Police Innovation

Over the last three decades American policing has gone through a period
of significant change and innovation. In what is a relatively short histor-
ical time frame the police began to reconsider their fundamental mis-
sion, the nature of the core strategies of policing, and the character of
their relationships with the communities that they serve. This volume
brings together leading police scholars to examine eight major innova-
tions which emerged during this period: community policing, broken
windows policing, problem oriented policing, pulling levers policing,
third party policing, hot spots policing, Compstat and evidence-based
policing. Including advocates and critics of each of the eight police inno-
vations, this comprehensive book assesses the evidence on impacts of
police innovation on crime and public safety, the extent of the implemen-
tation of these new approaches in police departments, and the dilemmas
these approaches have created for police management. This book will
appeal to students, scholars and researchers.

          is Walter E. Mayer Professor of Law and Criminal

Justice at the Hebrew University Law School in Jerusalem, and Professor
of Criminology and Criminal Justice at the University of Maryland,
College Park. He is the author or editor of eleven books and more than
sixty scientific articles that cover a wide range of criminal justice topics.

        .      is Senior Research Associate in the Program in

Criminal Justice Policy and Management at Harvard University’s John
F. Kennedy School of Government. His research focuses on working
with criminal justice agencies to develop crime prevention strategies to
deal with gang violence, illegal firearms markets, and violent crime hot
spots.
Cambridge Studies in Criminology

Edited by
Alfred Blumstein, H. John Heinz School of Public Policy and Management,
Carnegie Mellon University
David P. Farrington, Institute of Criminology, University of Cambridge

The Cambridge Studies in Criminology series aims to publish the highest qual-
ity research on criminology and criminal justice topics. Typical volumes report
major quantitative, qualitative, and ethnographic research, or make a substantial
theoretical contribution. There is a particular emphasis on research monographs,
but edited collections may also be published if they make an unusually distinctive
offering to the literature. All relevant areas of criminology and criminal justice are
included; for example, the causes of offending, juvenile justice, the development
of offenders, measurement and analysis of crime, victimization research, polic-
ing, crime prevention, sentencing, imprisonment, probation, and parole. The
series is global in outlook, with an emphasis on work that is comparative or holds
significant implications for theory or policy.

Other books in the series

Life in the Gang: Family, Friends, and Violence, by Scott H. Decker and Barrik
Van Winkle
Delinquency and Crime: Current Theories, edited by J. David Hawkins
Recriminalizing Delinquency: Violent Juvenile Crime and Juvenile Justice Reform, by
Simon I. Singer
Mean Streets: Youth Crime and Homelessness, by John Hagan and Bill McCarthy
The Framework of Judicial Sentencing: A Study in Legal Decision Making, by
Austin Lovegrove
The Criminal Recidivism Process, by Edward Zamble and Vernon L. Quinsey
Violence and Childhood in the Inner City, by Joan McCord
Judicial Policy Making and the Modern State: How the Courts Reformed America’s
Prisons, by Malcolm M. Feeley and Edward L. Rubin
Schools and Delinquency, by Denise C. Gottfredson
The Crime Drop in America, edited by Alfred Blumstein and Joel Wallman
Delinquent-Prone Communities, by Don Weatherburn and Bronwyn Lind
White-Collar Crime and Criminal Careers, by David Weisburd and Elin Waring,
with Ellen F. Chayet
Sex Differences in Antisocial Behaviour: Conduct Disorder, Delinquency, and
Violence in the Dunedin Longitudinal Study, by Terrie Moffitt, Avshalom Caspi,
Michael Rutter, and Phil A. Silva
Delinquent Networks: Youth Co-Offending in Stockholm, by Jerzy Sarnecki
Criminality and Violence among the Mentally Disordered: The Stockholm
Metropolitan Project, by Sheilagh Hodgins and Carl-Gunnar Janson
Why Corporations Obey the Law: Assessing Criminalization and Cooperative Models
of Crime Control, by Sally S. Simpson
Situational Prison Control: Crime Control in Correctional Institutions, by Richard
Wortley
Companions in Crime, by Mark Warr
The Criminal Career, by Britta Kyvsgaard
Gangs and Delinquency in Developmental Perspective, by Terence P. Thornberry
et al.
Early Prevention of Adult Antisocial Behaviour, edited by David P. Farrington and
Jeremy W. Coid
Violent Crime, by Darnell F. Hawkins
Errors of Justice: Nature, Sources and Remedies, by Brian Forst
Rethinking Homicide: Exploring the Structure and Process in Homicide Situations, by
Terance D. Miethe and Wendy C. Regoeczi
Understanding Police Use of Force: Officers, Suspects, and Reciprocity, by Geoffrey
P. Aipert, Roger G. Dunham
The Virtual Prison: Community Custody and the Evolution of Imprisonment, by
Julian V. Roberts
Marrying Time in the Golden State: Women’s Imprisonment in California, by
Cardace Kruttschnitt, Rosemary Gartner
Economic Espionage and Industrial Spying, by Hedieh Nasheri
Police Innovation
Contrasting Perspectives

Edited by
David Weisburd and Anthony A. Braga
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press

The Edinburgh Building, Cambridge cb2 2ru, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521836289

© Cambridge University Press 2006

This publication is in copyright. Subject to statutory exception and to the provision of

relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
First published in print format 2006

isbn-13 978-0-511-19176-3 eBook (NetLibrary)

isbn-10 0-511-19176-6 eBook (NetLibrary)

isbn-13 978-0-521-83628-9 hardback

isbn-10 0-521-83628-x hardback

isbn-13 978-0-521-54483-2 paperback

isbn-10 0-521-54483-1 paperback

Cambridge University Press has no responsibility for the persistence or accuracy of urls
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
David Weisburd: For Ariel Yehuda, who achieves
whatever he sets his mind to, yet always finds time
for the needs of others.
Anthony A. Braga: To Cassie, my friend and
companion
 Contents

List of figures page xii

List of tables xiii
Notes on contributors xiv

1 Introduction: understanding police innovation 1

                   
. 

Part I Community policing

2 Advocate
The promise of community policing 27
     .     
3 Critic
Community policing: a skeptical view 44
              

Part II Broken windows policing

4 Advocate
Of “broken windows,” criminology, and criminal
justice 77
        .             
. 
5 Critic
Incivilities reduction policing, zero tolerance, and the
retreat from coproduction: weak foundations and
strong pressures 98
     .   

ix
x Contents

Part III Problem-oriented policing

6 Advocate
Science, values, and problem-oriented policing: why
problem-oriented policing? 117
 . 
7 Critic
Problem-oriented policing: the disconnect between
principles and practice 133
        .                  

Part IV Pulling levers policing

8 Advocate
Old wine in new bottles: policing and the lessons of
pulling levers 155
    .      
9 Critic
Partnership, accountability, and innovation: clarifying
Boston’s experience with pulling levers 171
        .                         

Part V Third-party policing

10 Advocate
The case for third-party policing 191
                          
11 Critic
Third-party policing: a critical view 207
     .      

Part VI Hot spots policing

12 Advocate
Hot spots policing as a model for police innovation 225
                     .     
13 Critic
The limits of hot spots policing 245
      .      
 Contents xi

Part VII Compstat

14 Advocate
Compstat’s innovation 267
    .        
15 Critic
Changing everything so that everything can remain the
same: Compstat and American policing 284
         ,       .          ,
    .       ,                

Part VIII Evidence-based policing

16 Advocate
Evidence-based policing for crime prevention 305
 . 
17 Critic
Improving police through expertise, experience, and
experiments 322
 . 

18 Conclusion: Police innovation and the future of policing 339

        .                  

Index 353
 Figures

1.1 Total violent crime (trends in violent crime rates

by city size) page 10
7.1 Continuum of police strategies to control high-crime
places 146
9.1 Youth homicide in Boston, 1976–2003: victims ages 24
and under 175
9.2 Adult homicide in Boston, 1976–2003: victims ages 25
and older 177
12.1 The Minneapolis hot spots experiment 231
12.2 Cumulative distribution of computerized crime mapping 237
15.1 New York City homicide rate 1986–1998 296
15.2 Crime rates before and after Compstat implementation
in three model Compstat departments 297
18.1 Dimensions of policing strategies 341
18.2 Index crime rates per 100,000 residents in the United
States, 1990–2002 348

xii
 Tables

7.1 Problems and characteristics of twelve violent crime

places page 138
7.2 Responses to problems at violent crime places 142
10.1 Summary of third-party policing evaluation evidence 202
12.1 Results of hot spots policing evaluations 233
16.1 Sherman and Eck’s (2002) conclusions on the
effectiveness of police practices to prevent crime 311
16.2 Weisburd and Eck’s (2004) summary of findings on
police effectiveness research 314

xiii
 Notes on contributors

        .      is Senior Research Associate in the Program in

Criminal Justice Policy and Management of the Malcolm Wiener
Center for Social Policy at Harvard University’s John F. Kennedy
School of Government. He received his MPA from Harvard University
and Ph.D. in Criminal Justice from Rutgers University. His research
focuses on working with criminal justice agencies to develop crime
prevention strategies to deal with urban problems such as firearms
violence, street-level drug markets, and violent crime hot spots. He
has served as a consultant on these issues to the Rand Corporation;
National Academy of Sciences; US Department of Justice; US Depart-
ment of the Treasury; Bureau of Alcohol, Tobacco, and Firearms;
Boston Police Department; New York Police Department; and other
state and local law enforcement agencies.
       is a Professor in the Division of Criminal Justice at the Uni-
versity of Cincinnati where he teaches graduate courses on research
methods, police effectiveness, crime prevention, and criminal justice
policy. Dr. Eck is internationally known for his studies on problem-
oriented policing, the prevention of crime at places, the analysis and
mapping of crime hot spots, drug dealing and trafficking control, and
criminal investigations. He was a member of the National Academy
of Sciences Committee on Police Policy and Research, and a former
Director of Research for the Police Executive Research Forum, where
he helped pioneer the development and testing of problem-oriented
policing. Dr. Eck is an Individual Affiliate of the Center for Problem-
Oriented Policing and a judge for the British Home Office’s Tilley
Award for Excellence in Problem Solving. He earned his Ph.D. in
criminology from the University of Maryland and his bachelors and
masters degrees from the University of Michigan.
      .        , Ph.D., is a professor in the School of Criminal
Justice, Rutgers University, Newark, faculty chair of the Police Institute
at Rutgers University, Newark, and a Senior Fellow at the Manhattan

