats will capture more incidents than defi- not be collected. The danger here is that by nitions limited to physical contact. Likewise, restricting inquiry, we may unintentionally questions about lifetime experiences will cap- confirm our misconceptions. Nonetheless, ture more incidents than questions about the targeted samples may be a sensible tradeoff past 6 months or the past year. when research funds are modest and when Two strategies for addressing the inefficiency information is available to weigh the cost of of survey methods in studying police use of the tradeoff. force are using targeted samples to increase In the Phoenix study, Garner and his col- the yield of incidents and supplementing leagues surveyed all persons arrested over survey research projects on other topics with a 2-week period.4 A reanalysis of the data use-of-force questions. Both these techniques, could tell us if the number of use-of-force in- which offer cost-effective ways of studying cidents captured for the same expenditure of infrequent events, are underutilized and resources would be greater by limiting cover- could be used to greater advantage in age to Friday and Saturday. The data also research on use of force by police. could tell us how use of force differs by week- Targeted sampling. Targeted samples trade day and weekend, thus informing us about breadth of coverage for a focus on high-risk the biases of a targeted “weekend” survey. persons or situations. Rather than covering The compromise, of course, is that the re- the entire country, a survey might be results would only be generalizable to weekend stricted to big cities. Instead of surveying an arrests.

67 Use of Force by Police

The supplement approach. Another strat- instruments, NCVS is very cost efficient, egy for increasing return on research invest- functional, and expeditious. ments is “piggybacking” one’s research onto The BJS supplement also highlights two another project. The strategy is an opportu- important aspects of research on police use nistic one that involves locating another re- of force. First, surveys of the general popula- search project that can be supplemented to tion are inefficient at capturing encounters suit one’s needs. The “other” research project in which police use force against the public. might be in the early stages of planning or Of the 6,421 survey respondents nationwide, initiation, or it might be an ongoing longitu- only 14 reported that police used or threat- dinal effort. If one is starting a research ened to use force against them in the past project, finding “partners” for combining sev- year, leading to an annual estimate of eral research agendas into a single research 500,000 persons nationwide. Calculation of vehicle is economical because development the margin of error around this estimate was and implementation costs can be shared. not attempted for the pilot survey, but it is Working out differences among partners can likely to be large because of the small num- be difficult, however, and can lead to compro- ber of cases involved. BJS will address these mises that seriously detract from the utility issues in a second pilot survey, which will and integrity of the research when viewed use a sample more than 10 times larger than from the perspective of one or all of its com- that used in the first pilot test. ponent parts. Another important feature of the BJS survey With regard to an ongoing research project, a is that it illustrates the complexity of use- secondary or supplemental project can capi- of-force encounters. In the BJS pilot test, talize on the efforts of a primary project in respondents who indicated that police had such areas as sample selection and training used force against them could have been of interviewers. However, the secondary asked more than 100 questions, depending project will have little control over almost all on their experiences. Recommended changes Among the observational aspects of the research design. This is fine if to the survey, for the most part, include the methods alluded to on the design of the primary project meets the addition of more questions to describe use-of- this page is systematic needs of the secondary project. To the extent force incidents in greater detail. Although social observation, a field that this is not the case, the piggyback ap- the length of the survey is not problematic, research method used to proach extracts a cost from the secondary given that the average time to complete it study police. “Researchers project in terms of scientific value that needs for persons who had contact with police was record events as they see to be debited against the economic savings. 10 minutes, the survey makes it evident that and hear them and do not Also, researchers may give only limited a substantial amount of information has to rely on others to describe opportunity to others to become involved in be collected in order to eliminate ambiguity or interpret events. The their ongoing research because they are con- and ensure that the data are most pertinent researchers follow well- cerned about protecting the integrity of their to policy issues. specified procedures that project. can be duplicated.” The Police-Public Contact Survey (see chap- Observational methods —Mastrofski, Stephen D, ter 2) by the Bureau of Justice Statistics Roger B. Parks, and Observational methods are highly inefficient (BJS) is a good example of the “supplement” Albert J. Reiss, Jr., et al., at capturing use-of-force incidents not only strategy.5 BJS added a set of questions about Systematic Observation because these are infrequent events but also police contact and police use of force to of Public Police: Applying because a researcher can observe only one or the National Crime Victimization Survey Field Research Methods two officers on assignment at a time. Bayley (NCVS), which is a 3-year, seven-panel rotat- to Policy Issues, Washing- and Garofalo had six observers spend a total ing longitudinal design carried out by the ton, DC: U.S. Department of slightly more than 2,000 hours observing Bureau of the Census. In this collaboration, of Justice, National Insti- police officers in the field.6 These observa- which involved a pilot test of a survey ques- tute of Justice, December tions identified 37 use-of-force incidents by tionnaire, BJS used only one of the NCVS 1998: vii. police, the majority of which involved rela- panels. As a vehicle for testing new survey tively low levels of force.

