The Abuse of Police Authority

THE ABUSE OF POLICE AUTHORITY

A National Study of Police Officers’ Attitudes 2

This edition of The Abuse of Police Authority: A National Study of Police Officers Attitudes has been formatted for the Web in PDF format; the pagination differs from the printed version.

The Abuse of Police Authority 3

THE ABUSE OF POLICE AUTHORITY

A National Study of Police Officers’ Attitudes

David Weisburd

Rosann Greenspan Edwin E. Hamilton Kellie A. Bryant

Hubert Williams

Police Foundation 4

The Police Foundation is a private, independent, nonprofit organization dedicated to supporting innovation and improvement in policing. Established in 1970, the foundation has conducted semi- nal research in police behavior, policy, and procedure, and works to transfer to local agencies the best new information about practices for dealing effectively with a range of important police operational and administrative concerns. Motivating all of the foundation’s efforts is the goal of efficient, humane policing that operates within the framework of democratic principles and the highest ideals of the nation. The Police Foundation’s research findings are published as an information service.

The research findings in this publication were supported by Grant Number 97–CK–WX–0047, awarded on behalf of Community Oriented Policing Services, US Department of Justice. Findings, recommendations, and conclusions of the research reported here are those of the authors and do not necessarily reflect the official position or policies of the US Department of Justice.

©2001 by the Police Foundation. All rights, including translation into other languages, reserved under the Universal Copyright Convention, the Berne Convention for the Protection of Literary and Artistic Works, and the International and Pan American Copyright Conventions. Permission to quote is readily granted.

ISBN 1–884614–17–5

Library of Congress Catalog Card Number: 2001–130311

Police Foundation 1201 Connecticut Avenue, NW Washington, DC 20036-2636 (202) 833–1460 E-Mail: [email protected] www.policefoundation.org

The Abuse of Police Authority 5

Contents

Foreword ...... 9

Acknowledgments ...... 11

I Introduction ...... 12

II Methodology ...... 15

III Characteristics of the Sample ...... 19

IV Main Survey Results ...... 23

Abuse of Authority and the Use of Force 23

Code of Silence 25

Social Factors 29

Departmental Response 31

Controlling Abuse 32

Community-Oriented Policing 35

Subgroup Analysis 39

Race 39 Rank: Supervisors and Nonsupervisors 42 Region 46 Agency Size 48 Gender 50

V Conclusion ...... 52

Endnotes ...... 56

References ...... 62

Authors ...... 66

Police Foundation 6

Illustrations Tables 3.1 Officers’ Current Rank ...... 20 3.2 Education Level of Officers ...... 21 3.3 Racial Background of Officers...... 22 3.4 Hispanic vs. Non-Hispanic Officers ...... 22 4.1 Officers’ Attitudes Toward Limitations on Use of Force...... 24 4.2 Officers’ Perceptions of Use of Force Behavior in Their Department ...... 25 4.3 Code of Silence: Attitudes...... 27 4.4 Code of Silence: Perceptions of Behavior ...... 27 4.5 Scenario of an Unruly Suspect ...... 28 4.6 Perceptions of the Effects of Extra-Legal Factors on Police Behavior ...... 29 4.7 Police Perceptions of the Public’s Attitude Toward the Police ...... 30 4.8 Perceptions of Media and Citizens’ Concerns Toward Police Abuse ...... 31 4.9 Departmental Responses to Abuse of Authority ...... 32 4.10 The Role of Supervision in Controlling Abuse ...... 33 4.11 Officers’ Perceptions of the Effects of Training on Abuse of Authority ...... 34 4.12 The Community-Police Partnership ...... 35 4.13 Perceptions of the Effects of Community Policing on Abuse of Authority ...... 36 By Race 4.14 Police officers often treat whites better than they do African Americans and other minorities ...... 40

4.15 Police officers are more likely to use physical force against African Americans and other minorities than against whites in similar situations...... 41

The Abuse of Police Authority 7

4.16 Police officers are more likely to use physical force against poor people than against middle-class people in similar situations ...... 41 4.17 Community-oriented policing increases, decreases, or has no impact on the number of incidents of excessive force ...... 41

4.18 Community-oriented policing increases, decreases, or has no impact on the seriousness of excessive force incidents...... 42

4.19 Citizen review boards are effective means for preventing police misconduct...... 43

By Rank: Supervisors and Nonsupervisors 4.20 Good first-line supervisors can help prevent police officers from abusing their authority...... 43

4.21 If a police chief takes a strong position against abuses of authority, he or she can make a big difference in preventing officers from abusing their authority...... 43 4.22 Most police abuse of force could be stopped by developing more effective methods of supervision...... 44

4.23 Whistle blowing is not worth it...... 45

4.24 The code of silence is an essential part of the mutual trust necessary to good policing...... 45 4.25 It is sometimes acceptable to use more force than is legally allowable to control someone who physically assaults an officer...... 45

4.26 Police department rules about the use of force should not be any stricter than required by law...... 46

4.27 Community-oriented policing increases, decreases, or has no impact on the number of excessive force incidents...... 47 4.28 Community-oriented policing increases, decreases, or has no impact on the seriousness of excessive force incidents...... 47

Police Foundation 8

By Region 4.29 Frequent friendly contact with local residents and merchants increases the likelihood that police officers will accept free lunches, discounts, or gifts of appreciation for effective service...... 47 4.30 It is sometimes acceptable to use more force than is legally allowable to control someone who physically assaults an officer...... 48 4.31 Police officers always report serious criminal violations involving abuse of authority by fellow officers...... 49

By Agency Size

4.32 If a police chief takes a strong position against abuses of authority, he or she can make a big difference in preventing officers from abusing their authority...... 49 4.33 Good first-line supervisors can help prevent officers from abusing their authority...... 49 4.34 Most police abuse of force could be stopped by developing more effective methods of supervision...... 50 4.35 Police administrators concentrate on what police officers do wrong rather than what police officers do right...... 51

Figures 3.1 Officers’ Gender ...... 22

3.2 Officers’ Satisfaction With Career ...... 22

The Abuse of Police Authority 9

We expect our police “…to have the wisdom of Solomon, the courage of David, the patience of Job and the leadership of Moses, the kindness of the Good Samaritan, the strategy of Alexander, the faith of Daniel, the diplomacy of Lincoln, the tolerance of the Carpenter of Nazareth, and, finally, an intimate knowledge of every branch of the natural, biological, and social sciences. If he had all these, he might be a good policeman.” —August Vollmer, 1936

Foreword When the police fail to meet our expectations, we react with dismay, anger, and additional demands. Police corruption and abuse of authority have persisted since the beginning of policing, and were exacerbated late in the twentieth cen- tury by America’s drug epidemic. Every year, incidents of police abuse of authority cost local communities tens of millions of dollars in legal damages. Tax dollars are wasted. Careers are destroyed. The public trust is compromised.

Virtually every police department has policies prescribing officer conduct and regulating use of force. No police department or police chief should knowingly condone conduct that runs counter to either department policy or constitutional standards. While there is accountability for acts of corruption and other forms of wrongdoing in most police departments, there is little or no accountability for those who tolerate such an environment. How, for example, were a few officers able to brutalize Abner Louima within sound if not sight of first-line supervisors th and other department officials in New York’s 70 Precinct? Even good people, placed in the wrong situation, will do the wrong thing. Bad supervision, intense peer pressure, and an organizational culture that sends un- clear signals can cause honorable men and women to behave in dishonorable ways. The key moral problem for police departments is the same as it is for corporations, universities, labor unions, and government agencies: how can you create a culture that will induce members to strike the right balance between achieving an organizational goal and observing fundamental principles of de- cency and fairness? Values in police agencies come not just from documents that describe them but also from traditional police culture. Too often, there is a disconnect between policies and practices, a failure of police management to monitor behavior and to respond appropriately. If police leadership does not assume an aggressive role in ensuring that the police culture is one of integrity and accountability, officers will continue to cultivate their own culture in their own way.

Police Foundation 10

As this study reaffirms, commitment by the chief and command staff to uphold democratic values and eradicate discriminatory practices is key. Police administrators should proactively institute and enforce strong policies governing conduct, as well as systems to collect and analyze data relative to police-citizen contacts such as complaints, use of force incidents, and traffic stops. Such efforts would inform policy, guide recruitment and training, and build accountability necessary to restore and maintain public trust in the police. It is the lack of internal, systemic controls, and not “a few rotten apples,” that perpetuates problems of misconduct and abuse by police. Most of America’s police officers are honest, dedicated, hard-working public servants, and it is they, as well as the public they serve, who are victims of the “bad” cop.

Because of the nature of their responsibilities, the police have the power to intervene and become involved at very basic levels within the lives of American citizens. The nature of the police response—the manner in which officers inter- act with citizens and the methods by which they enforce the law—have critical implications for our democracy and the quality of life of our citizens. As Jerome Skolnick writes in his thoughtful essay, On Democratic Policing, “Order achieved through democratic policing is concerned not only with the ends of crime control, but also with the means used to achieve those ends.” Are police abuses inevitable in our efforts to control crime? What are police officers’ views on the code of silence, whistle blowing, and the ways in which race or class influence police behavior? What are effective means of preventing abuse of authority by police? This report provides a nationwide portrait of what America’s police officers think about these and other important questions of abuse of police authority. Hubert Williams President Police Foundation

The Abuse of Police Authority 11

Acknowledgments As in any large-scale project, we owe a debt of gratitude to many people who helped in developing our work. At the outset, we received much advice, encour- agement, and support, which continued throughout the project, from Dr. David Hayeslip (now at Abt Associates) and Joseph Kuhns from the Office of Commu- nity Oriented Policing Services. We would like to thank Mathematica Policy Research Inc. for setting and meeting the highest standards in survey research, especially project director Rhoda Cohen, senior statistician John Hall, telephone supervisor Deborah Reese, survey associate Phyllis Schulman, and all the telephone interviewers. We also thank Edward Maguire of George Mason University for sharing his police agency database and for assisting in developing the sampling frame, and we thank Colleen Cosgrove and Bill Matthews for assisting in selecting police departments for our focus groups. A number of policing scholars contributed greatly at various stages of the project, from initial design of the sample frame, through development of the survey instrument, and by participation in our focus group of police scholars and executives. Our gratitude goes to Carl Klockars, Peter Manning, Ramiro Martinez, Stephen Mastrofski, Albert Reiss, Jerome Skolnick, Alfred Slocum, and Robert Worden. We appreciate as well the contributions of Commissioner Thomas Frazier, Chief Jerry Oliver, Chief Jerry Sanders, and Director Robert Pugh in our focus group of scholars and executives. We learned a tremendous amount from the generous participation of the un- named officers who took part in our focus groups of rank-and-file officers and supervisors. We thank the officers who participated in the pretest, and especially the 925 officers from across the country who volunteered their cooperation in completing the survey interview. Finally, our thanks go to Fred Wilson, Chris Tutko, Rachel Dadusc, and Michael Clifton, formerly of the Police Foundation, and Kenneth Brunk of the Police Foundation, and especially to Mary Malina, Communications Director of the Police Foundation, who supervised the production of this report.

Police Foundation 12

INTRODUCTION

American society has long entrusted to Extreme examples of police abuse its police the authority to use force in often spark major public debate. …[P]otential the pursuit of justice, law, and order. Videotapes of Rodney King being abuse and This authority is often glorified in beaten by Los Angeles police offic- actual abuse books, television, and movies, where ers, as well as reports of the torture the police are seen as constantly of Abner Louima by New York City of [police] responding to violent felons with police, capture public attention and authority equally violent reactions. But the real- raise troubling questions about po- remain both ity of police use of force is much less lice abuse of force in a democratic dramatic and the boundaries of legiti- society. Are such events isolated ab- a central mate police use of force are much more errations in American policing, or are problem constrained than defined in popular they extreme examples of a more for police culture. The police indeed have general problem that plagues Ameri- discretion to use violence when it is can police departments? Does the fact agencies and required. However, the potential abuse that such events often involve minori- a central and actual abuse of such authority ties suggest important inequities of public policy remain both a central problem for po- law enforcement against particular ra- lice agencies and a central public policy cial, class, or ethnic groups? What concern. concern. measures can be taken to constrain police abuse, and which are likely to