xiv
 Notes on contributors xv

Institute. In 1972, he began work at the Police Foundation and con-

ducted several large-scale experiments in policing, most notably the
Kansas City Preventive Patrol Experiment and the Newark Foot Patrol
Experiment. The latter was the source of his contribution to his most
familiar publication in the Atlantic, “Broken Windows,” with James
Q. Wilson. During the late 1980s, Dr. Kelling developed the order
maintenance policies in the New York City subway that ultimately led
to radical crime reductions. His areas of special interest are the police,
the relationships among fear, crime, and disorder, community crime
control, and the evolution of policing strategies and tactics. He con-
sults with numerous major police departments in the United States,
along with Australian, British, and Canadian police departments and
oversight agencies.
    .       joined the faculty of the John Jay College of Crimi-
nal Justice at the beginning of 2005. From 1991 to 2004, he was a senior
researcher at the Program in Criminal Justice Policy and Management,
Kennedy School of Government, Harvard University. He has written
and consulted extensively in the areas of community and problem-
solving policing, police corruption, and neighborhood revitalization.
He has performed field work in police departments and troubled com-
munities in many American cities, London, Sydney, and Puerto Rico.
He is the co-author of a seminal work on community policing, Beyond
911: A New Era for Policing, and numerous articles on police manage-
ment, illicit drug markets, illicit firearms markets, youth violence, and
deterrence theory, including editorials in the New York Times, Wash-
ington Post, and elsewhere. He has advised the Justice Department,
the Department of the Treasury, the Bureau of Alcohol, Tobacco, and
Firearms, the Office of National Drug Control Policy, and the White
House on these issues.
      .          is Professor of Public and International
Affairs at George Mason University, where he directs the Administra-
tion of Justice Program and the Center for Justice Leadership and Man-
agement. Research interests include measuring police performance and
assessing police reforms. He has received the Academy of Criminal
Justice Science’s O. W. Wilson Award for outstanding contributions to
police education, research, and practice.
               is an Associate Professor in the Department of
Criminology and Criminal Justice at Griffith University. She received
her Ph.D. from Rutgers University, New Jersey in 1993 and spent an
additional seven years as an academic in the USA (at Northeastern
xvi Notes on contributors

University and the University of Cincinnati). She is the recipient

of numerous US and Australian research grants on topics such as
problem-oriented policing, police technologies (e.g., crime mapping,
gunshot detection systems, 3-1-1 call systems), community crime con-
trol, civil remedies, street-level drug enforcement, and policing public
housing sites. In 2003, Associate Professor Mazerolle was admitted as
a member to the Academy of Experimental Criminologists and now
serves as the Vice President of the Academy. She also serves on the
Board of Studies for the Australian Institute for Police Management,
on the Capital Cities Lord Mayors Drug Advisory Board and as an
Associate Editor of the Journal of Experimental Criminology. Associate
Professor Mazerolle is the author of Policing Places with Drug Problems
and a co-editor, with Jan Roehl, of Civil Remedies and Crime Prevention.
She has written many scholarly articles on policing, drug law enforce-
ment, displacement of crime, and crime prevention.
          is Max Pam Professor of Law and Director, Center for
Studies in Criminal Justice at the University of Chicago. She received
her B.S. in General Engineering from the University of Illinois, and her
J.D. from the University of Chicago Law School. She joined the Uni-
versity of Chicago faculty in 1994 after serving as an Honors Program
Trial Attorney in the Antitrust Division of the United States Depart-
ment of Justice. Prior to serving as a Department of Justice prosecutor,
Ms. Meares clerked for Judge Harlington Wood, Jr. of the US Court
of Appeals for the Seventh Circuit. Ms Meares’s teaching and research
interests center on criminal procedure and criminal law policy, with
a particular emphasis on empirical investigation of these subjects. In
addition to teaching at the Law School, Ms Meares has an appoint-
ment as a Research Fellow at the American Bar Foundation. She is also
a faculty member of the University of Chicago Center for the Study
of Race, Politics, and Culture and an executive committee member
of the Northwestern/University of Chicago Joint Center for Poverty
Research.
     .      is the Guggenheim Professor of Criminal Justice
Policy and Management at Harvard’s Kennedy School of Govern-
ment. He has written extensively in the fields of criminal justice, public
management, and the voluntary sector. In the field of policing, his
significant publications include Beyond 911: A New Era for Policing?,
Recognizing Value in Public Policing: The Challenge of Measuring Police
Performance, What Citizens Should Value (and Measure!) in Police Depart-
ment Performance. He was also one of the principal authors of the Urban
 Notes on contributors xvii

Institute’s evaluation of the Federal Government’s COPS Program

responsible for the section of the report that used detailed case stud-
ies of twelve different police agencies to determine whether and how
the grants they received from the COPS program helped them make
the transition to a new strategy of policing. He was also a member
of the National Academy of Sciences Panel on Policing that produced
the report entitled: Fairness and Effectiveness in Policing.
         is a Senior Lecturer in the School of Criminology
and Criminal Justice, Griffith University. Her areas of interest include
policing and governance, white-collar crime, forensic mental health,
and commissions of inquiry. Recent publications include: Third Party
Policing, Cambridge University Press, forthcoming; and “Miscarriages
of Justice” in T. Prenzler and J. Ransley (eds.) Police Reform: Building
Integrity (2002). She is currently researching a book on government
inquiries, and a project on counter-terrorist policing.
      .       is Professor of Criminal Justice and Direc-
tor of the Center for Research in Law and Justice at the University
of Illinois at Chicago. Previously, he served as Director of Graduate
Studies, Department Head, and Dean. His primary areas of interest
include the evaluation of police, community, school, and multiagency
initiatives to prevent crime and delinquency. In the absence of a real
hobby, he is working with the Chicago Police Department to develop
geo-based Internet surveys that will “measure what matters” to com-
munity residents with respect to neighborhood problems, police per-
formance, community performance, police–community relations, and
related topics. He has written various books and articles on commu-
nity crime prevention and community policing, and teaches graduate
courses in policing, research methods, program evaluation, and com-
munity processes.
    .         , Ph.D., is Professor Emeritus at John Jay Col-
lege of Criminal Justice and the Graduate Center of City Univer-
sity of New York. He has previously served with the US Depart-
ment of Justice and the National Academy of Public Administration in
Washington, DC and was a Visiting Exchange Professor at the Police
Staff College in Bramshill, England. He has served as an expert wit-
ness, consultant to, and trainer with numerous criminal justice agen-
cies. His areas of interest include: community policing, crime analy-
sis and information systems, police management, training, operations,
use of force, integrity control, comparative policing, policy analysis,
xviii Notes on contributors

Compstat, and crime mapping. His recent publications include:

“Compstat” in Encyclopedia of Law Enforcement (2005); “Policing a
Diverse Community: A Case Study,” in Policing and Minority Com-
munities, Dolores Jones-Brown and Karen Terry (eds.) (2003); “Zero
Tolerance,” in Encyclopedia of Crime and Punishment, David Levinson,
(ed.) (2002).
     .      is Professor of Political Science and a member of
the research faculty at the University’s Institute for Policy Research.
His research focuses on public encounters with the institutions of jus-
tice, including the police, crime prevention projects, and community-
oriented policing. His most recent books on policing are Community
Policing: Can it Work? (2004), On the Beat: Police and Community Problem
Solving (1999) and Community Policing, Chicago Style (1997). Professor
Skogan chaired the National Research Council’s Committee to Review
Research on Police Policy and Practices. The Committee’s report,
Fairness and Effectiveness in Policing: The Evidence, was published in
2004.
        .      is an Assistant Professor in the Department of
Criminal Justice at the University of Nevada Las Vegas. He received
his Ph.D. from Rutgers University (2003) and his M.S. in Criminal
Justice from Northeastern University (1995). From 2000 to 2004,
he was the Director of Evaluation for the Police Institute at Rutgers-
Newark where he participated in studies related to violence and disor-
der in New Jersey neighborhoods. His current research projects involve
police order-maintenance practices, police management, and commu-
nity crime prevention.
       received his Ph.D. in social psychology at Johns Hopkins
University in 1977. He has previously held positions at Virginia Tech
(1977–1979), and Johns Hopkins University (1978–1985). He was a
Visiting fellow at the National Institute of Justice in 1997. He currently
teaches and researches in the Department of Criminal Justice at Tem-
ple University where he has been since 1984. He served as department
chair from July 2000 through July 2004. He edited Urban Neighborhoods
(1986) and authored Human Territorial Functioning (1988), Research
Methods in Criminal Justice (1994) and Breaking Away from Broken
Windows (2000). His most recent refereed publications include
Robinson, Lawton, Taylor, and Perkins (2003), “Multilevel Longitudi-
nal Impacts of Incivilities”; Taylor, Anderson, and McConnell (2003),
“Competencies and interest in a problem-focused undergraduate
 Notes on contributors xix

research methods criminal justice course”; and Taylor (2002), “Fear

of crime, local social ties, and collective efficacy: Maybe masquerading
measurement, maybe déjà vu all over again.” His research has been
funded by the National Science Foundation, the National Institute
of Mental Health, the National Institute of Justice, and the National
Institute of Corrections.
          is the Walter E. Meyer Professor of Law and Criminal
Justice at the Hebrew University Law School in Jerusalem and Profes-
sor of Criminology and Criminal Justice at the University of Maryland,
College Park. Professor Weisburd also serves as a Senior Fellow at the
Police Foundation, and Chair of its Research Advisory Committee.
He was a member of the National Research Council working group
on Evaluating Anti-Crime Programs and its panel on Police Practices
and Policies. Professor Weisburd is author or editor of 11 books and
more than 60 scientific articles, and is founding editor of the Journal of
Experimental Criminology. He received his Ph.D. from Yale University.
        .      , Ph.D., is an Associate Professor in the Depart-
ment of Criminal Justice, University of Massachusetts Lowell. His
research interests include the prevention of delinquency and crime
and economic analysis of prevention programs. Funded research has
included studies on the monetary costs of juvenile offending in urban
areas and the role of public health in the prevention of juvenile
criminal violence. Dr. Welsh is an author or editor of six works,
including Evidence-Based Crime Prevention (2002, with Lawrence
Sherman, David Farrington, and Doris MacKenzie). Recent articles
have appeared in Criminology & Public Policy, Justice Quarterly, Crime
and Justice: A Review of Research, and Annals of the American Academy of
Political and Social Science. He was a member of OJJDP’s Study Group
on Very Young Offenders and a contributing author to its volume, and
has been a consultant to the United Nations Office on Drugs and Crime
and the British Home Office.
                 is the Norman Tishman and Charles M.
Diker Professor of Sociology in the Faculty of Arts and Sciences and
a Fellow of the Hauser Center for Nonprofit Organizations. Prior to
coming to Harvard in 1992, he was Professor of Sociology, Statis-
tics, and Economics at Northwestern University and a senior faculty
research associate in the Institute for Policy Research there. He is the
author of a variety of articles on various statistical issues – the analysis
of qualitative dependent variables, selection bias, and counterfactual
xx Notes on contributors

causal analysis. His research has also focused on changes in the social
and economic status of African-Americans during the twentieth cen-
tury. In particular, he has examined changes in youth unemployment,
marital behavior, and prison incarceration. For the past eight years he
has been working with and studying a group of black inner city minis-
ters known as the Ten Point Coalition. He holds a Ph.D. in Sociology
from Harvard.
1 Introduction: understanding police
innovation