68 Chapter 6: A Research Agenda on Police Use of Force

A development worth mentioning is the use force is negatively correlated with levels of of video cameras in police cars. In many police officer supervision, increasing supervi- cases, the camera is set to operate automati- sion of line officers may reduce or eliminate cally when an officer leaves the car. The pur- the problem. Finally, correlates can help us to pose is to capture police-civilian interactions understand why things happen as they do, so that they are available for later review in and so they are the building blocks of theo- case there is a complaint about the officer’s ries. Again, if we know that the incidence of behavior. Although some details or aspects of excessive force is highest among new officers, the encounter may be missing, incidents so we might explain this relation by pointing to recorded can be considered as observational a lack of experience and training. The corre- data. These data can provide information on lates of excessive force might be quite differ- the frequency and characteristics of various ent from the correlates of excessive use of types of police-public encounters within the force, however. limited context of automatic recording from a patrol car. Police management might use Geographic and temporal variation the recordings in its efforts to oversee the In searching the list of possible correlates of behavior of front-line officers. These video excessive force, some factors stand out as recordings also might be used in training potentially more important than others. In programs and in research, when a high particular, variations in the incidence of degree of verisimilitude is required. A limita- excessive force that are associated with time tion of this strategy, however, is that it can and place are especially significant. Geo- be fairly labor intensive to review the many graphic variation, in terms of region of coun- videos. try or size of place, allows us to focus on problems of excessive force and excessive Identifying Important Variation and use of force in concrete terms. Correlates There is a tendency in police research to The main purpose of scientific research is to concentrate on a handful of departments in identify variation in things of interest and large metropolitan areas. There are several link this variation to other factors or vari- practical reasons for this. Big cities are ables, which are called correlates. In the be- hotbeds for all sorts of social problems, and ginning stages of research, scientists usually their large populations make them good focus on accurate measurement of variation. places to study infrequent events. Also, po- Thus, priority should be given to develop- lice departments in large cities tend to have ment of tools that allow us to determine how resources that facilitate research, such as often excessive force occurs. computerized recordkeeping systems. For these reasons, which are largely matters of Social science research, however, moves practicality and convenience, police depart- down several paths at the same time. While ments in medium- and small-sized cities researchers are busy trying to measure often are overlooked as potential research variation, they quickly jump to questions of sites. what factors are associated with higher or lower levels of excessive force. Larger metropolitan areas will contain a greater number of use-of-force incidents, Correlates, if reliable and substantial, allow and probably excessive force incidents, than us to predict. If, for example, the incidence smaller areas. However, this may be the of excessive force is positively related to the result of more police-public interactions incidence of violent crime, we know where or more police officers. When measured as problems of excessive force will be greatest. rates per 1,000 arrests or per 100 front-line Another practical side of correlates is that officers, use-of-force problems actually may they can allow us to take steps to mitigate be greater in smaller areas. Also, taken as a problems. Thus, if the incidence of excessive group, the many medium and small cities