The Abuse of Police Authority 13 be most effective? Such questions have sought to tighten the bonds between been raised and debated in the media, police and community and to empower by our politicians, and by police schol- both groups to act effectively against ars and administrators. However, the community problems. voices of rank-and-file police officers With the support of the Office of Com- and supervisors have not been heard. munity Oriented Policing Services of the U.S. Department of Justice, the This silence is particularly important, Police Foundation undertook to given the vast changes in organization, conduct a representative national …[V]iews of tactics, and philosophy that have survey that would uncover the attitudes rank-and-file occurred in American policing over the of American police about sensitive past three decades. At the forefront of police have questions of police abuse of authority. those changes has been the transition How do contemporary police view special from the use of traditional military and abuses of police authority? Do police significance professional models of policing to the see them as an inevitable by-product creation of innovative models of in this age of increased efforts to control crime and community policing. While the police disorder? What forms do they take? of community had earlier defined professionalization How common do police believe them policing, which as limiting the role of the community to be? What strategies and tactics do in American policing, today police seek has sought to police view as most effective in to work closely with the public in preventing police abuses of authority? tighten the defining and responding to problems Given the importance of the movement bonds between of crime and disorder. In turn, the mili- toward community-oriented policing, tary model of police supervision that police and we sought to define whether commu- gave little autonomy or authority to nity-oriented policing is seen to encour- community street-level officers has begun to be age or constrain the boundaries of and to replaced by more flexible modes of acceptable use of police authority. Has supervision that allow rank-and-file empower community policing enhanced the officers the freedom to develop con- movement toward police respect for both groups to tacts with the public and to define the rights of citizens, or has it fostered act effectively innovative problem-solving strategies. new police skepticism about the rule The police and the community are seen against of law in a democratic society? as partners in emerging models of com- community munity policing. Rank-and-file police— The following is our report on a tele- problems. as those closest to the public—have, phone survey of a representative in turn, become central actors in the sample of more than 900 police offic- movement toward community-oriented ers who were drawn from an estimated policing. The views of rank-and-file population of 350,000 American mu- 1 police have special significance in this nicipal and county police. Ours is the age of community policing, which has first national study of this type and,

Police Foundation 14

therefore, has particular significance for from abusing authority. They also understanding the attitudes of Ameri- argue that their own department can police toward abuse of authority takes a “tough stand” on the issue in the age of community policing. The of police abuse. Finally, they argue report examines the questions raised that a department’s chief and first- above using the survey responses of line supervisors can play an impor- police officers, as well as the insights tant role in preventing abuse of …[T]he first gained from focus groups conducted authority. earlier in the study (see Appendices B, national study C, and D). The major findings of the • Police officers believe that the pub- of this study are as follows: lic and the media are too concerned type …has with police abuses of authority. • American police believe that extreme particular cases of police abuse of authority • American police officers support occur infrequently. However, a sub- significance core principles of community polic- stantial minority of officers believe for under- ing; they generally believe that that it is sometimes necessary to use community policing reduces or has standing the more force than is legally allowable. no impact on the potential for attitudes • Despite strong support for norms police abuse. of American recognizing the boundaries of police toward police authority, officers revealed • A majority of African-American police officers believe that police abuse of that it is not unusual for police to ignore improper conduct by their treat whites better than African authority in fellow officers. Americans and other minorities, and the age of that police officers are more likely • American police believe that train- to use physical force against minori- community ing and education programs are ties or the poor. Few white police policing. effective means of preventing police officers, however, share these views.

The Abuse of Police Authority 15 II

METHODOLOGY

Our findings are based on a telephone only one previous national survey—a …[W]e had survey that was conducted by 1985 study of police officers’ attitudes a basic Mathematica Policy Research Inc. of toward issues related to rape—used a Princeton, New Jersey, under the di- randomly selected sample of police requirement rection of the Police Foundation. The officers (LeDoux and Hazelwood, to obtain a survey instrument was developed by 1985). In selecting our sample, we had representative the Police Foundation’s staff after con- a basic requirement to obtain a repre- sulting a wide range of earlier studies sentative sample of police officers na- sample of 4 and after conducting a series of focus tionwide. We designed a two-step pro- police officers groups composed of police scholars, cess. First, we sought the most accurate nationwide. police managers, and rank-and-file listing of police agencies throughout the 2 police. The survey itself took an av- country. Second, after selecting a sample erage of 25 minutes to complete and of participating agencies, we began our was carried out with careful concern task of procuring lists of officers from for protecting the anonymity, privacy, those agencies. 3 and confidentiality of participants. A recent study by Maguire, Snipes, As the sample design was developed, Uchida, and Townsend (1998) concluded background research revealed that al- that the sources generally relied upon though a number of studies have ran- for national-level information about domly sampled police departments, police agencies are inadequate. Maguire

Police Foundation 16

et al. explained the limitations of and The sampling frame, as thus defined, discrepancies between the FBI Uniform consisted of 5,042 police departments Crime Reports and the 1992 Census of that employ between 91.6 percent and Law Enforcement Agencies that had been 94.1 percent of all full-time sworn compiled by the Bureau of Justice Statis- officers who serve in local police agen- tics with the Census Bureau, as compre- cies in the United States. Applying the hensive lists of all police agencies in the regional classification system used in United States. Their study developed a the FBI Uniform Crime Reports (FBI, The sampling more reliable list of police agencies by 1994), we see that the officers repre- frame… combining the information contained in sented 1,377 departments from the the Uniform Crime Reports, the 1992 Northeast, 1,547 from the South, 1,383 consisted of Census of Law Enforcement Agencies, from the North Central, and 735 from 5 5,042 police and a third list of police departments pro- the West. Maguire (1997) estimates departments vided by the Office of Community Ori- the number of officers in these 5,042 ented Policing Services. This newly departments at about 350,000. that employ created list, with further corrections by We followed a method of multistage, between Maguire, served as the universe of or clustered, sampling, whereby the police departments for the Police 91.6 percent sampling frame was divided into sam- Foundation study. and pling units that were based on depart- 6 94.1 percent The Police Foundation, in consultation ment size. Those units were then dis- with several policing experts and stat- tributed into three strata, or groups, by of all full-time isticians, identified criteria for inclusion size of department and organized by sworn offic- in the sampling frame. The criteria geographic region. One stratum (the ers who serve established were as follows: “certainty” stratum) consisted of the nine largest departments. The second • The police department has primary in local police stratum contained 84 randomly selected responsibility for providing police departments with 25 or more full-time agencies in services to a residential population sworn officers (the “midsize” stratum). the United (thus eliminating special police The third group included 28 randomly forces). States. selected departments with at least 10, • The department has a minimum of but no more than 24, full-time sworn 10 full-time sworn officers. officers (the “small” stratum).

• The department is either a munici- To draw the random samples of offic- pal or county police agency (state ers of all ranks from each of the 121 police and sheriff departments, with departments, and then to contact the their wide range of responsibilities officers selected to be interviewed, the that may or may not include polic- Police Foundation contacted the 121 ing residential populations, were selected departments and requested the excluded from the sampling frame). following information:

The Abuse of Police Authority 17

• A roster with the names and ranks participate for a number of reasons (i.e., of all full-time sworn personnel, were not full-time sworn officers, were on suspension, were on long-term dis- • A badge or employee identification ability, etc.). They were, therefore, number for each officer, removed from the sample. Their elimi- • A phone number at the department nation left a final sample size of 1,060. where each officer could be con- Response rates in social science re- tacted, search are often used as the benchmark Of the • An address at the department where for evaluating the representativeness of each officer could be contacted, and, the sample and for determining the 1,060 eligible degree to which one can generalize • If possible, the shift each officer is officers,… from the survey results to the survey assigned to. 925 completed population. A generally accepted rule As each department’s list became avail- of thumb is that response rates of 70 the survey…. able in the form necessary, the random percent or above are viewed as “very This rate [87%] samples were drawn, advance letters good” (see Babbie, 1990; Babbie, 1992; is one of the were sent to the selected officers, and Maxfield and Babbie, 1995). Of the the process of phoning and conduct- 1,060 eligible officers in the sample, highest… ing the surveys was carried out. 925 completed the survey, for a in surveys completion rate of 87.3 percent. This Of the 121 departments contacted, 113 of police, rate is one of the highest achieved in ultimately agreed to participate, for an surveys of police, whether on the na- whether on overall departmental participation rate tional or state level (see, for example, of 93.4 percent. The eight departments the national LeDoux and Hazelwood, 1985; Pate that declined were from all three strata. or state level.… and Fridell, 1993; Martin and Bensinger, Thus, we lost (a) one department (from 1994; McConkey, Huon, and Frank, the nine) in the certainty stratum, for a 1996; and Amendola, Hockman, and participation rate of 89 percent; (b) six Scharf, 1996). Even when we combine from the 84 in the midsize stratum, for the departmental participation rate of a participation rate of 93 percent; and 93.4 percent with the officer comple- (c) one from the 28 in the small stra- tion rate of 87.3 percent, the combined tum, for a participation rate of 96.5 overall response rate of 81.5 percent is percent. The participating departments still well above the accepted standard. cooperated by submitting rosters of all full-time sworn personnel, with rank, In survey research, it is traditional to contact address, and telephone num- report the level of statistical confidence, 7 bers. From those lists, 1,112 officers sometimes referred to as sampling error, were randomly selected. As initial that can be applied to the estimates contacts were made, it was determined reported. For our study, that level of con- that 60 officers were ineligible to fidence was very high for percentages

Police Foundation 18

relating to the full sample. The 95 per- intervals and significance statistics were cent confidence intervals for responses adjusted according to the sampling 9 in the survey were generally between procedures we used. 8 2 and 4 percent. This figure suggests that we can be very confident that the Because of the stratified and clustered population characteristics associated with sampling procedures used in the study, the survey responses were generally it is necessary to include a correction within plus or minus 2 to 4 percent of when reporting survey responses. This those reported. If we were hypotheti- correction is based on weighting each cally to observe repeated samples like department and police officer accord- that drawn in our study and to calculate ing to the proportion of the actual a confidence interval for each, then only population of American police that 10 about 5 in 100 would fail to include each represents. In practice, weight- the true population percentage (see ing in the survey does not greatly alter Weisburd, 1998). This statistic is some- the majority of estimates that we times defined as the margin of error or report. Nonetheless, the weighted the sampling error of a study. Confidence estimates provide a more accurate intervals for subsamples in the study, picture of the true population of such as women or minorities, were larger. responses than that provided by the In those cases, we generally compare raw estimates. We report only weighted subgroups and report significance lev- percentages in the discussion and els. It should be noted that the standard tables below. errors used for calculations of confidence

The Abuse of Police Authority 19

III

CHARACTERISTICS OF THE SAMPLE

The survey represents a broad popula- The length of service of the sworn po- The survey tion of officers and reflects the diverse lice officers in the sample ranged from represents composition of American police. For less than 1 year to 35 years, with about example, 56 percent of the officers sur- 25 percent at 5 years or less. One in five a broad veyed defined themselves as “patrol officers had served from 6 to 10 years, population officers.” Another 16 percent were de- almost one in five had served from 11 to of officers tectives, criminal investigators, and 15 years, and more than one-third had corporals. Sergeants constituted about served 16 years or more. Most officers and reflects 15 percent of the sample, and another had patrol responsibilities (60 percent). the diverse 13 percent held the rank of lieutenant Some 30 percent were involved in other 11 composition or above (see Table 3.1). About 3 of field operations such as gang, juvenile, 10 officers in the sample noted that they robbery, and homicide, including 7 per- of American served as “supervisors.” While more cent who identified themselves as as- police. than 2 of 10 officers were under signed to community policing. More than 30 years old, more than 8 percent were 10 percent did not have field assign- over 50 years old. Officers ranged in ments, but served in administration, com- age from 22 to 66 years old. Regarding munications, technical support, and other marital status, almost three of four (74 jobs. This proportion is similar to that percent) were either married or living reported in the 1993 Law Enforcement with someone as if married.

Police Foundation 20

Officers’ Current Rank

Number Percentage Rank of Officers of Officers

Patrol Officer 514 55.7

Table 3.1 Table Detective/Criminal Investigator 110 12.0 Corporal 36 4.0 Sergeant 142 15.3 Lieutenant 56 6.1 Captain 17 1.7 Inspector 2 0.2 Major 3 0.3 Deputy Chief 6 0.6 Chief 14 1.5 Other 24 2.4

N = 924

Management and Administrative Statis- growth in professional police educa- This survey tics (LEMAS) survey of agencies with tion, more than half of those who had reinforces 100 or more officers, where 11 percent attended college reported that they had earlier studies of county police and 9 percent of majored in criminology, criminal municipal police did not have field justice, or police science. Some 15 per- that suggest assignments (BJS, 1995). cent of the weighted sample were that American continuing their education in pursuit Many scholars and policy makers have of a degree. policing emphasized the importance of educa- reflects the tion in developing a modern police (see This survey reinforces earlier studies racial and National Advisory Commission on that suggest that American policing Criminal Justice Standards and Goals, reflects the racial and ethnic composi- ethnic 1973; Carter and Sapp, 1990; Worden, tion of the U.S. population (see composition 1990; Travis, 1995). Almost one-third BJS, 1995). In the weighted sample, of the U.S. of the sample had a bachelor’s degree 80.8 percent of the officers were white, or higher (see Table 3.2). Additionally, as compared with 80.3 percent of the population. 52 percent had a two-year degree or population (Bureau of the Census, some college education, and almost 1991), and 10.7 percent were African 15 percent had graduated from high American, as compared with 12 per- school (or had a GED). Only five of cent in the national population (see 12 the officers surveyed had only some Table 3.3). Also, 9.6 percent of the high school education. Reflecting the weighted sample, compared with

The Abuse of Police Authority 21

Education Level of Officers

Number Percentage Highest Level Attained of Officers of Officers

Some High School 5 0.5

High School Graduate/GED 133 14.7 3.2 Table Some College 303 33.1 Associate’s Degree (2 year) 174 18.6 Bachelor’s Degree (4 year) 258 27.6 Some Graduate School 19 2.0 Master’s Degree 29 3.2 Doctoral Degree or Law Degree 3 0.3

N = 924

8.8 percent of the U.S. population, iden- Despite the controversies that surround While the tified themselves as of Hispanic, Latino, American policing, our survey shows that or Spanish origin (see Table 3.4). American police officers are generally racial satisfied with their career choice. Indeed, composition While the racial composition of Ameri- almost all of the officers we surveyed of American can policing may reflect the nation from (94 percent) indicated that they were which it is drawn, American policing satisfied and over half of those said that policing may remains a predominantly male profes- they were “extremely” satisfied with their reflect the sion. Only 8.5 percent of the sample choice of policing as a profession (see nation…, were women (see Figure 3.1). Other Figure 3.2). Only two officers described sources provided similar estimates. themselves as extremely dissatisfied with American According to the National Center for their career choice. Even when asked policing Women and Policing (1998), “Women about their satisfaction with their current remains a currently make up less than 10 percent assignment, more than 90 percent of the of sworn police officers nationwide.” sample indicated that they were predominantly This figure was also consistent with the satisfied, of whom 40 percent were male 1993 LEMAS survey, which reported “extremely” satisfied. Nevertheless, profession. that 8 percent of officers in municipal 46 percent of police officers described police departments and 10 percent of their work as extremely stressful (16 per- officers in county police departments cent) or quite stressful (30 percent). were women (BJS, 1995).