David Weisburd and Anthony A. Braga

Introduction
Over the last three decades American policing has gone through a period
of significant change and innovation. In what is a relatively short histori-
cal time frame the police began to reconsider their fundamental mission,
the nature of the core strategies of policing, and the character of their
relationships with the communities that they serve. Innovations in polic-
ing in this period were not insular and restricted to police professionals
and scholars, but were often seen on the front pages of America’s news-
papers and magazines, and spoken about in the electronic media. Some
approaches, like broken windows policing – termed by some as zero toler-
ance policing – became the subject of heated political debate. Community
policing, one of the most important police programs that emerged in this
period, was even to give its name to a large federal agency – The Office of
Community Oriented Policing Services – created by the Violent Crime
Control and Law Enforcement Act of 1994.
Some have described this period of change as the most dramatic in the
history of policing (e.g., see Bayley 1994). This claim does not perhaps
do justice to the radical reforms that led to the creation of modern police
forces in the nineteenth century, or even the wide-scale innovations in
tactics or approaches to policing that emerged after the Second World
War. However, observers of the police today are inevitably struck by the
pace and variety of innovation in the last few decades. Whether this period
of change is greater than those of previous generations is difficult to know
since systematic observation of police practices is a relatively modern
phenomenon. But there is broad consensus among police scholars that
the last three decades have “witnessed a remarkable degree of innovation
in policing” (Committee to Review Research 2004: 82).
In this volume we bring together leading police scholars to examine
the major innovations in policing that emerged during the last decades of
the twentieth century. We focus on eight innovations that are concerned
with change in police strategies and practices: community policing,

1
2 David Weisburd and Anthony A. Braga

broken windows policing, problem-oriented policing, pulling levers polic-

ing, third-party policing, hot spots policing, Compstat and evidence-
based policing. This is of course not an exhaustive list of innovation in
policing during this period. For example, we do not examine technolog-
ical innovations such as advances in computerized crime mapping or the
use of DNA in criminal investigations. We also do not examine innova-
tions in tactics and strategies that affected only specialized units or were
applied to very specific types of crimes. Our approach was to identify
innovations that had influence on the broad array of police tasks and on
the practices and strategies that broadly affected the policing of American
communities.
We title the chapters examining each police innovation reviewed in this
volume under the heading “Advocate” and “Critic.” In this context our
book seeks to clarify police innovation in the context of chapters writ-
ten by those who have played important roles in developing innovation,
and those who have stood as critics of such innovation. Nonetheless, we
do not take a debate format in our book. Authors did not respond to
each other’s papers, but rather sought to present a perspective that would
clarify the benefits of the innovation examined, or the potential problems
that the innovation raises for policing. The critics often identify promising
elements of innovation while pointing out the difficulties that have been
encountered in the applications of innovation in the field. The advocates
often note the drawbacks of particular strategies, while arguing that they
should be widely adopted. Accordingly, our chapters represent serious
scholarly examination of innovations in policing, recognizing that estab-
lished scholars may disagree about the directions that policing should
take while drawing from the same empirical evidence.
By design the essays in this volume take a “micro” approach to the prob-
lem of police innovation, focusing on the specific components, goals, and
outcomes associated with a specific program or practice. In this introduc-
tory chapter, we would like to take a “macro” approach to the problem
of police innovation that allows us to see how innovation more gener-
ally emerged and developed during this period. We do not think that the
dramatic surge in police innovation of the last few decades occurred as a
matter of chance. Our approach is to see the development of innovation
in policing as a response to a common set of problems and dilemmas.
This approach can also help us to understand the broad trends of police
innovation that we observe.

Understanding innovation and policing

Many scholars seem to take for granted what strikes us as a central
problem in understanding the broader phenomenon that our volume
 Introduction 3

examines. Why did we observe a period of significant innovation in polic-

ing in the last decades of the twentieth century? One simple answer to
this question would be to note that institutions change, and that when
faced with new ideas that have potential to improve their functioning
they will naturally choose what is innovative. However, those who have
studied the diffusion of innovation have been led to a very different view
of the processes that underlie the adoption of new products, programs,
or practices. Everett Rogers, who pioneered the scientific study of diffu-
sion of innovation, argues, for example, that “more than just a beneficial
innovation is necessary” to explain its widespread diffusion and adoption
(1995: 8). Indeed, there are many examples of innovations that represent
clear improvements over prior practice, yet fail to be widely adopted.
Rogers brings the example of the “Dvorak Keyboard” named after
a University of Washington researcher, who sought to improve on
the “Qwerty” keyboard in use since the late nineteenth century. The
“Qwerty” keyboard was engineered to slow down typists in the nine-
teenth century in order to prevent jamming of keys that was common
in the manual typewriters in that period. However, as the engineering
of typewriters improved in the twentieth century, there was no longer a
need for a keyboard engineered to slow typists down. Indeed, it seemed
natural that a better arrangement of the keyboard would be developed
that would allow for quicker typing that would cause less fatigue. Dvorak
developed such a keyboard in 1932 basing his arrangement of the keys on
time and motion studies. Dvorak’s keyboard was clearly an improvement
on the Qwerty keyboard. It allowed for more efficient and faster typing,
and led to less fatigue on the part of typists. But today more than seventy
years after Dvorak’s development of a better and more efficient keyboard,
the Qwerty keyboard remains the dominant method. Indeed, Dvorak’s
keyboard is merely an interesting historical curiosity.
The diffusion of innovation requires that there be a “perceived need”
for change in the social system in which an innovation emerges (Rogers
1995: 11). That need can be created by industries or interest groups, for
example through advertisements that lead consumers to believe that they
must have a particular new product or service. Often in social systems,
the recognition that something must change is brought about by a period
of crisis or challenge to existing programs or practices (see e.g., Rogers
1995; Altschuler and Behn 1997). In this context, we think that the key to
understanding the emergence of a period of rapid innovation in policing
in the last decades of the twentieth century lies in a crisis in policing that
emerged in the late 1960s. Identifying that period of crisis can help us
to understand not only why we observe so much police innovation in
recent decades, but also why that innovation follows particular patterns
of change.
4 David Weisburd and Anthony A. Braga

The crisis of confidence in American policing

The decade of the 1970s began with a host of challenges to the police as
well as the criminal justice system more generally (LaFree 1998). This
was the case in part because of the tremendous social unrest that char-
acterized the end of the previous decade. Race riots in American cities,
and growing opposition, especially among younger Americans, to the
Vietnam War, often placed the police in conflict with the young and with
minorities. But American fears of a failing criminal justice system were
also to play a role in a growing sense of crisis for American policing. In
1967, a presidential commission report on the Challenge of Crime in a
Free Society reinforced doubts about the effectiveness of criminal justice
in combating crime in the United States:

In sum, America’s system of criminal justice is overcrowded and overworked,

undermanned, underfinanced, and very often misunderstood. It needs more
information and more knowledge. It needs more technical resources. It needs
more coordination among its many parts. It needs more public support. It needs
the help of community programs and institutions in dealing with offenders and
potential offenders. It needs, above all, the willingness to reexamine old ways of
doing things, to reform itself, to experiment, to run risks, to dare. It needs vision.
(President’s Commission 1967, 80–81)

Shortly after the presidential report on the Challenge of Crime in a Free

Society, the Kerner Commission on Civil Disorders published a report
which was also to raise significant questions about the nature of criminal
justice in the United States, and the organization of American policing.
However, in this case it was the question of race and the relationship
between police and minority communities that was to have center stage.
The challenges to patterns of American discrimination against African
Americans were not focused primarily on the police, but the police, in
addition to other criminal justice agencies, were seen as “part of the
problem” and not necessarily working to help in producing a solution to
difficult social issues:

In Newark, Detroit, Watts and Harlem, in practically every city that has experi-
enced racial disruption since the summer of 1964, abrasive relationships between
police and Negroes and other minority groups have been a major source of
grievance, tension and ultimately disorder. (Kerner Commission 1968: 157)

The concerns of the commission reports in the 1960s and the sense
of growing alienation between the police and the public in the latter half
of that decade led policymakers, the police, and scholars to question the
nature of American policing, and in particular the strategies that were
 Introduction 5

dominant in policing since World War II. A recent National Research

Council Committee to Review Research on Police Policy and Practices
has termed these approaches as the “standard model of policing” (Com-
mittee to Review Research 2004; see also Weisburd and Eck 2004):
This model relies generally on a “one size fits all” application of reactive strategies
to suppress crime, and continues to be the dominant form of police practices in
the United States. The standard model is based on the assumption that generic
strategies for crime reduction can be applied throughout a jurisdiction regardless
of the level of crime, the nature of crime, or other variations. Such strategies
as increasing the size of police agencies, random patrol across all parts of the
community, rapid response to calls for service, generally applied follow-up inves-
tigations, and generally applied intensive enforcement and arrest policies are all
examples of this standard model of policing. (Weisburd and Eck 2004: 44)

A number of important questions about the standard model of polic-

ing had been raised in the 1960s. Nonetheless, there was little serious
academic inquiry into the impact of policing strategies on crime or on
public attitudes. The need for such research was apparent, and in the
1970s serious research attention was to begin. One important impetus
for such studies of the police came from the federal government. With
the Omnibus Crime Control and Safe Streets Act of 1968 a research
arm of the Department of Justice was established, eventually to become
the National Institute of Justice, which was to invest significant resources
in research on police and other components of the criminal justice sys-
tem. But important funding for research on policing was also to come
from private foundations. Perhaps the most important contribution to
policing was made by the Ford Foundation in 1970 when it established
the Police Development Fund. The Fund and the Police Foundation
which it established, was to foster a series of large-scale studies on Amer-
ican policing. McGeorge Bundy, then president of the Ford Foundation,
argued in announcing the establishment of a Police Development Fund in
1970:
The need for reinforcement and change in police work has become more urgent
than ever in the last decade because of rising rates of crime, increased resort to
violence and rising tension, in many communities, between disaffected or angry
groups and the police. (Bundy 1970)