69 Use of Force by Police

that dominate our country may constitute factors will help fill out the use-of-force the greater part of the use-of-force problem. picture in terms of social classifications. Thus, there is a need to broaden the scope of Criminal justice-related factors also are rel- police research to include departments in evant. Researchers would be interested in medium and small cities. Without this re- work-related variables, such as officers’ expe- search, we will not know the answers to rience, training, rank, and nature of work many important questions about the scope assignments. Also of interest would be use-of- of the excessive force problem and about the force histories, civilian complaints, disciplin- changing nature and distribution of the ary actions, and work performance ratings. problem across the country. Relevant criminal justice information about Studying temporal variation is important suspects would include prior criminal record, because it allows us to determine whether history of violence, and gang involvement. use-of-force problems are getting better or Situational factors deal with the social and worse. This information permits forecasts physical context in which use-of-force events about the future, which may be important in take place. Some important factors to be in- galvanizing the perceived need for action. It vestigated are characteristics of the physical also may provide a general sense of how well setting, circumstances regarding the initia- efforts to curb use-of-force problems are tion of police-civilian encounters, possible doing. drug and alcohol use by the suspect, presence An ability to investigate geographic and of weapons, number and types of primary temporal variation requires the collection actors involved in the situation, number and of standardized data across places and over types of secondary actors (e.g., bystanders or time. The IACP project (see chapter 3) is a witnesses), and immediate actions that pre- move in this direction, although the volun- cipitate use of force, with a focus on criminal tary nature of the reporting system limits activity, provocation, and threats, if any. the utility of the data. A voluntary system— Organizational factors concentrate on as- even one incorporating anonymity—allows pects of police departments. Although formal the worst departments to avoid participa- aspects of the organization are easier to tion, leading to underestimation of the document, informal aspects of the organiza- problem. The need for a fully nationwide re- tion are relevant as well. Some factors to porting system on police use of force already examine are: has been discussed at length, and viable recommendations on how to implement the ● Number of employees. Do larger depart- system already have been made.7 Significant ments tend to have more use-of-force improvement in understanding police use-of- problems? force problems depends in large measure on ● Number and types of administrative units, significant advancement toward this goal, with a focus on departments that control which has been slow to date. misconduct and violence. Does an ethics Correlates and use of force unit or an integrity unit make a difference? Does civilian review matter? There are correlates that span individual, ● Personnel allocation. Is the ratio of super- situational, and organizational factors that visors to line officers related to the fre- may be related to the incidence of excessive quency with which force is used? force. At the individual level, personal characteristics of police officers and civilians ● Expenditures. Do departments with higher involved in use-of-force incidents, viewed pay scales have lower rates of use-of-force separately and in combination, may be problems? important. Age, race, gender, education, and ● Workforce characteristics. Do departments economic status are typical correlates of with a large proportion of minority social behavior, and information on these

70 Chapter 6: A Research Agenda on Police Use of Force

officers, female officers, experienced both officers and suspects.8 In addition to officers, or college-educated officers have confirming these findings, research should lower rates of excessive force? investigate where these technologies fit into the use-of-force continuum. To what extent ● Training and supervision policies. Are the are these new technologies used as alterna- frequency, length, type, and quality of tives for higher or lower levels of force? Also, training programs related to the incidence what evidence is there to suggest that these of excessive force? new technologies are abused and thus are ● Departmental philosophy. Does commu- contributing in some way to the excessive nity policing bring fewer use-of-force force problem? problems? Do broken-windows or zero- The Christopher Commission report, through tolerance philosophies increase these its “list of 44” problem officers, drew attention problems? to the fact that a handful of Los Angeles po- Under the heading of organizational factors lice officers had extensive histories of use-of- might be included items external to the orga- force problems and that management often nization that shape the work environment. was unresponsive to this situation.9 Since These factors include crime rates and char- that time, there has been an increased inter- acteristics of the population being served, est in information systems that can identify such as poverty rates. Alternatively, these officers who are at high risk for using exces- might be labeled community characteristics. sive force. In essence, these systems are prediction devices that are updated with new Evaluating Efforts to Manage information on a regular basis. Criminal jus- Use-of-Force Problems tice researchers have a long history of experience with prediction devices as applied to The last part of the proposed research offender behavior, and many of the same agenda focuses on evaluation of attempts to problems and issues that surfaced in that manage police use of force. Here the concern context are germane here. Practical questions focuses on activities that are specifically tar- (such as availability and quality of data), geted at dealing with use-of-force problems. methodological issues (such as selection of These activities might include training, tech- statistical techniques and prediction error), nologies that provide police officers with new and ethical issues (such as consequences of options for safely and effectively controlling potentially noxious classification and privacy recalcitrant suspects, methods for identify- of information), all remain to be explored in ing rogue officers in advance, management relation to using information systems to help strategies for controlling officer behavior in manage use-of-force problems. the field, or programs designed to change Another technology that may influence use officer attitudes and behaviors in ways that of force is video cameras mounted inside reduce the use of excessive force. police cars to record police-public interac- Technologies tions. It was mentioned previously that video recordings could be used for research and Among the new technologies developed for administrative purposes and as evidence controlling criminal suspects, pepper spray in lawsuits. A possible consequence of this and stun guns have received considerable technology is that it may make officers more attention. These technologies often are circumspect in their use of force, which may promoted as alternatives to lethal force, show up as fewer use-of-force incidents and although they have utility in their own right fewer complaints about excessive force. Com- as tools for behavior control. Research by parisons between departments that are us- Kaminski and colleagues suggests that pep- ing this video technology and those that are per spray is effective in controlling disor- not could address this question. derly suspects and may reduce injuries to