Police Foundation 22

Racial Background of Officers

Number Percentage Race of Officers of Officers

White 748 80.8

Table 3.3 Table African American 94 10.7 American Indian or Alaskan Native 8 0.8 Asian 8 0.8 Native Hawaiian or Pacific Islander 3 0.3 Other 36 4.3 Mixed Race 24 2.4

N = 921

Hispanic vs. Non-Hispanic Officers

African- White American Other Officers Officers Officers TOTAL

Hispanic 44 (6.2%) 2 (1.8%) 38 (51.6%) 84 (9.6%)

Table 3.4 Table Non-Hispanic 703 (93.8%) 92 (98.2%) 41 (48.4%) 836 (90.4%)

N = 920

Figure 3.1 Figure 3.2 Officers’ Gender Officers’ Satisfaction With Career

Female Dissatisfied 76 (8.5%) 52 (5.6%)

Figures

3.1 & 3.2

Male Satisfied 848 (91.5%) 871 (94.4%)

The Abuse of Police Authority 23 IV

MAIN SURVEY RESULTS

The survey consisted of more than 80 Abuse of Authority …[M]ost police and the Use of Force questions that relate to the problem of officers in the abuse of authority (see Appendix A). The use of force may be a relatively Below, we summarize the main find- rare occurrence in American policing United States ings of the study. First, we examine re- (Worden and Shepard, 1996), but those disapprove sults across the entire sample, focus- incidents that do occur escalate too 13 of the use ing on six central concerns: (a) abuse often to the level of excessive force. of authority and the use of force, (b) In trying to understand why, we asked of excessive the code of silence, (c) social factors, a series of questions that address the force. (d) departmental responses, (e) con- attitudes that police officers have trolling abuse, and (f) community toward the use of force and their policing. We then turn to comparisons perceptions of the behavior of their of subgroup responses according to re- fellow officers. gional variation, size of department, Our survey shows that most police supervisory status, racial variation, and officers in the United States disapprove gender variation. For example, are the of the use of excessive force. Nonethe- perceptions of white officers different less, a substantial minority believe that from those of African Americans or they should be permitted to use more other minorities? Does it matter if the force than the law currently permits, and officer is from the Midwest or the South? they consider it acceptable to sometimes

Police Foundation 24

Officers’ Attitudes Toward Limitations on Use of Force

It is sometimes acceptable Police to use more Police Police

Table 4.1 Table are not force than department officers permitted is legally rules about should be to use as allowable to the use of allowed to much force control some- force should use physical as is often one who not be any force in necessary physically stricter than response in making assaults an required to verbal arrests. officer. by law. abuse. (N=912) (N = 912) (N = 915) (N = 920)

Percent

Strongly Agree 6.2 3.3 6.7 0.4 Agree 24.9 21.2 58.9 6.6 Disagree 60.5 55.2 32.3 67.6 Strongly Disagree 8.4 20.3 2.1 25.4

…[T]he…majority use more force than permitted by the required by law. More than 65 percent laws that govern them. The officers re- of the sample officers were content that …did not believe vealed this attitude in their responses to police department rules about the use that officers several questions that were presented to of force not be stricter than required ...engage in them. More than 30 percent of the sample by law. Still, almost 35 percent did feel expressed the opinion that “police that departmental rules should be an excessive officers are not permitted to use as much stricter than required by law. And when use of force force as is often necessary in making asked whether police officers should on a regular arrests” (see Table 4.1). Almost 25 per- be allowed to use physical force in cent felt that it is sometimes acceptable response to verbal abuse, a very small basis. to use more force than legally allowable number, only 7 percent, thought that to control a person who physically as- this clear violation of current norms saults an officer. A very substantial mi- should be allowed. nority, more than 4 of 10, told us that Although a substantial minority ex- always following the rules is not com- pressed the view that the police should patible with getting the job done (see be permitted to use more force, the Appendix A, a19). overwhelming majority of the sample Most officers are not interested in hold- did not believe that officers do engage ing themselves to higher standards than in an excessive use of force on a

The Abuse of Police Authority 25

Officers’ Perceptions of Use of Force Behavior in Their Department

Police officers in your Police officers department in [city] use respond to more force than verbal abuse

necessary to with physical 4.2 Table make an arrest. force. (N = 922) (N = 922)

Sometimes, Often, or Always 196 (21.7%) 137 (14.7%) Seldom 581 (62.4%) 497 (53.5%) Never 145 (16.0%) 288 (31.8%)

regular basis. A mere 4 percent thought and only 16 percent reported that they …[O]fficers that police officers regularly used more never did so (see Table 4.2). Although do not believe physical force than was necessary in the large majority of respondents felt making arrests (see Appendix A, a10). that it is inappropriate to respond to ver- in protecting And almost everyone (97 percent) agreed bal abuse with physical force, almost wrongdoers; that serious cases of misconduct, such 15 percent thought that officers in their nevertheless, as the Rodney King case in Los Angeles department engaged in such behavior and the Abner Louima case in New sometimes (or often, or always). they often York, are “extremely rare” in their de- Code of Silence do not turn partments (see Appendix A, a40). Some of the most strongly held and them in. varied responses addressed the trou- Still, they did not give their fellow offic- bling area of whether officers should ers a completely clean report. When tell when they know that misconduct asked about their perceptions of the has occurred. The responses suggest behavior of officers in their own depart- the possibility of a large gap between ments, almost 22 percent of the weighted attitudes and behavior. That is, officers sample suggested that officers in their do not believe in protecting wrong- department sometimes (or often, or doers; nevertheless, they often do not always) use more force than necessary, turn them in.

Police Foundation 26

The survey shows that more than 80 their house because of you.” However, percent of American police do not they admitted that in individual cases, accept that the code of silence is an it is very difficult to betray fellow essential part of the mutual trust nec- officers even when those officers are essary to achieve good policing (see involved in criminal matters. Table 4.3). However, about a quarter In the survey we presented several sce- of the sample told us that whistle blow- narios involving misconduct, and we ing is not worth it, and more than two- asked the officers a series of questions thirds reported that police officers were …[M]ore than about the seriousness of the conduct, likely to be given a “cold shoulder” by 80 percent the consequences that should and fellow officers if they reported incidents would follow that conduct, and of American of misconduct. A majority felt that it whether they or others in the depart- was not unusual for police officers to police do not ment would report such conduct. In turn a “blind eye” to improper conduct accept that the one scenario, “An officer has a hand- by other officers (see Table 4.4). Even cuffed suspect sitting at his desk while code of silence when it came to reporting serious crimi- he fills out the necessary paperwork. nal violations, a surprising 6 in 10 is an essential With no provocation from the officer, report that police officers did not al- part of the the suspect suddenly spits in the face ways report serious criminal violations of the officer. The officer immediately mutual trust involving abuse of authority by fellow pushes the suspect in the face causing officers. necessary to the suspect to fall from the chair onto achieve good During the focus groups, officers re- the floor.” There was wide variation in policing. sisted the notion of a code of silence, perception of the offense’s seriousness, but agreed in the end that the code from 15 percent of the sample consid- stands except in the case of criminal ering it not serious at all to 16 percent violations. For instance, one supervi- considering it very serious (see Table sor suggested, “I don’t think there’s a 4.5). But would respondents report an code of silence at all when we are talk- officer who engaged in this behavior? ing about criminal conduct. And if it Only 3 in 10 stated that they would is, those people are part of a criminal definitely report. Even fewer, only mind.” Another said, “I think that the 11 percent of the sample, thought that wall of silence, as far as criminal things, most officers in their agency would is a thing of the past. I hear a lot of definitely report the offense. cops saying they are not going to lose

The Abuse of Police Authority 27

Code of Silence: Attitudes

An officer who reports another The code of officer’s mis- silence is an conduct is essential part likely to be

of the mutual given the 4.3 Table trust necessary Whistle blowing “cold shoulder” to good policing. is not worth it. by fellow officers. (N = 905) (N = 904) (N = 908)

Percent

Strongly Agree 1.2 3.1 11.0 Agree 15.7 21.8 56.4 Disagree 65.6 63.5 30.9 Strongly Disagree 17.5 11.7 1.8

Code of Silence: Perceptions of Behavior

It is not unusual for a police Police officers officer to turn always report a blind eye serious violations to improper involving abuse

conduct by of authority by 4.4 Table other officers. fellow officers (N = 908) (N = 899)

Percent

Strongly Agree 1.8 2.8 Agree 50.6 36.2 Disagree 43.3 58.5 Strongly Disagree 4.4 2.5

Police Foundation 28

Scenario of an Unruly Suspect: “An officer has a handcuffed suspect at his desk while he fills out the necessary paperwork. With no provocation from the officer, the suspect suddenly spits in the face of the officer. The officer immediately pushes the suspect in the face causing the suspect to fall from the chair onto the floor.”

Table 4.5 Table How serious do you consider the officer’s behavior to be? (N = 914)

Very Serious 135 (15.6%) Quite Serious 188 (20.2%) Moderately Serious 249 (27.3%) Not Very Serious 201 (21.7%) Not Serious at All 141 (15.3%)

Do you think you would report a fellow officer who engaged in this behavior? (N = 914)

Definitely Yes 262 (28.9%) Possibly Yes 207 (22.6%) Probably Not 254 (27.7%) Definitely Not 191 (20.8%)

Do you think most officers in your agency would report a fellow officer who engaged in this behavior? (N = 908)

Definitely Yes 94 (10.8%) Possibly Yes 270 (29.7%) Probably Not 378 (41.3%) Definitely Not 166 (18.3%)

The Abuse of Police Authority 29

Perceptions of the Effects of Extra-Legal Factors on Police Behavior

Police officers are Police A police more likely to officers are officer is Police use physical more likely to more likely officers force against use physical

to arrest a often treat African force against 4.6 Table person who whites better Americans poor people displays what than they and other than against he or she do African minorities middle-class considers Americans than against people in to be a and other whites in sim- similar bad attitude. minorities. ilar situations. situations. (N=917) (N = 914) (N = 916) (N = 918)

Percent

Strongly Agree 2.1 1.2 1.7 1.9 Agree 46.7 15.8 9.4 12.2 Disagree 45.1 57.8 55.6 57.9 Strongly Disagree 6.1 25.2 33.3 27.9

Social Factors and will increase the likelihood of an …[P]olice…are The question of the role of extralegal arrest (Klinger, 1996; Lundman, 1996; almost evenly factors in law enforcement has long Worden and Shepard, 1996). Our sur- been a concern among criminologists. vey shows that police in the U.S. are divided in Although sociologists since the 1950s almost evenly divided in their opinions their opinions (Westley, 1953) have suggested that a of whether a police officer is more likely of whether a citizen’s demeanor affects police be- to arrest a person who displays what he havior, recently some authors have or she considers to be a bad attitude. police officer called into question the importance of Some 49 percent of the sample thought is more likely being “in contempt of cop” and have that a bad attitude could affect the likeli- to arrest a argued for a more precise definition of hood of arrest, while 51 percent dis- the term “demeanor,” one that limits agreed (see Table 4.6). person who its meaning to verbal behavior (Klinger, displays… 1994; Lundman, 1994). However, even Do other extralegal factors, such as a bad attitude. with a more careful definition of terms, whether citizens are African American the consensus seems to have returned or white, or poor or middle class, make to the view that a disrespectful or hos- a difference in the treatment they tile demeanor displayed by a citizen receive from the police? The crimino- will affect the police-citizen encounter logical literature is split on the extent

Police Foundation 30

Police Perceptions of the Public’s Attitude Toward the Police

The relationship between the Most people do police and not respect the citizens in [city] police. is very good.