With the establishment of the Police Foundation and the newly estab-
lished federal support for research on the criminal justice system, the
activities of the police began to come under systematic scrutiny by
researchers. Until this time, there had been a general assumption that
policing in the post-World War II era represented an important advance
over previous decades and was effective in controlling crime.
6 David Weisburd and Anthony A. Braga

For example, perhaps the dominant policing strategy in the post-World

War II period was routine preventive patrol in police cars. It was drawn
from a long history of faith in the idea of “police patrol” that had become
a standard dogma of policing for generations. As George Kelling and
his colleagues wrote in their introduction to the Kansas City Preventive
Patrol Experiment, a study conducted by the Police Foundation:

Ever since the creation of a patrolling force in 13th century Hangchow, preven-
tive patrol by uniformed personnel has been a primary function of policing. In
20th century America, about $2 billion is spent each year for the maintenance
and operation of uniformed and often superbly equipped patrol forces. Police
themselves, the general public, and elected officials have always believed that the
presence or potential presence of police offices on patrol severely inhibits criminal
activity. (Kelling, Pate, Dieckman, and Brown 1974: 1)

Preventive patrol in police cars was the main staple of police crime
prevention efforts at the beginning of the decade of the 1970s. As Kelling
and colleagues noted in the Police Foundation report on the Kansas City
study, “(t)oday’s police recruits, like virtually all those before them, learn
from both teacher and textbook that patrol is the ‘backbone’ of police
work” (Kelling et al. 1974: 1). The Police Foundation study sought
to establish whether empirical evidence actually supported the broadly
accepted assumptions regarding preventive patrol. The fact that ques-
tions were raised about routine preventive patrol suggests that the con-
cerns about the police voiced in the decade before had begun to impact
the confidence of police managers. As Kansas City Police Chief Clarence
M. Kelley, later to become director of the Federal Bureau of Investiga-
tion (FBI), said in explaining the need for the Kansas City experiment:
“Many of us in the department had the feeling we were training, equip-
ping, and deploying men to do a job neither we, nor anyone else, knew
much about” (Murphy 1974: v).
To understand the impact of the Kansas City study on police man-
agers and researchers, it is important to recognize not only that the study
examined a core police practice but that its methodological approach
represented a radical departure from the small-scale evaluations of police
practices that had come earlier. The Kansas City Preventive Patrol Exper-
iment was a social experiment in policing on a grand scale, and it was
conducted in a new Foundation that had significant resources and was
backed by the well-established and respected Ford Foundation. Patrick
Murphy, the distinguished police manager, and president of the Police
Foundation at the time, suggests just how much the Foundation itself
saw the experiment as a radical and important change in the quality of
police research.1
 Introduction 7

This is a summary report of the findings of an experiment in policing that ranks

among the few major social experiments ever to be completed. The experiment
was unique in that never before had there been an attempt to determine through
such extensive scientific evaluation the value of visible police patrol. (Murphy
1974: v)
This context, both in terms of the centrality of the strategy examined,
the scale of the research, and the prestige of the institutions that supported
the study, including the Kansas City Police Department and its chief,
Clarence Kelley, were to give the findings of the study an impact that is
in retrospect out of proportion to the actual findings. One study in one
jurisdiction, no matter how systematic, cannot provide a comprehensive
portrait of the effects of a strategy as broad as routine preventive patrol.
Moreover, the study design was to come under significant academic crit-
icism in later years (Minneapolis Medical Research Foundation 1976;
Larson and Cahn 1985; Sherman and Weisburd 1995). Nonetheless, in
the context of the decade in which it was conducted, this study was to
have a critical impact upon the police and police researchers. This was
especially the case since the research findings were to be consistent with
a series of other studies of core police practices.
Kelling and his colleagues, in cooperation with the Kansas City Police
Department, took fifteen police beats and divided them up into three
groups. In five of these, called “reactive” beats, “routine preventive patrol
was eliminated and officers were instructed to respond only to calls for
service” (Kelling et al. 1974: 3). In five others, defined as “control” beats,
“routine preventive patrol was maintained at its usual level of one car per
beat” (ibid.: 3). In the remaining five beats, termed “proactive” beats,
“routine preventive patrol was intensified by two to three times its usual
level through the assignment of additional patrol cars” (Kelling et al.
1974: 3). When Kelling and his colleagues published the results of their
study in 1974 it shattered one of the bedrock assumptions of police
practitioners – that preventive patrol was an effective way to prevent
crime and increase citizens’ feelings of safety. They concluded simply
that increasing or decreasing the intensity of routine preventive patrol
in police cars did not affect either crime, service delivery to citizens, or
citizens’ feelings of security.
Another large-scale study conducted by William Spelman and Dale
Brown and published in 1984 was also to challenge a core police assump-
tion of that period – that improvement in rapid response to calls for
service would lead to improvements in crime fighting. This study was
developed in good part because of the findings of a prior investigation in
Kansas City that found little support for the crime control effectiveness of
rapid response to calls for service (Kansas City Police Department 1977).
8 David Weisburd and Anthony A. Braga

With support from the National Institute of Justice, Spelman and Brown
interviewed 4000 victims, witnesses, and bystanders in some 3300 serious
crimes in four American cities. This was another major study in terms of
the resources brought to bear and the methods used. Again it examined
a strategy that was aided by technological advances in the twentieth cen-
tury and that was a central dogma of police administrators – that police
must get to the scene of a crime quickly if they are to apprehend criminal
offenders. Spelman and Brown explained:
For at least half a century, police have considered it important to cut to a min-
imum their response times to crime calls. The faster the response, they have
reasoned, the better the chances of catching a criminal at or near the scene of the
crime. (Spelman and Brown 1984: xxi)

Based on the data they collected, however, Spelman and Brown pro-
vided a very different portrait of the crime control effectiveness of rapid
response to calls for service:
Rapid police response may be unnecessary for three out of every four serious
crimes reported to police. The traditional practice of immediate response to
all reports of serious crimes currently leads to on-scene arrests in only 29 of
every 1,000 cases. By implementing innovative programs, police may be able to
increase this response-related arrest rate to 50 or even 60 per 1000, but there is
little hope that further increases can be generated. (Spelman and Brown 1984:
xix)

These findings based on a host of systematic data sources from multi-

ple jurisdictions provided little support for the strategy of rapid response
as a police practice to do something about crime. Indeed, Spelman and
Brown found that citizen reporting time, not police response time, most
influenced the possibility of on-scene arrest. Marginal improvement in
police response times was predicted to have no real impact on the appre-
hension or arrest of offenders.
The Kansas City Preventive Patrol Experiment and the National Insti-
tute of Justice study of police response time were not the only studies to
“debunk” existing police practices. James Levine, for example, analyzed
national crime data on the effectiveness of increasing the number of police
in an article published in 1975. His title sums up his findings: “The Inef-
fectiveness of Adding Police to Prevent Crime.” Despite the fact that
this effort and many others that challenged conventional police practices
did not represent the kind of systematic data collection or analysis of the
Police Foundation and National Institute of Justice studies, they followed
a similar “narrative” which became increasingly common as the 1990s
approached. Levine, for example, begins by noting the broad consensus
for the principle that adding more police will make cities safer. He then
 Introduction 9

goes on to note that “(s)ensible as intensified policing may sound on the

surface, its effectiveness in combating crime has yet to be demonstrated”
(Levine 1975: 523). Finally, drawing upon simple tabular data on police
strength and crime rates over time, he concludes:
It is tempting for politicians and government leaders to add more police: it is an
intuitively sensible and symbolically satisfying solution to the unrelenting problem
of criminal violence . . . The sad fact is, however, that they receive a false sense
of security; in most situations they are just as vulnerable with these extra police
as without them. (Levine 1975: 544)

Follow-up investigations were also the subject of critical empirical

research during this period. The standard model of policing had assumed
that general improvements in methods of police investigations would
lead to crime control gains both because more active offenders would
be imprisoned and thus unable to commit crime, and because poten-
tial offenders would be deterred by the prospect of discovery and arrest
(Committee to Review Research 2004). But a series of studies in the
1970s and early 1980s suggested that investigations had little impact
upon crime (Greenwood et al. 1975; Greenwood, Petersilia, and Chaiken
1977; Skogan and Antunes 1979; Eck 1983). This was the case in good
part because many crimes, especially property crimes, were found to be
unlikely to be solved by police investigations. These studies consistently
showed that if citizens did not provide information about suspects to
first responding officers, follow-up investigations were unlikely to lead to
successful outcomes.
In retrospect, many of these studies overstated what could be learned
about standard police practices from the findings gained (Weisburd and
Eck 2004). And, in practice, there were evaluations in this period that
produced more promising findings regarding standard police practices
such as routine preventive patrol (e.g., see Press 1971; Schnelle, Kirch-
ner et al. 1977; Chaiken 1978). Nonetheless, as the United States entered
the decade of the 1990s there appeared to be a general consensus that tra-
ditional police practices did not work in preventing or controlling crime.
As Michael Gottfredson and Travis Hirschi wrote in their classic book A
General Theory of Crime in 1990: “No evidence exists that augmentation
of patrol forces or equipment, differential patrol strategies, or differential
intensities of surveillance have an effect on crime rates” (Gottfredson and
Hirschi 1990: 270).
David Bayley wrote even more strongly in 1994:
The police do not prevent crime. This is one of the best kept secrets of modern
life. Experts know it, the police know it, but the public does not know it. Yet
the police pretend that they are society’s best defense against crime . . . This
10 David Weisburd and Anthony A. Braga

2,500

2,000
Per 100,000 Population

1,500

1,000

500

0
1973 1976 1979 1982 1985 1988
Year
1,000,000+ 500,000–999,999 250,000–499,999

Figure 1.1 Total violent crime (trends in violent crime rates by city
size)
Source: Reiss and Roth, 1993

is a myth. First, repeated analysis has consistently failed to find any connection
between the number of police officers and crime rates. Secondly, the primary
strategies adopted by modern police have been shown to have little or no effect
on crime. (Bayley 1994: 3)

This view of the ineffectiveness of policing strategies was reinforced

by official crime statistics. These statistics, widely available to the public,
suggested that the police were losing the “war on crime.” In particular,
in America’s largest cities, with their well-established professional police
forces, crime rates and especially violent crime rates were rising at alarm-
ing rates. Between 1973 and 1990 violent crime doubled (Reiss and Roth
1993). It did not take a statistician to understand that the trends were
dramatic. For example, in Figure 1.1 we report the trends in violent crime
rates by city size per 100,000 population. Clearly crime was on the rise,
and the trend had been fairly consistent over a long period. Thus, not
only were scholars showing that police strategies did little to impact upon
crime, but the overall crime statistics commonly used by the government
 Introduction 11

and community to define police effectiveness were providing a similar

message.