71 Use of Force by Police

Training judges, and offenders, among others, to understand how they think about excessive force. Some attempts to deal with use-of-force prob- This research is important because social lems are based on strategies that target the problems often require shared solutions, and problem at a broader level. A good example of shared solutions require a common basis of this approach is training programs designed understanding and mutual respect for differ- to promote ethical behavior, enhance police- ences in views. community relations, or train officers in verbal techniques for achieving compliance from Second, more and better data on police use civilians. Excessive force issues often are of force is needed. Most discussions about addressed in these programs. police use of force occur in an empirical vacuum where arguments are made without In some instances, training programs that the benefit of solid, useful information. Vari- address use-of-force issues are part of a more ous suggestions for a national reporting sys- general effort to change police organization tem on police use of force have been made, and culture. Perhaps the most popular of but progress toward this goal has been slow, these efforts is community policing, which even though the information collected by emphasizes joint working relationships such a reporting system could be extremely between police and community to maintain valuable. As an interim step, it is recom- order and promote justice. To the extent that mended that all police departments maintain community policing leads to a sense of part- records documenting all instances of police nership that includes feelings of fellowship use of force and all injuries to civilians, and based on a sense of common humanity, that this information be reported regularly antagonism between police and the public to the public. Maintaining these records will should be mitigated and propensities for po- not pose a great burden to police agencies lice to abuse authority or for the public to because use of force is infrequent. Further- perceive abuse by police should be reduced. more, the knowledge gained should be well Part of the agenda for community policing worth the effort. evaluation research should be to determine whether a reduction in use of force and com- Third, also required is research on how use plaints of excessive force occurs. Research of force by police varies across time, across also should investigate whether changes in cities, and across individual police depart- attitudes about the use of force occur for ments. Further, we need research on indi- police and civilians. vidual, situational, and organizational factors related to variations in use-of-force Summary levels, along with excessive force levels. In the process of identifying correlates of use This chapter outlines a research agenda on of force and excessive force, we will increase police use of force, giving special attention to our theoretical understanding of these problems of excessive force. A variety of ques- events and advance our ability to predict tions are raised, both reflecting the complex- problematic situations. At an aggregate level, ity of use-of-force issues and the relative research also should concentrate on the rela- paucity of our knowledge about use-of-force tion between excessive use of force, meaning transgressions. Four sets of research objec- types of situations in which force is used tives are identified. with disturbing frequency, and excessive First, more work is required on what various force, meaning types of instances in which people have in mind when they refer to police use more force than necessary. excessive force and on how they adjudge Finally, we need to identify, document, and specific instances of police behavior when evaluate interventions, changes, and reforms questions of excessive force arise. We need to that may mitigate police use-of-force prob- study the general public, members of minority lems. Although our understanding of these groups, police administrators, patrol officers, problems is incomplete, the urgency of the