Table 4.7 Table (N = 924) (N = 923)

Percent

Strongly Agree 5.6 18.7 Agree 19.1 69.4 Disagree 65.2 9.9 Strongly Disagree 10.1 2.1

…Eighty-eight to which race affects everyday polic- What were the police officers’ views percent… ing (Mastrofski, Parks, DeJong, and of how the public perceives the Worden, 1998), the likelihood of being police? More than 75 percent did not described the arrested (Tonry, 1995; Black and Reiss, feel that “most people do not respect relationship 1970; Lundman, Sykes, and Clark, 1978; the police” (see Table 4.7). Put more between the Smith and Visher, 1981; Smith, Visher, positively, more than 75 percent of and Davidson, 1984; Worden, 1996; officers felt that most people respect the police and Lundman, 1996), and the use of exces- police. Indeed, 88 percent of police in the citizens sive force (Adams, 1996; Worden, 1996; our sample described the relationship in their Reiss, 1971; Walker, Spohn, and between the police and the citizens in DeLone, 1996; Ogletree, Prosser, Smith, their locality as very good. However, locality as and Talley, 1995). According to our more than half of our sample thought very good. sample, almost 2 in 10 police officers that the “public is too concerned with in the U.S. believe that whites are police brutality” (see Table 4.8), and treated better than African Americans more than 80 percent of police officers and other minorities (see Table 4.6). told us that the newspapers and TV in More than 1 in 10 said that there is this country are too concerned with more police violence against African police brutality (see Table 4.8). As one 14 Americans than against whites. More- officer in the focus group of police over, 14 percent of the sample believed supervisors noted in regard to the that police use physical force against poor media, “They are absolutely ruthless people more often than against middle- when it comes to police officers.” class people in similar situations.

The Abuse of Police Authority 31

Perceptions of Media and Citizens’ Concern Toward Police Abuse

The newspapers and TV in this The public is too country are too concerned with concerned with police brutality. police brutality.

(N = 918) (N = 920) 4.8 Table

Percent

Strongly Agree 13.4 36.0 Agree 41.6 44.2 Disagree 42.5 19.0 Strongly Disagree 2.5 0.8

Supervisors and rank-and-file officers ior by police (see Table 4.9). And they Supervisors alike complained that they are judged overwhelmingly (94 percent) disagreed on the sensational misdeeds of offic- with the suggestion that investigations and rank- ers from cities far away from their own. of police misconduct are usually biased and file… As one said, “We’re judged on Rodney in favor of the police. complained… King, Fuhrman.” Another officer put it When asked about the effectiveness of they are this way, “And as far as the Detroit deal, different institutional procedures for yeah, we caught heat behind that; L.A., addressing abuses of authority, most judged on the we caught heat behind that; and New people considered internal affairs units sensational York, yeah, we caught heat behind effective (79 percent), while a much misdeeds of that.” Still another presented the mi- smaller percentage (38 percent) con- nority view that the media do treat them sidered citizen review boards an effec- officers from fairly, “Our department has a great deal tive means for preventing police mis- cities far of credibility and respect from the conduct. This preference for internal away from media.” review was consistent with views expressed during the focus groups. One their own. Departmental Response rank-and-file officer argued that law- We polled the officers for their views yers and doctors police themselves so of how their departments handle cases why shouldn’t police, “Who is on the of abuse of authority. Officers in the bar association? Who is on doctors’ as- sample overwhelmingly (93 percent) sociations? Doctors judging doctors; reported that their departments take a doctors policing doctors. We are very tough stance on improper behav- special[ists]; we’ve got training; we deal

Police Foundation 32

Departmental Responses to Abuse of Authority

Your police Investiga- Internal Citizen department tions of affairs units review takes a police are not boards are very tough misconduct effective effective stance on are usually means for means for

Table 4.9 Table improper biased preventing preventing behavior in favor police police by police. of police. misconduct. misconduct. (N = 921) (N = 914) (N = 910) (N = 872)

Percent

Strongly Agree 35.2 0.4 2.4 3.1 Agree 57.4 5.1 19.0 34.7 Disagree 6.6 72.4 66.2 48.4 Strongly Disagree 0.9 22.0 12.4 13.9

Eighty-five with other people just like them. Why authority? American police overwhelm- are we different?” One supervisor sug- ingly told us that leadership makes a percent…said… gested, “Internal affairs works. Civilian difference. Eighty-five percent of the a police chief’s review authority—as soon as you men- officers said that a police chief’s tak- …strong position tion civilian review, the knee-jerk re- ing a strong position against abuses of authority can make a big difference in against abuses action is no way; yadda yadda, they go on and on. If they only knew, civil- preventing officers from abusing their …can make a ian review authority is nothing more authority (see Table 4.10). Policing big difference than a toothless tiger. They’re easier on scholars have long recognized the importance of the chief’s role. Skolnick in preventing cops than the departments are themselves. Bottom line.” Another supervi- and Fyfe (1993, p. 136) for example officers from sor agreed, “I think internal affairs is argue, “[T]he chief is the main abusing their more threatening because we’re police architect of police officers’ street behavior. This is so because the strength authority officers. We’ve all been out there. So we know how to play the game.” and direction of street-level police peer pressures ultimately are determined by Controlling Abuse administrative definitions of good and Can leadership make a difference in bad policing and by the general tone preventing police officers’ abuse of that comes down from the top.”

The Abuse of Police Authority 33

The Role of Supervision in Controlling Abuse

If a police chief takes a strong position against abuses of authority, he or she can Most police

Table 4.10 Table make a big Good first-line abuse of force difference in supervisors could be stopped preventing can help prevent by developing officers from police officers more effective abusing their from abusing methods of authority. their authority. supervision. (N = 920) (N = 921) (N = 913)

Percent

Strongly Agree 24.5 22.9 7.3 Agree 60.3 66.9 48.0 Disagree 13.8 9.3 39.5 Strongly Disagree 1.4 0.9 5.2

Elsewhere, Skolnick and Bayley (1986, pervisors can help prevent police offic- It is the p. 220) suggest that executive leader- ers from abusing their authority (see supervisor ship might be even more important in Table 4.10). As an officer who partici- police departments, with their “tradi- pated in one of the focus groups ex- as “role tional paramilitary character,” than in pressed it, “The supervisor, the first-line model” who other organizations: “Police depart- supervisor, the sergeant, is so critically surfaces as ments are not democratically run orga- important in how he sets the tone, the nizations. Everyone within them is ei- expectations. How he says things and the critical ther aware or attuned to the chief’s supports department programs or doesn’t aspect in preferences, demands, and expecta- support them. If not by what he says, good first-line tions.” Skolnick and Bayley (1986, then by body language and tone of voice. p. 6) argue, “[A]dministrative leadership, How he sells it or doesn’t sell it. That leadership. an animating philosophy of values, can sort of thing, I think, is real.” It is the indeed effect change.” supervisor as “role model” who surfaces as the critical aspect in good first-line As important as the role of the chief may leadership. Following this, 55 percent be in preventing abuse, an even greater of those surveyed thought that develop- majority—90 percent of police in the ing more effective means of super- sample—told us that good first-line su- vision would prevent abuse of force.

Police Foundation 34

Officers’ Perceptions of the Effects of Training on Abuse of Authority

Do you think Do you think interpersonal Do you think ethics in law skills or interper- human diversity enforcement sonal relations or cultural aware- training is training is ness training effective in effective in is effective in

Table 4.11 Table preventing abuse preventing abuse preventing abuse of authority? of authority? of authority? (N = 576) (N = 674) (N = 807)

Yes 472 (82.2%) 544 (80.3%) 603 (74.9%) No 104 (17.8%) 130 (19.7%) 204 (25.1%)

…[O]fficers In the focus group of chiefs of police of supervision was “conducive to who have and policing scholars, concern was ex- corruption.” The opinions expressed by pressed over the changing role of the the officers in our survey—about the dif- received supervisor in community policing. ference that good supervisors can make training in Chief Jerry Sanders of San Diego sug- in controlling abuse of authority—sug- ethics, in gested that by creating “teams” and gest that such concern by police execu- reducing “spans of control all of a sud- tives and academics is well placed. They interpersonal den we find the sergeants are closer to believed that good supervision matters. the team members, the officers, than skills, and in Contrary to the traditional view that they are to the department. They are cultural most important policing lessons are so close to the people on the team that obtained through experience in the sensitivity it creates problems.” field and not in the academy (Bayley report…such… and Bittner, 1984), scholars and police training can Commissioner Thomas Frazier of Bal- professionals have recently emphasized play a role in timore and Chief Jerry Sanders agreed the importance of changing models of that the management dynamics of the police training. This has led to a controlling department had been changed, and a renewal of commitment to training ef- abuses…. lieutenant with 24-hour responsibility forts and to exploring vastly different might not see his or her sergeants for a training curricula (e.g., see Grant and week or two at a time. Professor Carl Grant, 1996; Scrivner, 1994; Goldstein, Klockars suggested that community 1979; Trojanowicz and Bucqueroux, policing officers operate independently, 1994). The good news is that police almost without supervision, and Profes- officers who have received training in sor Alfred Slocum suggested that the lack ethics, in interpersonal skills, and in

The Abuse of Police Authority 35

The Community-Police Partnership

Citizens can be a vital source of Police should work Police should make information about with citizens to frequent informal the problems try and solve contact with in their problems on their people on their neighborhood. beat. beat. (N = 924) (N = 924) (N = 921) 4.12 Table

Percent

Strongly Agree 79.1 65.1 56.4 Agree 20.7 34.3 42.0 Disagree 0.1 0.4 1.2 Strongly Disagree 0.1 0.2 0.3 cultural sensitivity report that such policing idea into policing in the U.S. …[P]olice specialized training can play a role in The survey shows that police today today over- controlling abuses of police authority. overwhelmingly support a philosophy that looks to the public for advice and whelmingly A substantial majority (82 percent) of cooperation. Several statements formed support a those officers in the sample who have a group designed to measure officers’ philosophy received training in law enforcement opinions of the police-community ethics either in the academy or since partnership that is generally considered that looks becoming a police officer told us that a necessary component of community- to the public such training prevents abuse of author- oriented policing. Respondents over- for advice and ity (see Table 4.11). A similar majority whelmingly agreed that working with (80 percent) who have received police citizens was an important and effec- cooperation. training in interpersonal skills or tive means of solving neighborhood interpersonal relations felt that this problems. For example, nearly all training prevents abuse of authority. agreed that “[c]itizens can be a vital And 75 percent of officers who reported source of information about the prob- receiving training in human diversity, lems in their neighborhood,” that cultural differences, cultural awareness, “[p]olice should work with citizens to or ethnic sensitivity said that this try and solve problems on their beat,” training prevents abuse of authority. and that “[p]olice should make frequent informal contact with people on their Community-Oriented beat” (see Table 4.12). Policing The study provides strong evidence of But what of the relationship between the penetration of the community community policing and abuse of au-

Police Foundation 36

Perceptions of the Effects of Community Policing on Abuse of Authority

Do you think Do you think Do you think that community that community that community policing increases, policing increases, policing increases, decreases, or decreases, or decreases, or has no impact has no impact has no impact on the risk on the number on the seriousness

Table 4.13 Table of corrupt of excessive of excessive behavior force incidents? force incidents? (N=883) (N = 885) (N = 884)

Increases 63 (7.1%) 17 (2.0%) 32 (3.4%) Decreases 316 (35.8%) 450 (50.9%) 373 (42.2%) Has no impact 504 (57.1%) 418 (47.1%) 479 (54.4%)

…[A] close thority? Police in our sample generally 7 percent of the officers told us they indicated that a close relationship with thought community policing increases relationship the community does not increase the the risk of corruption. Over a third with the risk of police corruption. We asked this thought it decreases the risk of corrup- community question in two ways. Without referring tion, and another 57 percent thought it to community policing, we asked all of- had no effect (see Table 4.13). does not ficers whether they agreed that “[f]requent Some scholars have suggested that com- increase the friendly contact with local residents and munity policing may decrease the likeli- merchants increases the likelihood that risk of police hood of gross forms of corruption, such police officers will accept free lunches, corruption. as extortion, while increasing the temp- discounts, or gifts of appreciation for ef- tations toward softer forms of corrup- fective service” (see Appendix A, a34). tion, such as the free lunch, the “profes- Although one in five officers agreed with sional” discount, or the gift of apprecia- the statement, almost 80 percent tion for effective service (Weisburd, disagreed. Almost all the officers in the McElroy, and Hardyman, 1988). Others survey were familiar with the concept of suggest that community policing has no community-oriented policing (98 per- discernible impact on corrupt behavior cent). We asked those officers whether (McElroy, Cosgrove, and Sadd, 1990). they thought that community policing increases, decreases, or has no impact In the focus groups, there was consen- on the risk of corrupt behavior. Only sus among the officers that community