Crisis and change in American policing

It is against this backdrop that the innovations that we examine in this
volume develop. Our view is that the challenges to police effectiveness,
rising crime rates, and concerns about the legitimacy of police actions
that developed in the late 1960s created a perceived need for change in
what some have described as the industry of American policing (Ostrom,
Whitaker, and Parks 1978; Committee to Review Research 2004). Unfor-
tunately, there is no hard empirical evidence that would allow us to make
this link directly, since the study of the adoption of innovation has only
recently become a subject of interest for police scholars (e.g., see Weiss
1997; Klinger 2003; Weisburd, Mastrofski, McNally et al. 2003). Accord-
ingly, there have been few systematic studies of these processes and schol-
ars were generally not concerned about the emergence of innovation as a
research problem when these innovations were being developed.
Nonetheless, we think it reasonable to make a connection between the
perceived failures of the standard model of American police practices
and the experimentation with innovation, and openness to the adoption
of innovation that occurred in the last decades of the twentieth century.
Certainly, such a link is made by many of those who fostered innovation
in policing. For example, in his proposal for a problem-oriented policing
in 1979, Herman Goldstein referred directly to the growing evidence of
the failures of traditional police practices:

Recently completed research questions the value of two major aspects of police
operations – preventive patrol and investigations conducted by detectives. Some
police administrators have challenged the findings; others are awaiting the results
of replication. But those who concur with the results have begun to search for
alternatives, aware of the need to measure the effectiveness of a new response
before making a substantial investment in it. (1979: 240)

William Bratton (1998a) in describing the emergence of Compstat in

New York City, also refers to the failures of traditional approaches, and
the need for innovation that would allow the police to be more effective
in doing something about crime problems:

The effects of rapidly responding to crimes were muted because research showed
it took people almost 10 minutes to decide to call the police in the first place.
And police riding in air-conditioned squad cars, rapidly going from call to call,
did not make people feel safer. In fact, it further separated the police from the
public, the consumers of police services.
12 David Weisburd and Anthony A. Braga

Fortunately, the researchers and practitioners did not stop their work at finding
what was not working, but began to look at how to think differently about crime
and disorder and develop strategies that would work. (1998: 31)

More generally, the turn of the last century was a period of tremen-
dous change in police practices. This is perhaps most evident in the
development of community-oriented policing, which was aided by finan-
cial support from the Office of Community Oriented Policing Services
established in 1994. As Wesley Skogan reports in this volume, com-
munity policing in some form has been adopted by most police agen-
cies in the United States. In a Police Foundation survey conducted in
1997, 85 percent of surveyed police agencies reported they had adopted
community policing or were in the process of doing so (Skogan forth-
coming). A Bureau of Justice Statistics survey (2003) conducted at the
turn of the century found that more than 90 percent of departments
in cities over 250,000 in population reported having full-time, trained
community policing officers in the field (Bureau of Justice Statistics
2003).
The openness of police agencies to innovation is perhaps even more
strongly illustrated by the sudden rise of Compstat as a police practice.
Compstat was only developed as a programmatic entity in 1994 and was
not encouraged financially by federally funded programs. Nonetheless
by the turn of the century more than a third of larger police agencies
had claimed to have implemented the program and a quarter of police
agencies claimed that they were planning to adopt a Compstat program
(Weisburd et al. 2003).
A number of police scholars have suggested that the changes that such
surveys observe are more cosmetic than substantive. Some studies have
documented the “shallow” implementation of police innovations, and
have suggested that in the end the police tend to fall back on traditional
methods of conducting police work (e.g., see Clarke 1998; Eck 2000).
For example, even in police agencies that have adopted innovations in
problem-oriented policing, careful analysis of the activities of the police
suggest that they are more likely to follow traditional police practices than
to choose innovative approaches (Braga and Weisburd in this volume).
Moreover, the main practices of the standard model of policing continue
to dominate the work of most police agencies (Committee to Review
Research 2004).
While the depth of innovation over the last few decades remains a
matter of debate, it is certainly the case that police agencies have become
open to the idea of innovation, and that new programs and practices
have been experimented with and adopted at a rapid pace over the last
 Introduction 13

few decades. We think this openness can be traced to the crisis in police
legitimacy and effectiveness that we have described.

Understanding the form and character

of police innovation
Recognizing the importance of the challenges to policing that began to
emerge in the late 1960s can help us to understand not only the cause
for a period of rapid innovation, but also the form and character of the
innovations that we observe in this period and which are the focus of
our volume. The innovations we study here represent different forms of
adaptation to similar problems. Overall, they seek to find a solution to a
set of challenges to the effectiveness and legitimacy of policing we have
reviewed.
Community policing, which is examined in the next section of our
volume, is one of the first new approaches to policing to emerge in this
modern period of police innovation. Community policing programs were
already being implemented and advocated in the 1980s (e.g., Trojanow-
icz 1982; Goldstein 1987; Cordner 1988; Green and Mastrofski 1988;
Weisburd and McElroy 1988; Trojanowicz 1989), and by the 1990s, as
we have already noted, the idea of community policing had affected most
American police agencies. Police practices associated with community
policing have been diverse and have often changed over time. Foot patrol,
for example, was considered an important element of community policing
in the 1980s, but has not been a core component of more recent commu-
nity policing programs. Community policing has often been implemented
in combination with other programs, such as problem-oriented policing,
thus making it difficult to distinguish the core components of commu-
nity policing from those of other innovations that developed during this
period.
Nonetheless, community policing represented a radical departure from
the professional model of policing that was dominant in the post-World
War II period. For decades the police had assumed that the main task of
policing was to fight crime, and that the police, like other professionals,
could successfully carry out their task with little help and preferably with
little interference from the public. The police were the experts in defining
the nature of crime problems and the nature of the solutions that could
be brought to do something about them. The community in this context
did not have a central role in the police function, and the responsibility
for crime problems lay squarely in the hands of the policing industry.
One core element of the community policing movement was that the
community should play a central role in defining the problems the police
14 David Weisburd and Anthony A. Braga

address, and that these problems should extend much beyond conven-
tional law enforcement. As Kelling and Moore (1988: 4) argue, “during
the 1950s and 1960s, police thought they were law enforcement agencies
primarily fighting crime.” In the “community era” or community polic-
ing era, the police function broadens and includes “order maintenance,
conflict resolution, provision of services through problem solving, as well
as other activities” (Kelling and Moore 1988: 2). The justification for
these new activities was drawn either from a claim that historically the
police had indeed carried out such functions, or that the community from
which the police gained legitimacy saw these as important functions of
the police.
David Bayley notes that this approach “creates a new role for police
with new criteria for performance”:

If police can not reduce crime and apprehend more offenders, they can at least
decrease fear of crime, make the public feel less powerless, lessen distrust between
minority groups and the police, mediate quarrels, overcome the isolation of
marginal groups, organize social services, and generally assist in developing “com-
munity.” These are certainly worthwhile objectives. But are they what the police
should be doing? They are a far cry from what the police were originally created
to do. (Bayley 1988: 228)

One way to understand the early development of community policing

is to recognize that it responds to the question: What is the justification
for the police if they cannot prevent crime? While crime fighting has
increasingly become a central concern in community policing over the
last decade, an important contribution of community policing to police
innovation was its recognition that there were many critical community
problems that the police could address that were not traditionally defined
as crime problems. The expansion of the police function was to become
an important part of many of the innovations that are discussed in this
volume. This definition of new tasks for policing can be seen in part as a
response to the failure of police to achieve the crime control goals of the
professional model of policing.
Other innovations in policing in this period also looked to redefine
the role of the police in one way or another. Broken windows policing,
the subject of Part II of our volume, also seeks to direct the police to
problems that had often been ignored in standard police practices. The
idea of broken windows policing developed out of a Police Foundation
study, the Newark Foot Patrol Experiment, published in 1981. From
that study James Q. Wilson and George Kelling (1982) identified a link
between social disorder and crime which suggested the importance of
police paying attention to many problems that were seen in earlier decades
 Introduction 15

as peripheral to the police function. Wilson and Kelling were impressed

by the activities of the police officers who walked patrol in the Police
Foundation study, and thought that what might be seen in traditional
policing as inappropriate behavior actually held the key to public safety
and crime reduction. Kelling and Coles (1996: 18) write:
Most New Jersey police chiefs were dismayed when they learned from program
evaluators what (anonymous) officers, who were supposed to be “fighting crime,”
were actually doing while on foot patrol. For example, after being called a second
time during the same evening to end brawls in the same bar, one foot patrol officer
had had enough: although the “bar time” was some hours away, he ordered the
bar closed for business as usual. The bartender grumbled, closed up, and opened
the next day for business as usual. When this incident was recounted to the chief
of the department in which it occurred . . . he responded, “that wouldn’t happen
in my department, the officer would be fired.”