72 Chapter 6: A Research Agenda on Police Use of Force situation requires that we seek out actions 3. McEwen, Tom, National Data Collection that are likely to have a beneficial impact. on Police Use of Force; and Geller, William A., In some cases, such as community policing, and Hans Toch, “Improving Our Understand- where a broad spectrum of change is in- ing and Control of Police Abuse of Force: volved, what is required is sensitivity to use- Recommendations for Research and Action,” of-force issues in evaluation studies. In other in And Justice for All: 277–337. cases, such as with police ethics and integ- 4. Garner, Joel, John Buchanan, Tom Schade, rity units, there may be a track record of im- and John Hepburn, Understanding the Use pact on the organization that is waiting to be of Force By and Against the Police. investigated. In addition, the effectiveness of activities that specifically target use-of-force 5. Greenfeld, Lawrence A., Patrick A. issues, such as officer training programs and Langan, and Steven K. Smith, Police Use of information management systems, needs to Force: Collection of National Data, Washing- be determined before we can genuinely advo- ton, DC: U.S. Department of Justice, Bureau cate widespread adoption of these programs. of Justice Statistics and National Institute of Justice, November 1997, NCJ 165040. In other chapters of this report, recently completed and ongoing research projects on 6. Bayley, David H., and James Garofalo, police use of force are described. As a group, “The Management of Violence by Police Pa- these research projects illustrate rigorous trol Officers,” Criminology, 27(1)(February methodologies, sophisticated measurement 1989): 1–27; and Bayley, David H., and strategies, cost-effective design features, and James Garofalo, “Patrol Officer Effectiveness the utility of comparisons across several in Managing Conflict During Police-Citizen police departments. These projects provide a Encounters,” in Report to the Governor, Vol. good foundation for developing a research III, Albany: New York State Commission on agenda on police use of force. Yet, substan- Criminal Justice and the Use of Force, 1987: tially more research is needed. Useful, reli- B1–B88. able, sound knowledge provides the best 7. McEwen, Tom, National Data Collection venue for society’s attempts to deal with the on Police Use of Force; and Geller, William A., pernicious consequences of use-of-force and Hans Toch, “Improving Our Understand- transgressions by police. ing and Control of Police Abuse of Force: Notes Recommendations for Research and Action.” 8. Kaminski, Robert J., Steven M. Edwards, 1. McEwen, Tom, National Data Collection and James W. Johnson, “Assessing the on Police Use of Force, Washington, DC: U.S. Incapacitative Effects of Pepper Spray Department of Justice, Bureau of Justice During Resistive Encounters With Police,” Statistics and National Institute of Justice, Policing: An International Journal of Police April 1996, NCJ 160113; and Adams, Ken- Strategies and Management, 22(1)(1999): neth, “Measuring the Prevalence of Police 7–29; and Kaminski, Robert J., Steven M. Abuse of Force,” in And Justice for All: Un- Edwards, and James W. Johnson, “The Deter- derstanding and Controlling Police Abuse of rent Effects of Oleoresin Capsicum on Force, ed. William A. Geller and Hans Toch, Assaults Against Police: Testing the Velcro- Washington, DC: Police Executive Research Effect Hypothesis,” Police Quarterly, Forum, 1995: 61–97. 1(2)(1998): 1–20. 2. Garner, Joel, John Buchanan, Tom Schade, 9. Independent Commission on the Los and John Hepburn, Understanding the Use Angeles Police Department, Report of the of Force By and Against the Police, Research Independent Commission on the Los Angeles in Brief, Washington, DC: National Institute Police Department, Los Angeles: Independent of Justice, November 1996, NCJ 158614. Commission on the Los Angeles Police De- partment, 1991.

73 Bibliography

he publications below are, in whole or Garner, Joel, John Buchanan, Tom Schade, T in part, related to police use of force and and John Hepburn. Understanding the Use are among those generated by research sup- of Force By and Against the Police. Research ported by the National Institute of Justice in Brief. Washington, DC: U.S. Department or the Bureau of Justice Statistics either of Justice, National Institute of Justice, through grants to outside researchers or November 1996. NCJ 158614. (The related through studies by inhouse staff. For infor- 60-minute videotape: Garner, Joel H. Use of mation on the availability of a publication Force By and Against Police. Washington, whose listing includes an NCJ or FS number, DC: U.S. Department of Justice, National please contact the National Criminal Justice Institute of Justice. NCJ 159739.) Reference Service (see inside back cover). Geller, William A., and Hans Toch, eds. And Alpert, Geoffrey P. Police Pursuit: Policies Justice for All: Understanding and Control- and Training. Research in Brief. Washington, ling Police Abuse of Force. Washington, DC: DC: U.S. Department of Justice, National Police Executive Research Forum, 1995. Institute of Justice, May 1997. NCJ 164831. Greenfeld, Lawrence A., Patrick A. Langan, (The related 60-minute videotape: Alpert, and Steven K. Smith. Police Use of Force: Geoffrey P. Police in Pursuit: Policy and Collection of National Data. Washington DC: Practice. Washington, DC: U.S. Department U.S. Department of Justice, Bureau of Jus- of Justice, National Institute of Justice. NCJ tice Statistics and National Institute of 161836.) Justice, November 1997. NCJ 165040. ———. Helicopters in Pursuit Operations. Jefferis, E., R.J. Kaminski, S. Holmes, and D. Research in Action. Washington, DC: U.S. Hanley. “The Effect of a Video-taped Arrest Department of Justice, National Institute of on Public Perceptions of Police Use of Force.” Justice, August 1998. NCJ 171695. Journal of Criminal Justice, 25(5)(1997). Edwards, Steven M., John Granfield, and Kaminski, Robert J., Steven M. Edwards, Jamie Onnen. Evaluation of Pepper Spray. and James W. Johnson. “The Deterrent Research in Brief. Washington, DC: U.S. Effects of Oleoresin Capsicum on Assaults Department of Justice, National Institute of Against Police: Testing the Velcro-Effect Justice, March 1997. NCJ 162358. Hypothesis.” Police Quarterly, 1(2)(1998). Fridell, Lorie A., and Antony M. Pate. “Death ———. “Assessing the Incapacitative Effects on Patrol: Killings of American Law Enforce- of Pepper Spray During Resistive Encoun- ment Officers.” In Critical Issues in Policing: ters With Police.” Policing: An International Contemporary Readings, ed. Geoffrey P. Journal of Police Strategies and Manage- Alpert and Roger G. Dunham. Prospect ment, 22(1)(1999). Heights, IL: Waveland Press, 1997.