The Abuse of Police Authority 37 policing does not lead to corrupt community policing.” behavior, and there was concern among Chief Jerry Sanders of San Diego said, officers that chiefs are inappropriately concerned about this possibility. One I think it’s just much more subtle now supervisor explained why community than it was before. And it’s hard to policing is not a return to the day of the talk in those shades because the of- corrupt beat officer: ficers get invited to dinner at people’s houses because they create friend- But I think we’re in a different day ships. The friendships are created, and age, and I’m not so sure we’re …[C]ommunity which is what we’re trying to do. And going to get community-oriented policing does when is it not? And when is it a gratu- policing to lead us into the corruption ity to go into a friend’s business and not lead that we saw back then, and the rea- get a cup of coffee and when is it not? son why I think [so] is we’ve had things to corrupt I mean, I just think these are like Rodney King and what’s happened behavior, and… really difficult issues for not only the in Chicago and what happened there police officers but for police manage- chiefs are and what’s described as happening in ment. Where do we draw the line? Is many cities. I think there is a different inappropriately it, as O. W. Wilson said, “The first cup emphasis on morality and ethics in law concerned of coffee you take for free is the start enforcement than we saw back 40, 50 of corruption,” or is it we need to be a about this years ago. I don’t think even the pub- little bit more understanding about the lic has a tolerance for the corruption possibility. motives that we’re talking about? that was a fact of daily life in New York 50 years ago. While such concerns are expressed by police scholars and executives, they are In the panel of police scholars and not seen as significant by the vast executives that we convened at the majority of American police. beginning of the study, concern was expressed about the potential for cor- What do officers think is the relation- ruption under community policing. As ship between community policing and Baltimore’s Commissioner Thomas excessive force? Almost no one told us Frazier said, “One of the things that that community policing would increase troubles me about community polic- the amount (2 percent) or seriousness ing is you talk about establishing rela- (3 percent) of excessive force incidents tionships. The longer the relationship (see Table 4.13). A majority said exists, I think the more opportunity for that community policing decreases corruption.” Professor Klockars pointed the incidents of excessive force out the irony of some situations, “So if (51 percent), and 42 percent thought it you run a McDonald’s and you give a would decrease the seriousness of ex- cop a free meal, that’s corruption. cessive force incidents. Many thought it But if you give a whole booth, that’s had no impact on either the amount

Police Foundation 38

of excessive force (47 percent) or the them against the wall and searches seriousness of excessive force incidents them. Finding nothing, the officer uses (54 percent). demeaning language, then tells them that this “will teach you to respect the The community policing partnership law” and “I’d better not see you here can be complicated. Almost all officers again” and gets in his patrol car and (97 percent) told us that “[p]olice offic- drives off. ers sometimes have to explain to individuals and groups of citizens that the In the other version (see p. 63), the “…[P]olice police are prohibited by law from us- other half (482) of the officers officers ing some of the tactics that citizens responded to a set of questions based sometimes encourage them to use” (see Appendix A, on the following scenario: a32). But 21 percent felt that they could In a community meeting, citizens told have to explain use more aggressive tactics than they police that they were very concerned otherwise would if the community had to individuals about groups of rowdy youths hang- asked them to do so (see Appendix A, and groups ing out on street corners. After the a33). Whether they might sometimes meeting, an officer who participated of citizens that cross the line to tactics prohibited by in the meeting notices several youths law remained unanswered. the police are standing on a corner smoking ciga- prohibited We presented the officers with one of rettes and talking to one another. The by law from two versions of a scenario that ad- officer tells the youths to break it up dressed, among other issues, whether and leave the area. The youths say, using some they would feel justified in using more “We’re not doing anything. Why are of the tactics aggressive tactics if asked by the com- you hassling us?” The officer gets out that citizens munity (see p. 61). In one version, a of the car and orders the youths to randomly assigned half (438) of the of- place their hands up against the wall encourage ficers responded to a set of questions of a building. They refuse. The officer them to use.” based on the following scenario: throws them against the wall and searches them. Finding nothing, the While patrolling his beat, an officer officer uses demeaning language, tells notices several youths standing on a them that this “will teach you to re- corner smoking cigarettes and talking spect the law” and “I’d better not see to one another. The officer tells the you here again,” and gets in his youths to break it up and leave the patrol car and drives off. area. The youths say, “We’re not doing anything. Why are you hassling With these scenarios, we could cap- us?” The officer gets out of the car ture whether officers felt justified in and orders the youths to place their taking certain questionable actions hands up against the wall of a build- when they had been asked by the com- ing. They refuse. The officer throws munity to do so. Interestingly, the

The Abuse of Police Authority 39 answers of the two randomly assigned survey were among police officers of groups of officers were virtually iden- different racial groups. Although we tical to the series of questions that fol- originally grouped the officers in two lowed the scenarios. Most officers told categories (white and non-white) so we us that a verbal or written reprimand could have larger numbers in each cat- would and should follow such an inci- egory, when strong differences accord- dent. A substantial minority thought ing to race emerged, we re-examined the discipline would and should be the data, peeling back the non-white suspension without pay. Slightly more category into two subcategories: blacks …[T]here is a than one in three said they definitely or African Americans, and other minor- racial divide would report a fellow officer who ity officers. In so doing, the significance engaged in this behavior, whereas only of the results increased, indicating that between 1 in 10 believed that most officers in African-American officers hold the most whites and their agency would report such an in- distinctive positions on these issues. African cident. These results suggest that Without meaning to overstate the police officers do not feel justified in generalizability of our findings beyond Americans in using more aggressive tactics if asked American policing, the survey tends to our society by the community to do so. corroborate the view that there is a that is not racial divide between whites and Afri- Subgroup Analysis: can Americans in our society that is not transcended Race, Rank, Region, transcended even by a culture as ap- even by a parently strong as the culture of polic- Agency Size, Gender culture as… Thus far, we have described what the ing. Not that those differences emerged survey suggests about the attitudes of across every item in our survey, but strong as the police generally toward abuse of au- when they did occur, the relationships culture of were strong, and the kinds of ques- thority. But the data can also reveal policing. something about how different sub- tions in which they emerged grouped groups within American policing view together in meaningful configurations. such issues. An analysis of subgroup differences is presented in cross-tabu- Earlier we reported that almost 2 in 10 lations below. In reporting on differ- officers in the weighted sample agreed ences in responses among different that police officers often treat whites subgroups of police officers, we note better than they do African Americans again that our statistics were adjusted and other minorities. When we consid- according to the sampling procedures ered this issue broken down by race, we used. we found that more than half of the Af- rican-American officers felt this way (see 15 Race Table 4.14). By comparison, fewer than By far the most striking differences we one in four among other minorities discovered among subgroups in our agreed with the statement, and fewer

Police Foundation 40

Police officers often treat whites better than they do African Americans and other minorities (by race). (N = 912)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.14 Table White Officers 0.7 11.2 60.5 27.7 African-American Officers 4.6 46.7 39.8 8.9 Other Minority Officers 2.4 21.0 53.8 22.9

2 x = 41.78 df = 6 p < .001

16 African- than one in eight white officers agreed. determined only by race. While only 2 percent of white officers in the American The divergence in views of African sample thought that police officers officers Americans and other officers continues were more likely to use physical did not see and grows when we examine whether force against poor people than they felt that police officers were more against middle-class people in simi- unequal likely to use physical force against Af- lar situations, 54 percent of the Afri- treatment rican Americans and other minorities can-American officers felt that way by police as than against whites in similar situations. (see Table 4.16). Again, other minori- While only 1 in 20 white officers in ties held a position between the determined the sample thought that African Ameri- white and African-American officers, only by race. cans and minorities received this but closer to the perspective of the 18 unfair treatment from police, well over white officers. half of the African-American officers thought unfair treatment was more While the survey suggests that African- likely. Other minorities were more in American officers may not trust their agreement with the white officers (see fellow officers to treat minority and poor 17 Table 4.15). citizens fairly, they did tend to respond more positively to the role of commu- African-American officers did not see nity policing in reducing police abuses unequal treatment by police as of authority. For example, we

The Abuse of Police Authority 41

Police officers are more likely to use physical force against African Americans and other minorities than against whites in similar situations (by race). (N = 914)

Strongly Strongly Agree Agree Disagree Disagree

Table 4.15 Table Percent

White Officers 0.6 4.5 58.0 37.0 African-American Officers 9.4 47.7 42.1 0.9 Other Minority Officers 2.4 10.0 50.7 36.9

2 x = 86.80 df = 6 p < .001

Police officers are more likely to use physical force against poor people than against middle-class people in similar situations (by race). (N = 916)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.16 Table White Officers 0.8 8.0 60.1 31.1 African-American Officers 9.1 45.3 43.6 2.0 Other Minority Officers 4.2 13.0 52.9 30.0

2 x = 85.42 df = 6 p < .001

Community-oriented policing increases, decreases, or has no impact on the number of incidents of excessive force (by race). (N = 883)

Has No Increases Decreases Impact

Percent

Table 4.17 Table White Officers 1.2 49.2 49.6 African-American Officers 6.6 65.4 28.1 Other Minority Officers 3.9 50.1 46.0

2 x = 20.92 df = 4 p < .001

Police Foundation 42

Community-oriented policing increases, decreases, or has no impact on the seriousness of excessive force incidents (by race). (N = 882)

Has No Increases Decreases Impact

Percent

Table 4.18 Table White Officers 3.2 39.0 57.9 African-American Officers 7.2 63.4 29.3 Other Minority Officers 1.0 46.8 52.3

2 x = 27.13 df = 4 p < .001

…[W]e found a statistically significant relation- duct. Almost 7 in 10 African-American ship between race and support for the officers in the sample believed in the found a view that community-oriented policing effectiveness of citizen review, …relationship decreases the number of incidents of compared with one-third of white of- 19 between race excessive force (see Table 4.17). ficers (see Table 4.19). For such rela- Although fewer than half of white tionships, other minority officers once and support officers believed this to be the case, again fell somewhere between African- for the view almost two-thirds of the African-Ameri- American and white police officers. can police officers surveyed agreed that As we continue to discuss relationships with this position. African-American community- among other subgroups in the weighted police officers are also more likely to sample, it will become clear that—while oriented say that community policing decreased other interesting differences occur— the seriousness of incidents of exces- policing no differences were as large as those sive force (see Table 4.18). Among decreases found among these racial groups. African-American police officers, 63 the number of percent expressed this view, as con- incidents of trasted with only 39 percent of white Rank: Supervisors and police officers. Finally, African-Ameri- excessive Nonsupervisors can officers also had more faith in While most officers in the sample— force citizen review boards as an effective those who were supervisors and those means for preventing police miscon- who were not—believed in the impor-

The Abuse of Police Authority 43

Citizen review boards are effective means for preventing police misconduct (by race). (N = 868)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.19 Table White Officers 2.5 30.8 52.2 14.6 African-American Officers 8.4 61.4 22.3 7.9 Other Minority Officers 2.4 38.9 43.6 15.1

2 x = 32.04 df = 6 p < .001

Good first-line supervisors can help prevent police officers from abusing their authority (by rank). (N = 921)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.20 Table Nonsupervisors 16.5 70.2 12.4 0.9 Supervisors 38.5 58.8 1.9 0.8

2 x = 76.12 df = 3 p < .001

If a police chief takes a strong position against abuses of authority, he or she can make a big difference in preventing officers from abusing their authority (by rank). (N = 920)

Strongly Strongly Agree Agree Disagree Disagree

Table 4.21 Table Percent

Nonsupervisors 18.3 62.6 17.4 1.6 Supervisors 39.6 54.6 5.0 0.8

2 x = 71.15 df = 3 p < .001

Police Foundation 44

Most police abuse of force could be stopped by developing more effective methods of supervision (by rank). (N = 913)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.22 Table Nonsupervisors 6.3 43.7 44.7 5.4 Supervisors 9.9 58.5 26.7 4.9

2 x = 33.01 df = 3 p <. 001

…[G]ood tance of supervision to good policing, could be stopped by developing more that belief was particularly strong effective methods of supervision (see first-line 22 among the supervisors themselves. Table 4.22). supervisors Thus, while more than 87 percent of Still in keeping with their role as su- could help nonsupervisors in the survey (primarily pervisors, but less predictable, was a patrol officers) said that good first-line prevent police series of questions that suggested that supervisors could help prevent police supervisors were very serious in their officers from officers from abusing their authority, 97 attitudes about reporting misbehavior percent of supervisors felt that way (see abusing their 20 and that they held police officers to a Table 4.20). Similar relationships are authority. very high standard. Well over 80 per- found in other questions directly related cent of supervisors believed in the to supervision. More than 80 percent of value of blowing the whistle on mis- nonsupervisors and almost 95 percent behavior by fellow officers, com- of supervisors believed that if a police pared with just over 70 percent of 23 chief took a strong position against nonsupervisors (see Table 4.23). abuses of authority, he or she could make Similarly, supervisors were much less a big difference in preventing officers likely to believe in the efficacy of the from abusing their authority (see Table 24 21 code of silence (see Table 4.24), and 4.21). And 50 percent of nonsupervisors supervisors disagreed to a much and 68 percent of supervisors were likely greater extent than non-super- to believe that most police abuse of force visors that it is sometimes accept

The Abuse of Police Authority 45

Whistle blowing is not worth it (by rank). (N = 904)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Nonsupervisors 3.9 24.4 61.4 10.3 4.23 Table Supervisors 1.1 15.6 68.4 15.0

2 x = 24.99 df = 3 p < .001

The code of silence is an essential part of the mutual trust necessary to good policing (by rank). (N = 905)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.24 Table Nonsupervisors 1.5 19.2 64.2 15.1 Supervisors 0.3 7.3 68.8 23.5