Wilson and Kelling argued that concern with disorder was an essential
ingredient for doing something about crime problems. Indeed, the broken
windows thesis was that serious crime developed because the police and
citizens did not work together to prevent urban decay and social disorder:
(A)t the community level, disorder and crime are usually inextricably linked, in
a kind of developmental sequence. Social psychologists and police officers tend
to agree that if a window in a building is broken and is left unrepaired, all of the
rest of the windows will soon be broken. (Wilson and Kelling 1982: 31)

In the context of crime, Wilson and Kelling argued “that ‘untended’

behavior also leads to the breakdown of community controls”:
A stable neighborhood of families who care for their homes, mind each other’s
children, and confidently frown on unwanted intruders can change in a few years
or even a few months, to an inhospitable and frightening jungle. A piece of prop-
erty is abandoned, weeds grow up, a window is smashed. Adults stop scolding
rowdy children; the children emboldened, become more rowdy. Families move
out, unattached adults move in . . . Such an area is vulnerable to criminal
invasion. Though it is not inevitable, it is more likely here, rather than in
places where people are confident they can regulate public behavior by informal
controls . . . (Wilson and Kelling 1982: 31)

Broken windows policing encourages the police to be concerned with

problems of disorder, and moves crime itself to a secondary, or at least
second-stage goal of the police. From the perspective of the crisis of
policing we have described, broken windows policing again responds to
the crisis of police effectiveness by expanding the police function. Both in
community policing and broken windows policing the failures of crime
fighting became less important, because the police function was seen to
lie in good part in other activities.
16 David Weisburd and Anthony A. Braga

Problem-oriented policing, the subject of Part III of our volume, also

sought to broaden the problems that police approached. In Herman
Goldstein’s original formulation of problem-oriented policing in 1979
he argued that the “police job requires that they deal with a wide range
of behavioral problems that arise in the community” (1979: 242). How-
ever, in this case, the solution for the crisis of policing was not found
in the definition of new tasks for the police, bur rather in a critique
of traditional police practice. Goldstein assumed that the police could
impact crime and other problems if they took a different approach, in
this case, the problem-oriented policing approach. Accordingly, a second
response to the crisis we have described is not to accept, as some academic
criminologists had, that the police could not do something about crime
and thus to search to define other important police functions as central
(Gottfredson and Hirschi 1990; Bayley 1994) but to argue that the strate-
gies of the standard policing model were flawed and that new more
effective models could be developed.
Problem-oriented policing sought to redefine the way in which the
police did their job. Goldstein argued that the police had “lost sight”
of their primary task which was to do something about crime and other
problems, and instead had become focused on the “means” of allocat-
ing police resources. He identified this pathology as a common one in
large organizations, and sought through the model of problem solving
to develop a more successful method of ameliorating crime and other
community problems.
Other innovations in policing that emerged fully only in the 1990s
also take the approach that the police can be effective in doing some-
thing about crime if they adopt innovative police practices. Pulling levers
policing discussed in Part IV, adopts a problem-oriented approach, but
provides a broader and more comprehensive combination of strategies
than more traditional problem-oriented policing programs. Pioneered in
Boston to deal with an “epidemic” of youth violence (Kennedy, Piehl, and
Braga 1996), the pulling levers approach begins by drawing upon a col-
lection of law enforcement practitioners to analyze crime problems and
develop innovative solutions. It seeks to develop a variety of “levers” to
stop offenders from continuing criminal behavior that include not only
criminal justice interventions, but also social services and community
resources.
Third-party policing discussed in Part V, offers another solution to
the failures of the standard policing model. It follows suggestions made
by Herman Goldstein (1979) that the “tool box” of police strategies be
expanded. In this case, however, the resources of the police are expanded
to “third parties” that are believed to offer significant new resources
 Introduction 17

for doing something about crime and disorder.2 The opportunity for a
third-party policing approach developed in part from more general trends
in the relationship between civil and criminal law (Mann 1992; Mazerolle
and Ransley in this volume). The expansion of the civil law and its use
in other legal contexts as a method of dealing with problems that were
once considered to be the exclusive province of criminal statutes cre-
ated important new tools for the police. Third-party policing asserts that
the police cannot successfully deal with many problems on their own,
and thus that the failures of traditional policing models may be found
in the limits of police powers. Using civil ordinances and civil courts,
or the resources of private agencies, third-party policing recognizes that
much social control is exercised by institutions other than the police and
that crime can be managed through agencies other than the criminal
law.
Hot spots policing discussed in Part VI, was first examined in the
Minneapolis Hot Spots Experiment (Sherman and Weisburd 1995). The
Minneapolis study was developed as a direct response to the findings of
the Kansas City Preventive Patrol Experiment. Drawing upon empirical
evidence that crime was clustered in discrete hot spots (e.g., see Pierce,
Spaar, and Briggs 1986; Sherman, Gartin, and Buerger 1989; Weisburd,
Maher, and Sherman 1992), Sherman and Weisburd argued that pre-
ventive patrol might be effective if it was more tightly focused. If “only
3 percent of the addresses in a city produce more than half of all the
requests for police response, if no police are dispatched to 40 percent
of the addresses and intersections in a city over one year, and, if among
the 60 percent with any requests the majority register only one request
a year, then concentrating police in a few locations makes more sense
than spreading them evenly through a beat” (Sherman and Weisburd
1995: 629). Hot spots policing does not demand that the police change
their strategies, but requires that they focus them more carefully at places
where crime is clustered.
Compstat, discussed in Part VII, also responds to the failures of the
standard model of policing by critiquing the ways in which the police carry
out their task. However, in the case of Compstat the focus is less on the
specific strategies that the police are involved in and more on the nature
of police organization itself. Herman Goldstein noted in 1979 that the
failures of the standard model of policing could be explained by the fact
that police organizations were poorly organized to do something about
crime. Compstat was designed to overcome that limitation. It sought to
empower the police command structure to do something about crime
problems. William Bratton, the New York City police chief who coined
the term and developed the program wrote:
18 David Weisburd and Anthony A. Braga

We created a system in which the police commissioner, with his executive core,
first empowers and then interrogates the precinct commander, forcing him or her
to come up with a plan to attack crime. But it should not stop there. At the next
level down, it should be the precinct commander, taking the same role as the
commissioner, empowering and interrogating the platoon commander. Then,
at the third level, the platoon commander should be asking his sergeants . . .
all the way down until everyone in the entire organization is empowered and
motivated, active and assessed and successful. It works in all organizations,
whether it’s 38,000 New York cops or Mayberry, R. F. D. (Bratton 1998b:
239)

Evidence-based policing, discussed in Part VIII also traces the failures

of traditional policing practices to the ways in which the police carry out
their tasks. It draws from a much wider set of public policy concerns,
and a broader policy movement than other police innovations examined
in this volume. There is a growing consensus among scholars, practition-
ers, and policymakers that crime control practices and policies should be
rooted as much as possible in scientific research (Sherman 1988; Cullen
and Gendreau 2000; MacKenzie 2000; Sherman et al. 2002). Over the
last decade there has been a steady growth in interest in the evalua-
tion of criminal justice programs and practices reflected in part by the
growth in criminal justice funding for research during this period (Visher
and Weisburd 1998; http://www.crimereduction.gov.uk/crimered.htm).
Increasing support for research and evaluation in criminal justice may
be seen as part of a more general trend toward utilization of scientific
research for establishing rational and effective practices and policies. This
trend is perhaps most prominent in the health professions where the idea
of “evidence-based medicine” has gained strong government and profes-
sional support (Millenson 1997; Zuger 1997), though the evidence-based
paradigm is also developing in other fields (e.g., see Nutley and Davies
1999; Davies, Nutley, and Smith 2000). The evidence-based approach
does not necessarily assume that the police can be more effective, but
it argues that a reliance on evidence in the police industry is a pre-
requisite for the development of effective policing practices (Sherman
1988).

Conclusion
In our Introduction we have traced the wide diffusion of innovations in
the last decades of the twentieth century to a crisis of police practices
that had begun to develop in the late 1960s. We have argued that it is
not accidental that so much innovation was brought to American polic-
ing during this period. Indeed, such innovation can be understood in the
 Introduction 19

context of a series of challenges to American policing that created a per-

ceived need for change among the police, scholars, and the public. We
have also argued that the paths of police innovation can be understood
in the context of the critiques that developed of standard policing mod-
els. In some cases, the innovations minimized the importance of crime
fighting, which had been the main focus of earlier policing models. Such
innovations responded to the crisis of policing by defining a broader or
new set of tasks which the police could perform more effectively. Other
innovations, however, started with a critique of the methods used in the
traditional policing models. These innovations assumed that the police
could be more effective, even in preventing or controlling crime, if the
tactics used were changed.
In the following chapters prominent police scholars examine each of
the innovations that we have discussed. The format of advocates and cri-
tiques provides a broad framework for assessing these innovations and
allows us to identify the major advantages and disadvantages of these
approaches. In our conclusions we try to draw more general lessons from
these contributions, and discuss the possible directions that police inno-
vation will take in the coming decades.

   
We would like to thank our colleagues John Eck, David Klinger, Cynthia Lum,
Lorraine Mazerolle, and Stephen Mastrofski for their thoughtful comments on
an earlier draft of this introductory chapter. Helpful comments were also pro-
vided by Kristen Miggans, a graduate student at the University of Maryland,
who also assisted us in preparing the manuscript for publication.
1. It is important to note that a number of large social experiments were con-
ducted during this period (e.g., see Bell et al. 1980; Struyk and Bendick
1981) and thus the Kansas City Preventive Patrol experiment can be seen
as part of a larger effort to subject social programs to systematic empirical
study.
2. It is important to note that the impetus for third-party policing does not nec-
essarily come from the police. Mazerolle and Ransley (2006) argues that a
variety of external demands have imposed third-party policing on the police
industry.


Altschuler, A. A. and Behn, R. D. (eds.) (1997). Innovation in American
government: Challenges, opportunities, and dilemmas. Washington, DC: The
Brookings Institution.
Bayley, D. H. (1988). Community policing: A report from the devil’s advocate.
In J. R. Greene and S. D. Mastrofski (eds.), Community policing: Rhetoric or
reality? (pp. 225–238). New York: Praeger.
(1994). Police for the future. New York: Oxford University Press.
20 David Weisburd and Anthony A. Braga

Bell, J. G., Robins, P. K., Spiegelman, R. G., and Weiner, S. (eds.) (1980).
A guaranteed annual income: Evidence from a social experiment. New York:
Academic Press.
Bratton, W. J. (1998a). Crime is down in New York City: Blame the police. In
W. J. Bratton and N. Dennis (eds.), Zero tolerance: Policing a free society.
London: Institute of Economic Affairs Health and Welfare Unit.
Bratton, W. J. with Knobler, P. (1998b). Turnaround: How America’s top cop
reversed the crime epidemic. New York: Random House.
Bundy, McGeorge. (1970, July). Press Conference. Presented in New York City,
NY.
Bureau of Justice Statistics. (2003). Local police departments 2000. Washington,
DC: Bureau of Justice Statistics, US Department of Justice.
Chaiken, J. (1978). What is known about deterrent effects of police activities.
In J. Cromer (ed.), Preventing crime (pp. 109–136). Beverly Hills, CA: Sage
Publications.
Clarke, R. (1998). Defining police strategies: Problem solving, problem-oriented
policing and community-oriented policing. In T. O’Connor Shelley and A. C.
Grant (eds.), Problem-oriented policing: Crime-specific problems, critical issues,
and making POP work. Washington, DC: Police Executive Research Forum.
Committee to Review Research on Police Policy and Practices. (2004). Fair-
ness and effectiveness in policing: The evidence. Washington, DC: The National
Academies Press.
Cordner, G. (1988). A problem-oriented approach to community-oriented polic-
ing. In J. Greene and S. Mastrofski (eds.), Community policing: Rhetoric or
reality? New York: Praeger.
Cullen, F. and Gendreau, P. (2000). Assessing correctional rehabilitation: Policy,
practice, and prospects. In J. Horney (ed.), Policies, processes, and decisions of
the criminal justice system: Criminal justice 3. Washington, DC: US Department
of Justice, National Institute of Justice.
Davies, H. T. O., Nutley, S., and Smith, P. (2000). What works: Evidence-based
policy and practice in public services. London: Policy Press.
Eck, J. E. (1983). Solving crime: A study of the investigation of burglary and robbery.
Washington, DC: Police Executive Research Forum.
(2000). Problem-oriented policing and it’s problems: The means over ends
syndrome strikes back and the return of the problem-solver. Unpublished
manuscript. Cincinnati, OH: University of Cincinnati.
Goldstein, H. (1979). Improving policing: A problem oriented approach. Crime
and Delinquency, 24, 236–258.
(1987). Toward community-oriented policing: Potential, basic requirements,
and threshold questions. Crime and Delinquency, 25, 236–258.
Gottfredson, M. and Hirschi, T. (1990). A general theory of crime. Palo Alto, CA:
Stanford University Press.
Greene, J. and Mastrofski, S. (eds.) (1988). Community policing: Rhetoric or reality?
New York: Praeger.
Greenwood, P. W., Chaiken, J., Petersilia, M., and Prusoff, L. (1975). Criminal
investigation process, III: Observations and analysis. Santa Monica, CA: Rand
Corporation.
 Introduction 21