75 Use of Force by Police

Kaminski, Robert J., and Eric Jefferis. “The Consequences. Vols. I and II. Washington, DC: Effect of a Violent Televised Arrest on Public Police Foundation, 1993. Perceptions of the Police: A Partial Test of Pinizzotto, Anthony J., Edward F. Davis, and Easton’s Theoretical Framework.” Policing: Charles E. Miller III. In the Line of Fire: Vio- An International Journal of Police Strategies lence Against Law Enforcement. Washington, and Management, 21(4)(1998). DC: U.S. Department of Justice, Federal Bu- Mastrofski, Stephen D., Roger B. Parks, reau of Investigation and National Institute Albert J. Reiss, Jr., Robert E. Worden, Chris- of Justice, October 1997. tina DeJong, Jeffrey B. Snipes, and William Scrivner, Ellen M. The Role of Police Psychol- Terrill. Systematic Observation of Public ogy in Controlling Excessive Force. Research Police: Applying Field Research Methods to Report. Washington, DC: National Institute Policy Issues. Research Report. Washington, of Justice, 1994. NCJ 146206. DC: U.S. Department of Justice, National Institute of Justice, December 1998. NCJ ———. Controlling Police Use of Force: The 172859. Role of the Police Psychologist. Research in Brief. Washington, DC: U.S. Department of McEwen, Tom. National Data Collection on Justice, National Institute of Justice, 1994. Police Use of Force. Washington, DC: U.S. NCJ 150063. Department of Justice, Bureau of Justice Statistics and National Institute of Justice, The following are among the objectives April 1996. NCJ 160113. of ongoing NIJ-supported use-of-force research projects, which may result in National Institute of Justice. Oleoresin future publications: Capsicum: Pepper Spray as a Force Alterna- tive. Technology Assessment Program. Wash- ● Measuring use of force relative to suspect ington, DC: U.S. Department of Justice, resistance in two police agencies. National Institute of Justice, March 1994. ● Identifying the frequency, rate, and type ———. Preliminary Investigation of Oleo- of force used in a county police depart- resin Capsicum. Law Enforcement and ment, including whether the introduction Corrections Standards and Testing Program. of pepper spray affected use of force or the Washington, DC: U.S. Department of Justice, extent of injuries to officers and civilians. National Institute of Justice, April 1995. NIJ ● Determining, through a national survey Report 100–95. of police agencies, how they organize their ———. Positional Asphyxia—Sudden Death. critical-incident response capabilities and National Law Enforcement Technology Cen- what types of actions, including use of ter Bulletin. Washington, DC: U.S. Depart- force, that officers and civilians take dur- ment of Justice, National Institute of Justice, ing the incidents. June 1995. ● Surveying a national sample of law en- ———. High Speed Pursuit: New Technolo- forcement agencies to determine the gies Around the Corner. National Law prevalence, distribution, and principal Enforcement and Corrections Technology features of early warning systems as a Center Bulletin. Washington, DC: U.S. De- response to the problem officer and evalu- partment of Justice, National Institute of ating the effectiveness of such systems in Justice, October 1996. three police departments.

———. Pursuit Management Task Force. ● Examining the use of force by police when Research Preview. Washington, DC: U.S. encountering persons with impaired Department of Justice, National Institute of judgment. Justice, August 1998. FS 000225. ● Examining serious assaults on police in Pate, Antony M., and Lorie A. Fridell, with two jurisdictions to identify neighborhood Edwin E. Hamilton. Police Use of Force: Offi- risk factors and areas of significant spa- cial Reports, Citizen Complaints, and Legal tial clustering.

76 To obtain an electronic version of this report, please visit the NIJ World Wide Web site at http://www.ojp.usdoj.gov/nij or contact:

National Criminal Justice Reference Service Box 6000 Rockville, MD 20849–6000 Phone: 800–851–3420 Phone: 301–519–5500 E-mail: askncjrs.org