2 x = 28.46 df = 3 p < .001

It is sometimes acceptable to use more force than is legally allowable to control someone who physically assaults an officer (by rank). (N = 912)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.25 Table Nonsupervisors 3.9 23.9 54.5 17.7 Supervisors 1.8 14.6 56.8 26.9

2 x = 21.09 df = 3 p < .001

Police Foundation 46

Police department rules about the use of force should not be any stricter than required by law (by rank). (N = 915)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.26 Table Nonsupervisors 7.4 62.8 28.6 1.2 Supervisors 4.9 49.6 41.4 4.1

2 x = 24.90 df = 3 p < .001

…[W]e did able to use more force than is legally one officer explained that in her part allowable to control someone who of the country, it would be considered find a physically assaults an officer (see Table an affront if a community policing 25 consistent 4.25). They were more interested than officer refused to accept an offer of a difference nonsupervisors in having departmen- cup of coffee. Those regional cultural tal rules about the use of force that are differences might explain why officers between the stricter than required by law (see Table from the southern region of the coun- 26 western 4.26). Supervisors in our sample were try were more likely to offer an opin- region of the also more likely to note that commu- ion that frequent friendly contact with nity-oriented policing could decrease local residents and merchants increased country and the number and the seriousness of ex- the likelihood that police officers would 27 others…. cessive force incidents (see Table 4.27 accept free lunches, discounts, or gifts 28 and Table 4.28). of appreciation for effective service (see 29 Table 4.29). Region While regional differences did not show As Professor Carl Klockars stated dur- up as clearly as one might have expected ing the focus group of police scholars for many questions in the survey, and executives at the outset of this we did find a consistent difference study, “There are right answers in dif- between the western region of the coun- ferent places.” During an untranscribed try and others on some specific break in our rank-and-file focus group, indicators. Police officers from the

The Abuse of Police Authority 47

Community-oriented policing increases, decreases, or has no impact on the number of excessive force incidents (by rank). (N = 885)

Has No Increases Decreases Impact TOTAL

Nonsupervisors 14 (2.3%) 285 (45.7%) 323 (52.1%) 622

Table 4.27 Table Supervisors 3 (1.4%) 165 (63.5%) 95 (35.1%) 263

2 x = 22.91 df = 2 p < .001

Community-oriented policing increases, decreases, or has no impact on the seriousness of excessive force incidents (by rank). (N = 884)

Has No Increases Decreases Impact TOTAL

Nonsupervisors 22 (3.3%) 244 (38.6%) 358 (58.1%) 624

Table 4.28 Table Supervisors 10 (3.7%) 129 (50.9%) 121 (45.4%) 260

2 x = 10.96 df = 2 p < .05

Frequent friendly contact with local residents and merchants increases the likelihood that police officers will accept free lunches, discounts, or gifts of appreciation for effective service (by region). (N = 916)

Strongly Strongly Agree Agree Disagree Disagree

Table 4.29 Table Percent

North Central 1.0 18.2 60.4 20.3 Northeast 0.4 14.1 67.8 17.8 South 2.8 25.2 55.4 16.7 West 0.2 15.6 58.1 25.5

2 x = 23.08 df = 9 p < .001

Police Foundation 48

It is sometimes acceptable to use more force than is legally allowable to control someone who physically assaults an officer (by region). (N = 912)

Agree Disagree TOTAL

Table 4.30 Table North Central 40 (20.4%) 149 (79.6%) 189 Northeast 67 (29.4%) 163 (70.7%) 230 South 82 (26.8%) 229 (73.2%) 311 West 32 (17.7%) 150 (82.3%) 182

2 x = 8.76 df = 3 p < .05

Officers West tended to have a more profes- ence behavior. Officers in the largest sional outlook about policing or tended departments more frequently demon- from small to view policing as constrained by the strated what arguably was a more departments law. Officers in the West were less likely cynical or alienated attitude about lead- believed more to agree that it is sometimes accept- ership in policing, although even in able to use more force than is legally those departments it was distinctly a strongly in allowable to control someone who minority viewpoint. The relationships the efficacy physically assaults an officer (see Table described were quite strong. 30 of good 4.30). And they are more likely to state While 94 percent of officers from small that police officers always report seri- supervision departments believed that if a police ous criminal violations involving abuse chief took a strong position against and in the of authority by fellow officers (see 31 abuses of authority, he or she could authority Table 4.31). make a big difference in preventing of the chief officers from abusing their authority, Agency Size only 68.2 percent of officers from the to influence Agency size also appeared relevant to largest departments agreed (see Table 32 behavior. police officers’ concerns about police 4.32). Similarly, while 97 percent of abuse of authority. Officers from small officers from small departments agreed departments believed more strongly in that good first-line supervisors could the efficacy of good supervision and help prevent officers from abusing their in the authority of the chief to influ- authority, only 80 percent of officers

The Abuse of Police Authority 49

Police officers always report serious criminal violations involving abuse of authority by fellow officers (by region). (N = 899)

Agree Disagree TOTAL

North Central 63 (33.5%) 124 (66.5%) 187

Northeast 79 (35.1%) 145 (64.9%) 224 4.31 Table South 121 (40.0%) 189 (60.1%) 310 West 83 (48.6%) 95 (51.4%) 178

2 x = 9.81 df = 3 p < .05

If a police chief takes a strong position against abuses of authority, he or she can make a big difference in preventing officers from abusing their authority (by agency size). (N = 920)

Strongly Strongly Agree Agree Disagree Disagree

Table 4.32 Table Percent

Small Departments 37.6 56.2 6.2 0.0 Medium Departments 24.9 62.7 11.3 1.1 Large Departments 14.0 54.2 28.4 3.4

2 x = 54.37 df = 6 p < .001

Good first-line supervisors can help prevent officers from abusing their authority (by agency size). (N = 921)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.33 Table Small Departments 26.1 70.4 2.7 0.8 Medium Departments 23.8 67.4 8.4 0.5 Large Departments 17.5 62.5 17.5 2.6

2 x = 25.93 df = 6 p < .001

Police Foundation 50

Most police abuse of force could be stopped by developing more effective methods of supervision (by agency size). (N = 913)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.34 Table Small Departments 9.9 61.1 25.2 3.8 Medium Departments 7.2 48.0 40.5 4.4 Large Departments 6.2 39.2 45.1 9.5

2 x = 16.44 df = 6 p < .05

from the largest departments agreed tistically significant results were found, …[O]fficers 33 (see Table 4.33). Following this pat- the sizes of the differences were small, from the largest tern, 71 percent of officers from small and no consistent theory or idea linked departments departments agreed that most police them or suggested that they were believed… abuse of force could be stopped by meaningful. We could argue that this developing more effective methods of police finding suggested that women adapt supervision, compared with 45 percent to the dominant culture of policing or administrators of officers from the largest departments that women who self-select to enter 34 concentrate on (see Table 4.34). Almost 8 in 10 policing are more like men in polic- officers from the largest departments what police ing. However, this conclusion may be believed that police administrators con- premature. As the National Center for officers do centrate on what police officers do Women and Policing (1998) reports, wrong rather wrong rather than what they do right, “Women police perform better than while just over half of the officers from than what they their male counterparts at defusing the small departments agreed (see potentially violent situations and do right. 35 Table 4.35). become involved in excessive use of force incidents less often.” We think it Gender is possible that we did not ask ques- The survey did not reveal meaningful tions in our survey that would reveal differences in the responses of officers specific differences between male and according to gender. While several sta- female police officers in regard to the problem of police abuse of authority.

The Abuse of Police Authority 51

Police administrators concentrate on what police officers do wrong rather than what police officers do right (by agency size). (N = 920)

Strongly Strongly Agree Agree Disagree Disagree

Percent

Table 4.35 Table Small Departments 14.0 38.6 42.7 4.7 Medium Departments 17.4 48.0 31.9 2.7 Large Departments 29.2 49.6 20.6 0.7

2 x = 32.92 df = 6 p < .001

Police Foundation 52 V

CONCLUSION

Police…support Over the past three decades, American licing idea have filtered down to rank- a philosophy policing has gone through vast changes and-file police officers. Police today over- in its organization, tactics, and philoso- whelmingly support a philosophy that that looks to phy. At the forefront of such changes looks to the public for advice and coop- the public for has been the transition from traditional eration in confronting problems of crime advice and military and professional models of and disorder. The survey was unambigu- policing, to innovative models of ous in this regard. Nearly all who were cooperation community policing. Our survey surveyed believed that citizens are vital in confronting focused on the attitudes of American to good policing and that police must problems of police toward abuse of authority in this work with citizens in solving crime prob- age of community policing. It is lems. In turn, contrary to what some have crime and the first truly representative study of feared about community policing, offic- disorder. police attitudes in many years, and thus ers were more likely to state that com- it provides an important set of findings munity policing reduces the potential for for understanding American police and police abuse than increases that poten- for developing public policy. tial. This finding, of course, does not mean that community policing has actu- It is clear from the survey that central ally reduced the level of abuse in components of the community po-

The Abuse of Police Authority 53

American policing, but rather that given a “cold shoulder” by fellow police officers believe this to be the case. officers if they reported incidents of misconduct. Most police officers in the Our portrait of attitudes toward police study reported that it is not unusual abuse is much more ambiguous. On for police officers to turn a “blind eye” the one hand, we have much positive to improper conduct by other officers. evidence regarding the attitudes of American police officers and their views These findings suggest that the culture about their colleagues. Most police we of silence, which has continually surveyed do not agree that it is accept- plagued reform in American policing, …[P]olice able to use more force than is legally continues. But it must be recognized officers necessary, even to control someone that from the perspective of police, the were likely who physically assaults an officer. The concern of the public with problems vast majority identified extreme brands of police abuse is not proportional to to be given a of police abuse such as that reported its incidence. The survey shows that “cold shoulder” in the Rodney King and Abner Louima most American police believe that the by fellow cases as being isolated and very rare public is too concerned with police occurrences. Most police surveyed told abuse. An even larger number believe officers if us that their police departments always that the media have paid too much they reported took a tough stand on the issue of attention to this question. From the incidents of police abuse of citizens. perspective of police, the public and the press have placed too much of their misconduct.… On the other hand, the survey suggests concern on police abuse. At the same that police abuse remains a problem [and] it is not time, the police we studied believed that must be addressed by policy mak- unusual for that the relationship between police ers and police professionals. While and community is a good one, and is police officers the survey suggests that most police one in which the community overall officers in the United States disapprove to turn a has respect for the police. of the use of excessive force, a sub- “blind eye” stantial minority consider it acceptable The survey suggests that race continues to improper to sometimes use more force than per- to be an issue for American police. One mitted by the laws that govern them. in five of those surveyed told us that conduct by They also believe that they should be whites are treated better by police than other officers. permitted to use more force than the African Americans and other minorities. law currently permits. The code of We cannot say whether this result rep- silence also remains a troubling issue resents a change in attitudes either in a for American police. It is still the case more positive or negative direction. that about a quarter of the police we However, we can conclude that a sub- surveyed told us that whistle blowing stantial number of police in the U.S. see is not worth it, and two-thirds reported race as an important factor in understand- that police officers were likely to be ing abuses of police authority.

Police Foundation 54

It is particularly troubling that this per- misbehavior. While those responses did ception was far more prevalent among not tell us whether indeed such pro- African-American police officers than grams are effective, they did tell us that among others. Comparing African- American police themselves view the American officers’ views about police programs as important and useful. abuse with those of white and other Police we surveyed also emphasized minority officers, we found significant the importance of police management and substantial differences. While a in preventing police violence and other small minority of white officers in the Comparing forms of abuse. A large majority of sample believed that police treat white African- police believe that when the chief of citizens better than African-American police takes a strong stand against American or other minority citizens in similar situ- police violence, other police officers ations, a majority of African-American officers’ views will follow his or her lead. Similarly, police officers held this view. Similar about police police officers told us that good first- differences were found between Afri- line supervision is an effective method abuse with can-American and other police for preventing police abuse. These find- officers in their views on the likelihood those of white ings reinforce the long-held view of of using physical force against minori- and other scholars and police professionals that de- ties and the poor. The magnitude of veloping effective methods of supervi- minority such race-based differences suggests a sion and effective supervisors should be large gap between African-American officers, a first priority in efforts to control and police officers and other officers. Such we found prevent abuses of police authority. a deep divide was not predicted at the significant outset of the study. While American police recognized the importance of supervision in prevent- and The survey also provides some surprising abuses, they continued to see a ing and important lessons regarding substantial tension between getting the police job how police think abuses of police done and controlling misbehavior. differences. authority can be controlled. Consistent Almost half of the police surveyed told with the suggestions of certain schol- us that “always following the rules” is ars and police professionals (Grant and not compatible with “getting the job Grant, 1996; Scrivner, 1994), most done.” More than half believed that su- officers believed that training and pervisors focus too much on what they education are effective methods for are doing wrong and not enough on reducing police abuse. A substantial what they are doing right. majority of those who have experienced training in interpersonal skills, Abuse of police authority continues to or have taken courses in ethics or be a major public policy concern. This diversity, said that such education and survey adds the voices of rank-and-file training is effective in preventing police and supervisors to the debate

The Abuse of Police Authority 55 over the nature of the problems that shields those who do abuse their …[P]olice American police face and the potential authority. What can be done to abuses…are a solutions that can be brought. Our prevent such abuses? According to study suggests that most police in the America’s police, education and train- continuing United States understand the impor- ing are effective means of preventing reality…as is tance of limits to police authority, and police abuse. They also recognize the the “code of are sensitive to the dangers of corrup- continuing importance of effective silence” that tion and abuse of force. Nonetheless, supervision, and the central role that police abuses of authority are a police executives play in sending shields those continuing reality in American polic- the message that police abuses of who do ing, as is the “code of silence” that authority cannot be tolerated. abuse their authority.