Greenwood, P. W., Petersilia, J., and Chaiken, J. (1977). The criminal investigation
process. Lexington, MA: D.C. Heath.
Kansas City Police Department. (1977). Response time analysis. Kansas City, MO:
Kansas City Police Department.
Kelling, G. L. and Coles, C. M. (1996). Fixing broken windows: Restoring order
and reducing crime in our communities. New York: The Free Press.
Kelling, G. L. and Moore, M. H. (1988). From political to reform to
community: The evolving strategy of police. In J. R. Greene and S. D.
Mastrofski (eds.), Community policing: Rhetoric or reality? New York: Praeger
Publishers.
Kelling, G. L., Pate, A., Dieckman, D., and Brown, C. E. (1974). The Kansas
City preventative patrol experiment: Technical report. Washington, DC: Police
Foundation.
Kelling, G. L., Pate, A., Ferrera, A., Utne, M., and Brown, C. E. (1981). Newark
foot patrol experiment. Washington, DC: The Police Foundation.
Kennedy, D., Piehl, A., and Braga, A. (1996). Youth violence in Boston: Gun
markets, serious offenders, and a use-reduction strategy. Law and Contem-
porary Problems, 59, 147–196.
Kerner Commission. (1968). National advisory commission on civil disorder. Wash-
ington, DC: US Government Printing Office.
Klinger, D. A. (2003). Spreading diffusion in criminology. Criminology and Public
Policy, 2(3), 461–468.
LaFree, G. (1998). Losing legitimacy: Street crime and the decline of social institutions
in America. Boulder, CO: Westview Press.
Larson, R. C. and Cahn, M. F. (1985). Synthesizing and extending the results of
police patrols. Washington, DC: US Government Printing Office.
Levine, J. P. (1975). Ineffectiveness of adding police to prevent crime. Public
Policy, 23, 523–545.
MacKenzie, D. (2000). Evidence-based corrections: Identifying what works.
Crime and Delinquency, 46, 457–471.
Mann, K. (1992). Punitive civil sanctions: The middleground between criminal
and civil law. Yale Law Journal, 101(8), 1795–1873.
Mazerolle, L. G. and Ransley, J. (2006). Third-party policing. Cambridge: Cam-
bridge University Press.
Millenson, M. L. (1997). Demanding medical excellence: Doctors and accountability
in the information age. Chicago: University of Chicago Press.
Minneapolis Medical Research Foundation, Inc. (1976). Critiques and commen-
taries on evaluation research activities – Russell Sage reports. Evaluation,
3(1–2), 115–138.
Murphy, P. V. (1974). Foreword. In G. L. Kelling, T. Pate, D. Dieckman, and
C. E. Brown, The Kansas City preventative patrol experiment: Technical report.
Washington, DC: Police Foundation.
Nutley, S. and Davies, H. T. O. (1999). The fall and rise of evidence in criminal
justice. Public Money and Management, 19, 47–54.
Ostrom, E., Whitaker, G., and Parks, R. (1978). Policing: Is there a system? In
J. May and A. Wildavsky (eds.), The policy cycle. New York: Russell Sage
Foundation.
22 David Weisburd and Anthony A. Braga

Pierce, G., Spaar, S., and Briggs, L. R. (1986). The character of police work: Strategic
and tactical implications. Boston, MA: Center for Applied Social Research,
Northeastern University.
President’s Commission on Law Enforcement and Administration of Justice.
(1967). The crime commission report: The challenge of crime in a free society.
Washington, DC: US Government Printing Office.
Press, S. J. (1971). Some effects of an increase in police manpower in the 20th precinct
of New York City. New York: New York City Rand Institute.
Reiss, A. J., Jr. and Roth, J. A. (eds.) (1993). Understanding and preventing violence:
Panel on the understanding and control of violent behavior. Washington DC:
National Academy Press.
Rogers, E. M. (1995). Diffusion of innovations (4th ed.). New York: Free Press.
Schnelle, J. F., Kirchner, R. E., Jr., Casey, J. D., Uselton, P. H., Jr., and McNees,
M. P. (1977). Patrol evaluation research: A multiple-baseline analysis of
saturation police patrolling during day and night hours. Journal of Applied
Behavior Analysis, 10, 33–40.
Sherman, L. W. (1998). Evidence-based policing. Ideas in American policing series.
Washington, DC: The Police Foundation.
Sherman, L. W. and Weisburd, D. (1995). General deterrent effects of police
patrol in crime “hot-spots”: A randomized controlled trial. Justice Quarterly,
12, 626–648.
Sherman, L. W., Gartin, P. R., and Buerger, M. E. (1989). Hot spots of preda-
tory crime: Routine activities and the criminology of place. Criminology, 27,
27–56.
Sherman, L. W., Farrington, D., Welsh, B., and MacKenzie, D. (eds.) (2002).
Evidence based crime prevention. New York: Routledge.
Skogan, W. G. (in Press). Impediments to community policing. In L. Fridell and
M. A. Wycoff (eds.), The future of community policing. Washington, DC: Police
Executive Research Forum.
Skogan, W. G. and Antunes, G. E. (1979). Information, apprehension, and
deterrence: Exploring the limits of police productivity. Journal of Criminal
Justice, 7, 217–241.
Spelman, W. and Brown, D. K. (1984). Calling the police: Citizen reporting of serious
crime. Washington: US Government Printing Office.
Struyk, R. J. and Bendick, M., Jr. (eds.) (1981). Housing vouchers for the poor:
Lessons from a national experiment. Washington, DC: The Urban Institute
Press.
Trojanowicz, R. (1982). An evaluation of the neighborhood foot patrol program in
Flint, Michigan. East Lansing, MI: National Neighborhood Foot Patrol Cen-
ter, Michigan State University.
(1989). Preventing civil disturbances: A community policing approach. East
Lansing, MI: Michigan State University, National Center for Community
Policing.
Visher, C. and Weisburd, D. (1998). Identifying what works: Recent trends in
crime prevention strategies. Crime, Law and Social Change, 28, 223–242.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder,
and fear? Annals of the American Academy of Political and Social Science, 593,
42–65.
 Introduction 23

Weisburd, D. and Lum, C. (2005). The diffusion of computerized crime mapping

policing: Linking research and practice. Police Practice and Research, 6(5),
433–448.
Weisburd, D. and McElroy, J. E. (1988). Enacting the CPO (Community Patrol
Officer) role: Findings from the New York City pilot program in community
policing. In J. R. Greene and S. D. Mastrofski (eds.), Community policing:
Rhetoric or reality? New York: Praeger Publishers.
Weisburd, D., Maher, L., and Sherman, L. W. (1992). Contrasting crime gen-
eral and crime specific theory: The case of hot-spots of crime. Advances in
Criminological Theory, 4, 45–70.
Weisburd, D., Mastrofski, S. D., McNally, A. M., Greenspan, R., and Willis, J. J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2(3), 421–456.
Weiss, A. (1997). The communication of innovation in American policing. Polic-
ing, 20, 292–310.
Wilson, J. Q. and Kelling, G. L. (1982). Broken windows: The police and neigh-
borhood safety. The Atlantic Monthly, March, 29–38.
Zuger, A. (1997). New way of doctoring: By the book. The New York Times.
December 16.
Part I

Community policing
2 Advocate
The promise of community policing

Wesley G. Skogan

Community policing is very popular. So popular is the concept with politi-

cians, city managers, and the general public, that few police chiefs want
to be caught without some program they can call community policing.
In a 1997 survey of police departments conducted by the Police Foun-
dation, 85 percent reported they had adopted community policing or
were in the process of doing so (Skogan 2004). The biggest reason for
not doing so was that community policing was “impractical” for their
community, and my own tabulations of the data found these replies were
mostly from small departments with only a few officers. Bigger cities
included in the survey (those with populations greater than 100,000) all
claimed in the 1997 survey to have adopted community policing – half
by 1991 and the other half between 1992 and 1997. By 2000, a federal
survey with a much larger sample found that more than 90 percent of
departments in cities over 250,000 in population reported having full-
time, trained community policing officers in the field (Bureau of Justice
Statistics 2003).
What do cities that claim they are “doing community policing” actually
do? They describe a long list of projects. Under the rubric of community
policing, officers patrol on foot (in the 1997 survey, 75 percent listed
this), or perhaps on horses, bicycles, or segways. Departments variously
train civilians in citizen police academies, open small neighborhood store-
front offices, conduct surveys to measure community satisfaction, can-
vass door-to-door to identify local problems, publish newsletters, conduct
drug education projects, and work with municipal agencies to enforce
health and safety regulations.
However, community policing is not defined by these kinds of activities.
Projects, programs, and tactics come and go, and they should as condi-
tions change. Communities with different problems and varied resources
to bring to bear against them should try different things. Community
policing is not a set of specific programs. Rather, it involves changing
decisionmaking processes and creating new cultures within police depart-
ments. It is an organizational strategy that leaves setting priorities and the

27
28 Wesley G. Skogan

means of achieving them largely to residents and the police who serve
in their neighborhoods. Community policing is a process rather than a
product. It has three core elements: citizen involvement, problem solving,
and decentralization. In practice, these three dimensions turn out to be
densely interrelated, and departments that shortchange one or more of
them will not field a very effective program.
This chapter sets the stage for a discussion of community policing. It
reviews the three core concepts that define community policing, describes
how they have been turned into concrete community policing programs,
and reports some of what we know about their effectiveness. It draws
heavily on my experience evaluating community programs in a number
of cities, as well as on what others have reported. It summarizes some
of the claims made for community policing, and some of the realities of
achieving them in the real world.