Police Foundation 56

ENDNOTES

1. Details regarding the sample are We also consulted a number of survey in- provided in Chapter II of this report. The struments that addressed police ethics. sampling frame of the survey as estimated They did not provide specific questions by Maguire (1997) is 351,480 officers. but assisted us in our thinking. The instru- ments included the Royal Barbados Police 2. A number of questions were Force survey by Richard R. Bennett (1994), adapted from two previous Police Foun- the Illinois Police Behavior Survey (Martin dation surveys (Pate and Hamilton, 1992; and Bensinger, 1994), an Australian sur- Wycoff and Oettmeier, 1993). We used sev- vey on police ethics (McConkey, Huon, eral questions from a survey on ethical and Frank, 1996), and a Police Founda- conduct and discipline from Queensland, tion survey of Oregon State Police officers Australia (Criminal Justice Commission, (Amendola, Hockman, and Scharf, 1996). 1995). We adopted the question structure (although not the content) for presenting We conducted three focus groups, which hypothetical scenarios from the survey in- provided a range of viewpoints from rank- strument on police integrity by Carl B. and-file officers to chiefs and academics: Klockars, Sanja Kutnjak-Ivkovic, and Will- The first group combined the knowledge iam E. Harver (no date). Scenarios were and experience of academic experts and provided by Chief Jerry Oliver during the police chiefs (for a full report, see Appen- expert focus group (see Appendix B) and dix B). The second group was composed by Earl Hamilton of the Police Foundation of rank-and-file police officers from depart- staff. Some of the demographic questions ments across the country (see Appendix C were developed with assistance from mem- for a full report, including selection crite- bers of Mathematica Policy Research Inc., ria). The third group consisted of sergeants who also assisted with editing the survey and lieutenants from departments nation- questions. wide (see Appendix D for a full report,

The Abuse of Police Authority 57 including selection criteria). In addition to the confidentiality of selected officers. In the important contributions of the police addition to restricting access to this infor- scholars, chiefs, and officers who partici- mation to key personnel at Mathematica, pated in our focus groups, as well as the the Police Foundation withheld the officers who participated in the pretest, we identity of selected officers from project consulted with a number of police research- personnel. ers as we developed the survey instrument. Professors Carl Klockars, Peter Manning, Interviewing staff members at Mathematica Stephen Mastrofski, Albert Reiss, Jerome signed a confidentiality pledge prior to the Skolnick, and Robert Worden provided surveying period. In signing this pledge, valuable criticism that impelled us to keep interviewers agreed not to divulge any pri- a sharp focus. vate, project-related information to any person not authorized to have access to such 3. Care was taken to protect the con- information. fidentiality and anonymity of all officers participating in the survey, from the de- Serial numbers were assigned to selected sign of the selection process through data officers before creating a machine-process- collection and analysis. Selected depart- ing record and identifiers (i.e., name, ad- ments were requested to provide to the dress, telephone number, etc.). Those num- Police Foundation a list of the names of all bers were not included in the machine their full-time, sworn personnel. From this record nor in the resulting database. Sur- roster, officers were randomly selected for vey data containing identifiers or potential participation. This methodology enabled us identifiers were kept secured and were de- to keep the names of the selected officers stroyed by Mathematica Policy Research confidential from their chief and other de- Inc. within 10 days of sending the data set partmental personnel. (There was some to the Police Foundation. variation in this procedure. In two cases, 4. We sought at the outset to gain an the department generated the random accurate list of all U.S. police officers. How- sample in their own computers in the pres- ever, attempts to acquire names of current ence of Police Foundation researchers. In police officers on a state-by-state basis from two others, departments provided serial state licensing boards, although promising numbers from which we generated the ran- at first, were ultimately frustrated and un- dom samples, and only then were we pro- successful. As a result, we turned to the vided names and contact information. Still, two-step process defined here. every effort was made to protect the privacy of the officers in each situation.) 5. The Northeastern region includes the states of Connecticut, Maine, Massa- All information received by the Police Foun- chusetts, New Hampshire, New Jersey, New dation from the departments, as well as in- York, Pennsylvania, Rhode Island, and terview and survey data, was kept in locked Vermont. The South includes Alabama, file cabinets. Access to such information was Arkansas, Delaware, District of Columbia, limited to key project personnel. Florida, Georgia, Kentucky, Louisiana, Under the terms of a subcontract, all infor- Maryland, Mississippi, North Carolina, Okla- mation furnished by the departments was homa, South Carolina, Tennessee, Texas, made available to Mathematica Policy Re- Virginia, and West Virginia. The North Cen- search Inc. for the selection of officers and tral region includes the states of Illinois, for subsequent interviews. Mathematica Indiana, Iowa, Kansas, Michigan, Minne- Policy Research Inc., in compliance with sota, Missouri, Nebraska, North Dakota, Police Foundation policy, agreed to main- Ohio, South Dakota, and Wisconsin. And tain strict procedures designed to protect the Western region includes Alaska, Ari-

Police Foundation 58

zona, California, Colorado, Hawaii, Idaho, The sample selected from those lists was Montana, Nevada, New Mexico, Oregon, of adequate size to allow completion of Utah, Washington, and Wyoming. 925 to 950 interviews. In certainty departments, the number of officers selected was 6. This method was developed in con- based on the proportion of the total popu- sultation with and conducted by John Hall, lation of officers represented by the de- senior sampling statistician with partment. For noncertainty departments Mathematica Policy Research Inc. with 25 or more officers, we sampled 10 per department. For the departments with The sample was selected using a stratified 10 to 24 officers, we sampled an average of multistage design. The sample initially in- 4.5 (a random half of the sample was allo- cluded 121 police departments that were cated 5 selections, and the other half, 4). selected by using probability proportional to size (PPS) methods. To use PPS selec- 7. As noted in endnote 3, there was tion, each department was assigned a mea- some variation in this procedure. In two sure of size based on an estimate of its cases, the departments generated the ran- number of full-time sworn officers. dom sample from their own computers in the presence of Police Foundation re- Departments that were so large that they searchers. In two others, departments pro- were certain to be sampled using PPS meth- vided serial numbers from which we gen- ods were selected with probability 1.0 and erated the random samples; only then were are called “certainty selections.” we provided names and contact information. Selection was made from a machine-read- able file (the sampling frame) that included 8. When the confidence interval ex- one record for each of the 5,014 eligible ceeds plus or minus 4 percent, we pro- police departments in the study. Each vide the exact interval in the endnotes. record contained department identification, department type, region, and the estimated 9. Because of the multistage sampling number of officers who were employed procedure used in our study, we could not by the department and were eligible for rely on standard estimates of standard er- the survey. The estimated number of offic- rors or statistical significance. Adjusted stan- ers was used as a measure of size in se- dard errors and observed significance lev- lecting the sample. els were estimated using the statistical analysis program Sudaan (see Babubhai, Before selection, the sample frame was Barnwell, and Bieler, 1997), after specify- stratified by size and region. The size cat- ing the specific sampling model used in egories were (a) certainty selections, (b) our study. other departments with 25 or more full- time sworn officers, and (c) other depart- 10. The weighting procedure was de- ments with 10–24 full-time sworn officers. veloped by John Hall of Mathematica Policy The sample included nine certainty selec- Research Inc. The weights for the Police tions: 84 from the middle-size group and Foundation Survey account for differences 28 from the smallest group. in (a) probabilities of selection among officers responding to the survey, (b) The Police Foundation contacted sampled nonresponse at the department level, and departments and, from each cooperating (c) response rates among groups of department, obtained a list of all officers officers. Without the weighting adjust- eligible for the survey. Those lists contained ments, some groups of officers would be identifying information that enabled inter- overrepresented (and others under- viewers to contact sampled officers. represented), leading to potentially biased survey estimates.

The Abuse of Police Authority 59

Weighting took place in seven steps: (a) a cell is the ratio of the sum of IDW for all each sampled department was assigned an departments in the cell to the sum of IDW initial weight equal to the inverse of its for responding departments in the cell, and probability of selection; (b) cells were the final department weight is formed for department-level nonresponse FDW(dept) = IDW(dept) x DNRA(dcell) adjustment; (c) department-level nonresponse weights were computed; (d) each Each officer was then assigned an initial sampled officer was assigned an initial weight (IWO), where weight, which was the product of the officer’s department’s weight and the in- IWO(officer) = FDW(dept) x 1/P(officer|dept) verse of the officer’s probability of selec- where P (officer dept) is the probability of tion within the department; (e) new cells an officer being selected for the sample were formed for officer-level nonresponse within the department. adjustments; (f) officer-level nonresponse weights were computed; and (g) each Officer Nonresponse Adjustment: Compu- officer’s final weight was the product of tation of the officer nonresponse adjust- the initial officer weight and the officer- ment (DNRAO) was similar to that for de- level nonresponse weight. partments, except the cells were defined differently. For adjustments at the individual Initial Department Weights: Initially weight- level, one cell comprised four departments ing departments by the inverse of the prob- selected with certainty within a region. One ability of selection was required because cell included three certainty departments departments were sampled with probabil- in two neighboring regions. One cell com- ity proportional to size. Thus the initial prised a certainty selection and two other department weight (IDW) is large departments within a region. Eight IDW(dept) = 1/P(dept) other cells were defined by the intersection of region and the two major where P (dept) is the department’s prob- noncertainty strata. ability of selection. ONRA (ocell) for a cell is the ratio of the When department-level nonresponse sum of IWO (officer) for all sampled offic- adjustments were made, the initial depart- ers in the cell to the sum of IWO (officer) ment weight allowed each department to for responding officers in the cell, and the represent its appropriate share of the final weight is population. Finalwt(officer) = IWO(officer) x ONRA(ocell). Department Nonresponse Adjustment: One large department (selected with certainty 11. The survey instrument and re- for the sample survey) chose not to responses, including raw frequencies and spond. Because of this, we defined one weighted percentages, are provided in cell for department-level nonresponse to Appendix A. Throughout this report, un- include all sampled departments within less otherwise indicated, percentages pre- those departments having at least 400 full- sented in the text are weighted percent- time sworn officers (FTSW). Other cells ages. were defined by the intersection of region 12. Officers could identify themselves and major stratum (selected with certainty, as belonging to more than one racial cat- noncertainty with more than 24 FTSW and egory. 10–24 FTSW). Departments assigned to the first cell described above were not also as- 13. For example, in their re-examina- signed to other cells. The departmental tion of 5,688 cases in the 1977 Police Ser- nonresponse adjustment, DNRA (dcell), for vices Study data, Worden and Shepard

Police Foundation 60

(1996) found that reasonable force was confidence intervals were ±5.21 percent for used in 37 cases, and improper force was “increases,” ±9.88 percent for “decreases,” used in 23 cases. We note that improper and ±8.31 percent for “no impact.” For force was thus used in 38 percent of en- other minority officers, confidence inter- counters that involved force. Similarly, in vals were ±4.29 percent for “increases,” his re-analysis of 1,565 cases in Albert ±11.37 percent for “decreases,” and ±11.76 Reiss’s 1967 data, Friedrich (1980) found percent for “no impact.” that reasonable force was used in 52 cases, and excessive force was used in 28 cases. 20. For Table 4.20, confidence intervals Excessive force was thus used in 35 per- range from ±5.53 percent to ±5.84 percent cent of encounters that involved force. for supervisors.

14. These issues are revisited later, 21. For Table 4.21, confidence intervals when we take up questions of differences range from ±5.88 percent to ±6.03 percent among subgroups of officers and consider for supervisors. whether police officers of different races have differing views of racial bias by po- 22. For Table 4.22, confidence intervals ± ± lice officers. range from 5.25 percent to 6.50 percent for supervisors 15. We are beginning to examine whether the influence of race might be 23. For Table 4.23, confidence intervals explained by other factors, such as the for supervisors ranged from ±4.98 percent concentration of minority police officers in to ±5.57 percent. specific parts of the country. Our findings suggest that the importance of race is main- 24. For Table 4.24, confidence intervals ± tained even when controlling for other rel- for supervisors ranged from 5.99 percent ± evant demographic characteristics. to 6.66 percent.

16. In Table 4.14, confidence intervals 25. For Table 4.25, confidence intervals ± ranged from ±4.61 percent to ±13.43 per- for supervisors ranged from 5.84 percent ± cent for African Americans, and from ±5.86 to 5.90 percent. percent to ±13.74 percent for other minori- 26. For Table 4.26, confidence intervals ties. We remind the reader that we report for supervisors were ±5.55 percent. confidence intervals only if they are greater ± than 4 percent. 27. For Table 4.27, confidence intervals ± 17. Confidence intervals ranged from for supervisors were 6.08 percent for “de- ± ±5.86 percent to ±10.76 percent for Afri- creases” and 6.35 percent for “no impact.” can Americans and from ±9.8 percent to For nonsupervisors, confidence intervals ± ±11.19 percent for other minorities. were 4.27 percent for “decreases” and ±4.31 percent for “no impact.” 18. Confidence intervals ranged from ±6.02 percent to ±10.94 percent for Afri- 28. For Table 4.28, confidence intervals can Americans and from ±10.49 percent to for supervisors were ±6.25 percent for “de- ±12.23 percent for other minorities. creases” and ±6.35 percent for “no impact.” For nonsupervisors, confidence intervals 19. For the number of incidents of ex- were ±4.29 percent for “no impact.” cessive force (Table 4.17), confidence intervals for white officers were ±4.12 per- 29. For Table 4.29, confidence intervals cent for “decrease” and ±4.19 percent for for the North Central region ranged from “no impact.” For African-American officers, ±4.50 percent to ±6.30 percent. For the

The Abuse of Police Authority 61

Northeast they ranged from ±4.50 percent cent for the largest departments and from to ±5.70 percent. For the South they ranged ±6.00 percent to ±8.00 percent for the small from ±4.30 percent to ±5.70 percent, and departments. for the West they ranged from ±7.90 percent to ±8.00 percent. 33. In Table 4.33, confidence intervals ranged from ±5.50 percent to ±7.50 per- 30. For Table 4.30, confidence intervals cent for the largest departments. for the North Central region were ±5.70 percent. For the Northeast they were ±6.02 35. In Table 4.34, confidence intervals percent. For the South they were ±6.31 ranged from ±4.82 percent to ±7.93 per- percent, and for the West they were ±6.31 cent for the largest departments and from percent. ±5.02 percent to ±11.56 percent for the 31. For Table 4.31, confidence intervals small departments. for the North Central region were ±8.57 35. In Table 4.35, confidence intervals percent. For the Northeast they were ±5.88 ranged from ±7.29 percent to ±8.02 per- percent. For the South they were ±5.57 cent for the largest departments, ±3.39 per- percent, and for the West they were ±7.02 cent to ±4.25 percent for the medium percent. departments, and ±7.56 percent to ±7.80 32. In Table 4.32, confidence intervals percent for the small departments. ranged from ±5.00 percent to ±8.00 per-

Police Foundation 62

REFERENCES

Adams, Kenneth. 1996. “Measuring the Triangle Institute. Prevalence of Police Abuse of Force,” in Police Violence: Under- Bayley, David H., and Egon Bittner. 1984. standing and Controlling Police “Learning the Skills of Policing.” Abuse of Force, ed. William A. Law and Society Review 30(3): Geller and Hans Toch. New 585–606. Haven, CT: Yale University Press. Bennett, Richard R. 1994. Royal Barbados Police Force Survey. (Not pub- Amendola, Karen, Cielle Hockman, and lished.) Peter Scharf. 1996. Assessing Law Enforcement Ethics: Technical Black, Donald, and Albert J. Reiss, Jr. 1970. Report on the Study Conducted “Police Control of Juveniles.” with the Oregon Department of American Sociological Review 35: State Police. Washington, DC: 63–77. Police Foundation. Bureau of the Census. 1991. Census of Babbie, Earl. 1992. The Practice of Social Population and Housing, 1990: Research, 6th ed. Belmont, CA: Summary Tape File 3. Washington, Wadsworth. DC: Bureau of the Census. ———. 1990. Survey Research Methods, Bureau of Justice Statistics (BJS). 1995. Law 2nd ed. Belmont, CA: Wadsworth. Enforcement Management and Administrative Statistics, 1993. Babubhai, Shai V., Beth G. Barnwell, and Washington, DC: U.S. Department Gayle S. Bieler. 1997. Sudaan Re- of Justice, Bureau of Justice lease 7.5: User’s Manual. Research Statistics. Triangle Park, NC: Research

The Abuse of Police Authority 63

Carter, David L., and Allen D. Sapp. 1990. LeDoux, John C., and Robert R. Hazelwood. “The Evolution of Higher Educa- 1985. “Police Attitudes and Beliefs tion in Law Enforcement: Prelimi- Toward Rape.” Journal of Police nary Findings From A National Science and Administration 13(3): Study.” Journal of Criminal Justice 211–21. Education 1(1): 59–85. Lundman, Richard J. 1996. “Demeanor and Criminal Justice Commission. 1995. Ethi- Arrest: Additional Evidence from cal Conduct and Discipline in the Previously Unpublished Data.” Queensland Police Service: The Journal of Research in Crime and Views of Recruits, First Year Delinquency 33(3): 306–53. Constables, and Experienced Officers. Brisbane, Australia: ———. 1994. “Demeanor or Crime? The Criminal Justice Commission. Mid-West City Police-Citizen Encounters Study.” Criminology Federal Bureau of Investigation (FBI). 1994. 32(4): 631–56. Crime in the United States: Uniform Crime Reports, 1993. Lundman, Richard, Richard E. Sykes, and Washington, DC: U.S. Government John P. Clark. 1978. “Police Con- Printing Office. trol of Juveniles: A Replication.” Journal of Research in Crime and Friedrich, Robert J. 1980. “Police Use of Delinquency 33: 306–353. Force: Individuals, Situations, and Organizations.” The Annals of the Maguire, Edward R. 1997. “The Police American Academy of Political and Foundation Use of Force Study: Social Science 452 (November): Sampling Frame Design Issues.” 82–97. Memorandum prepared for the Police Foundation, 20 October Grant, J. Douglas, and Joan Grant. 1996. 1997. “Officer Selection and the Preven- tion of Abuse of Force,” in Police Maguire, Edward R., Jeffrey B. Snipes, Violence: Understanding and Craig D. Uchida, and Margaret Controlling Police Abuse of Force, Townsend. 1998. “Counting Cops: ed. William A. Geller and Estimating the Number of Police Hans Toch. New Haven, CT: Yale Depts and Police Officers in the University Press. USA” Policing: An International Journal of Police Strategies and Goldstein, Herman. 1979. “Improving Management. 21(1): 97–120. Policing: A Problem-Oriented Ap- proach.” Crime and Delinquency Martin, Christine, and Peter B. Bensinger. 25(2): 236–58. 1994. Illinois Municipal Officers’ Perceptions of Police Ethics. Chi- Klinger, David A. 1996. “More on Demeanor cago: Illinois Criminal Justice In- and Arrest in Dade County.” formation Authority. Criminology 34(1): 61–82. Mastrofski, Stephen D., Roger B. Parks, ———. 1994. “Demeanor or Crime? Why Christina DeJong, and Robert E. ‘Hostile’ Citizens Are More Likely Worden. “Race and Every-Day Po- to Be Arrested.” Criminology 32(3): licing: A Research Perspective,” 475–93. paper delivered at the Twelfth In- Klockars, Carl B., Sanja Kutnjak-Ivkovic and ternational Congress on Criminol- William E. Harver. (no date). ogy, Seoul, Korea, August 24–28, Police Integrity Survey. 1998: 14.

Police Foundation 64

Maxfield, Michael G., and Earl Babbie. Scrivner, Ellen M. 1994. The Role of Police 1995. Research Methods for Crimi- Psychology in Controlling Exces- nal Justice and Criminology. sive Force. Washington, DC: US Belmont, CA: Wadsworth. Department of Justice, Office of Justice Programs, National Institute McConkey, Kevin M., Gail F. Huon, and of Justice. Mark G. Frank. 1996. Practical Eth- ics in the Police Service: Ethics and Skolnick, Jerome H., and David H. Bayley. Policing—Study 3. Payneham, 1986. The New Blue Line: Police South Australia: National Police Innovation in Six American Cit- Research Unit. ies. New York: The Free Press. Skolnick, Jerome H. 1999. On Democratic McElroy, Jerome, Colleen A. Cosgrove, and Policing. Ideas in American Polic- Susan Sadd. 1990. CPOP: The Re- ing series. Washington, DC: Police search, An Evaluative Study of the Foundation. New York City Community Patrol Officer Program. New York: The Skolnick, Jerome H., and James J. Fyfe. Vera Institute of Justice. 1993. Above the Law: Police and the Excessive Use of Force. New National Advisory Commission on Crimi- York: The Free Press. nal Justice Standards and Goals. Smith, Douglas A., and Christy A. Visher. 1973. The Police. Washington, DC: 1981. “Street-Level Justice: U.S. Government Printing Office. Situational Determinants of Police National Center for Women and Policing. Arrest Decisions.” Social Problems [electronic bulletin board]; [cited 29(2): 167–77. 13 April 1998]; available from Smith, Douglas A., Christy A. Visher, and Laura (www.feminist.org/police/ A. Davidson. 1984. “Equity and Dis- ncwpAbout.html); INTERNET. cretionary Justice: The Influence of Race on Police Arrest Decisions.” Jour- Ogletree, Charles J. Jr., Mary Prosser, Abbe nal of Criminal Law and Criminol- Smith, and William Talley Jr. 1995. ogy 75(1): 234–49. Beyond the Rodney King Story: An Investigation of Police Misconduct Travis, Jeremy. 1995. “Education in Law in Minority Communities. Boston: Enforcement: Beyond the College Northeastern University Press. Degree.” Criminal Justice of the Americas 8: 1–6. Pate, Antony M., and Edwin E. Hamilton. 1992. The New York City Police Ca- Trojanowicz, Robert, and Bonnie det Corps Technical Report. Wash- Bucqueroux. 1994. Community ington, DC: Police Foundation. Policing: How to Get Started. Cin- cinnati: Anderson Publishing. Pate, Antony M., and Lorie A. Fridell. 1993. Tonry, Michael. 1995. Malign Neglect— Police Use of Force: Official Re- Race, Crime, and Punishment in ports, Citizen Complaints, and Le- America. New York: Oxford gal Consequences. Washington, University Press. p. 71. DC: Police Foundation. Walker, Samuel, Cassia Spohn, and Miriam Reiss, Albert J. Jr. 1971. The Police and DeLone. 1996. The Color of Justice: the Public. New Haven: Yale Race, Ethnicity, and Crime in University Press. America. Belmont, CA: Wadsworth.

The Abuse of Police Authority 65

Weisburd, David. 1998. Statistics in Crimi- nal Justice. New York: Wadsworth. Weisburd, David, Jerome McElroy, and Patricia Hardyman. 1988. “Challenges to Supervision in Community Policing: Observations on a Pilot Project.” American Journal of Police, 7(2): 29–50. Westley, William A. 1953. “Violence and the Police.” American Journal of Sociology, 59: 34–41. Worden, Robert E. 1990. “A Badge and a Baccalaureate: Policies, Hypotheses, and Further Evidence.” Justice Quarterly 7(3): 565–92. Worden, Robert E. 1996. “The ‘Causes’ of Police Brutality: Theory and Evidence on Police Use of Force” in Police Violence: Understanding and Controlling Police Abuse of Force. Ed. William A. Geller and Hans Toch. New Haven, CT: Yale University Press. Worden, Robert E., and Robin L. Shepard. 1996. “Demeanor, Crime, and Po- lice Behavior: A Reexamination of the Police Services Study Data.” Criminology 34(1): 83–105. Wycoff, Mary Ann, and Timothy N. Oettmeier. 1993. Evaluating Patrol Officer Performance Under Community Policing: The Houston Experience. Washington, DC: Police Foundation.

Police Foundation 66

AUTHORS

David Weisburd, PhD, is Senior Fellow, foundation research projects, and co-authored and formerly Senior Research Scientist, at several foundation research reports. the Police Foundation. He also is a profes- Kellie A. Bryant, MS, formerly a research sor of criminology at the Hebrew Univer- associate at the Police Foundation, is a sity Law School in Jerusalem, and a Senior Deputy Marshal with the United States Research Fellow in the Department of Marshals Service. Criminology and Criminal Justice at the Uni- versity of Maryland, College Park. Hubert Williams, JD, has been President of the Police Foundation since 1985. In Rosann Greenspan, PhD, formerly research 1962, he joined the Newark, New Jersey, director at the Police Foundation, is Assis- Police Department, was later promoted to tant Director and Visiting Scholar at the Cen- sergeant, then lieutenant, and was granted ter for the Study of Law and Society at the a leave of absence from 1972 to 1974 to University of California at Berkeley, where direct the Newark High Impact Anti-Crime she also teaches in the Legal Studies Pro- Initiative. In 1974, Williams was selected to gram. Dr. Greenspan continues to collabo- head the Newark Police Department, and rate on foundation research projects. served as its director until 1985. During his Edwin E. Hamilton, MA, is Senior Research career with the Newark Police Department, Analyst at the Police Foundation. Since com- Williams’s assignments included undercover ing to the Police Foundation in 1983, Mr. narcotics, patrol, field operations, commu- Hamilton has worked on dozens of nity relations, and the office of the chief.

The Abuse of Police Authority