Community involvement
Community policing is defined in part by efforts to develop partner-
ships with community members and the civic organizations that represent
many of them collectively. It requires that police engage with the public as
they set priorities and develop their tactics. Effective community policing
requires responsiveness to citizen input concerning both the needs of the
community and the best ways by which the police can help meet those
needs. It takes seriously the public’s definition of its own problems. This
is one reason why community policing is an organizational strategy but
not a set of specific programs – how it looks in practice should vary con-
siderably from place to place, in response to unique local situations and
circumstances. Listening to the community can produce new policing
priorities. Officers involved in neighborhood policing quickly learn that
many residents are deeply concerned about problems that previously did
not come to police attention. To a certain extent they define things differ-
ently. The public often focus on threatening and fear-provoking conditions
rather than discrete and legally defined incidents. They can be more con-
cerned about casual social disorder and the physical decay of their com-
munity than they are about traditionally defined “serious crimes.” They
worry about graffiti, public drinking, and the litter and parking problems
created by nearby commercial strips. The public sometimes define their
problem as people who need to be taught a lesson. In Chicago, a well-
known social type is the “gangbanger (gang member),” and people want
them off the street. The police, however, are trained to recognize and
organized to respond to crime incidents, and they have to know what
people do, not just who they are. Given these differences, community
 The promise of community policing 29

residents are unsure if they can (or even should) rely on the police to
help them deal with these problems. Many of these concerns thus do not
generate complaints or calls for service, and as a result, the police know
surprisingly little about them. The routines of traditional police work
ensure that officers will largely interact with citizens who are in distress
because they have just been victimized, or with suspects and troublemak-
ers. Accordingly, community policing requires that departments develop
new channels for learning about neighborhood problems. And, when
they learn about them, they have to have systems in place to respond
effectively.
Civic engagement usually extends to involving the public in some way
in efforts to enhance community safety. Community policing promises to
strengthen the capacity of communities to fight and prevent crime on their
own. The idea that the police and the public are “co-producers” of safety,
and that they cannot claim a monopoly over fighting crime, predates the
community policing era. In fact, the community crime prevention move-
ment of the 1970s was an important precursor to community policing.
It promoted the idea that crime was not solely the responsibility of the
police. The police were quick to endorse the claim that they could not
solve crime problems without community support and assistance, for it
helped share the blame for crime rates that were rising at the time (cf.
Skogan, Hartnett, DuBois et al. 1999). Now police find that they are
expected to lead community efforts. They are being called upon to take
responsibility for mobilizing individuals and organizations around crime
prevention. These efforts include neighborhood watch, citizen patrols,
and education programs stressing household target-hardening and the
rapid reporting of crime. Residents are asked to assist the police by report-
ing crimes promptly when they occur and cooperating as witnesses. Com-
munity policing often involves increases in “transparency” in how depart-
ments respond to demands for more information about what they do and
how effective they are. A federal survey of police agencies found that by
1999 more than 90 percent of departments serving cities of 50,000 or
more were giving residents access to crime statistics or even crime maps
(Bureau of Justice Statistics 2001). Even where efforts to involve the
community were already well established, moving them to center stage as
part of a larger strategic plan showcasing the commitment of the police
to community policing.
All of this needs to be supported by new organizational structures and
training for police officers. Departments need to reorganize in order to
provide opportunities for citizens to come into contact with their officers
under circumstances that encourage these exchanges. There has to be a
significant amount of informal “face time” between police and residents,
30 Wesley G. Skogan

so that trust and cooperation can develop between the prospective part-
ners. To this end, many departments hold community meetings and form
advisory committees, establish store front offices, survey the public, and
create informational web sites. In Chicago they hold about 250 small
police–public meetings every month. They began doing so in 1995, and
by mid-2003 residents had shown up on more than half a million occa-
sions to attend almost 25,000 community meetings (Skogan and Steiner
2004). In some places, police share information with residents through
educational programs or by enrolling them in citizen–police academies
that give them in-depth knowledge of law enforcement. By 1999, almost
70 percent of all police departments – and virtually every department
serving cities of 50,000 or more – reported regularly holding meetings
with citizen groups (Bureau of Justice Statistics 2001).
What are the presumed benefits of citizen involvement? Community
policing aims at rebuilding trust in the community and ensuring support
for the police among taxpayers. Opinion polls document the fact that
Americans have given up thinking that politicians and government ade-
quately represent them; for example, Moy and Pfau (2000: 13) note that
there is “a profound and consistent lack of confidence in the executive
branch of the federal government,” and document that the proportion
of Americans who believe that “the government in Washington” can be
trusted “to do what is right” dropped from 76 percent in 1964 to just
14 percent in 1994. Police come off better than most government insti-
tutions, when Americans are asked how much confidence they have in
them. My own review of public opinion polls indicates that during the
1990s police did better than the presidency and the Supreme Court, and
Americans had much more confidence in the police than they had in the
Congress. In May 2004, Americans were most confident in their military
(85 percent had “a great deal” or “quite a lot” of confidence in them),
but police came next, at 64 percent. Less than half of Americans had
that much confidence in the Supreme Court, and only 30 percent in the
Congress.
However, community policing is especially about recapturing the legit-
imacy that police have in large measure lost in many of America’s minority
communities. The same opinion polls show that African Americans and
recent immigrants have dramatically less confidence in the police, and
are much more likely to believe that they are brutal and corrupt (Skogan
and Steiner 2004). They are the only growing part of the population in
a surprisingly large number of American cities, and civic leaders know
that they have to find ways to incorporate them into the system. Police
take on community policing in part because they hope that building a
reservoir of public support may help them get through bad times (see the
 The promise of community policing 31

discussion of “nasty misconduct” below) when they occur. Community

policing might help police be more effective. It could encourage witnesses
and bystanders to step forward in neighborhoods where they too often
do not, for example. More indirectly, it might help rebuild the social and
organizational fabric of neighborhoods that previously had been given
up for lost, enabling residents to contribute to maintaining order in their
community (Sampson, Raudenbush, and Earls 1997).
However, I do not know of a single police department that adopted
community policing because they thought that it was a direct route to get-
ting the crime rate down. Efforts to evaluate it need to focus as well on the
important community and governance processes that it is intended to set
in motion, because they represent potentially important “wins” on their
own. The extent of the public’s trust and confidence in the police is an
obvious first issue. For example, after eight years of citywide community
policing, Chicagoans’ views of their police improved by 10–15 percentage
points on measures of their effectiveness, responsiveness, and demeanor.
Latinos, African Americans and whites all shared in these improvements
(Skogan and Steiner 2004). Evaluators also should look into the “mobi-
lizing” effects of programs, including the extent of parallel community
self-help efforts and extending even to the possible development of orga-
nizational and leadership capabilities among newly activated residents.
Sociological research indicates that “collective efficacy” (a combination of
trust among neighborhood residents and the expectation that neighbors
will intervene when things go wrong) plays an important role in inhibiting
urban crime. However, the same work indicates that it is mostly white,
home-owning neighborhoods that currently have it, and researchers have
yet to document how neighborhoods that do not have collective efficacy
can generate it for themselves (Sampson, Raudenbush, and Earls 1997).
Although its effects on collective efficacy are undocumented, the rhetoric
of community policing and its accomplishments in turning out residents
point in this direction, and this should be an important focus of evaluation
in this area.
An important spin-off of civic engagement is that the adoption of com-
munity policing almost inevitably leads to an expansion of the police man-
date, and this further expands the list of points on which it should be eval-
uated. Controlling serious crime by enforcing the criminal law remains
the primary job of the police. But instead of seeing the police exclusively
in these terms, and viewing activities that depart from direct efforts to
deter crime as a distraction from their fundamental mission, advocates
of community policing argue that the police have additional functions to
perform and different ways to conduct their traditional business. As a
practical matter, when police meet with neighborhood residents in park
32 Wesley G. Skogan

buildings and church basements to discuss neighborhood problems, the

civilians present are going to bring up all manner of problems. If the police
who are present put them off, or have no way of responding to their con-
cerns, they will not come back the next month. Community policing takes
seriously the public’s definition of its own problems, and this inevitably
includes issues that lie outside the traditional competence of the police.
Officers can learn at a public meeting that loose garbage and rats in an
alley are big issues for residents, but some other agency is going to have to
deliver the solution to that problem. When police meet with residents in
Chicago, much of the discussion focuses on neighborhood dilapidation
(including problems with abandoned buildings and graffiti) and on pub-
lic drinking, teen loitering, curfew and truancy problems, and disorder
in schools. There is much more talk about parking and traffic than about
personal and property crime, although discussion of drug-related issues
comes up quite often (Skogan, Steiner, Hartnett et al. 2003). The broad
range of issues that concern the public requires in turn that police form
partnerships with other public and private agencies that can join them
in responding to residents’ priorities. They could include the schools
and agencies responsible for health, housing, trash pickup, car tows, and
graffiti cleanups.
In practice, community involvement is not easy to achieve. Ironically, it
can be difficult to sustain in areas that need it the most. Research on par-
ticipation in community crime prevention programs during the 1970s and
1980s found that poor and high-crime areas often were not well endowed
with an infrastructure of organizations that were ready to get involved,
and that turnout for police-sponsored events was higher in places honey-
combed with block clubs and community organizations (Skogan 1988).
In high-crime areas people tend to be suspicious of their neighbors, and
especially of their neighbors’ children. Fear of retaliation by gangs and
drug dealers can undermine public involvement as well (Grinc 1994).
In Chicago, a 1998 study of hundreds of community meetings found
that residents expressed concern about retaliation for attending or work-
ing with the police in 22 percent of the city’s beats (Skogan and Steiner
2004). In addition, police and residents may not have a history of get-
ting along in poor neighborhoods. Residents are as likely to think of the
police as one of their problems as they are to see them as a solution to
their problems. It probably will not be the first instinct of organizations
representing the interests of poor communities to cooperate with police.
Instead, they are more likely to press for an end to police misconduct.
They will call for new resources from the outside to address commu-
nity problems, for no organization can blame its own constituents for
their plight (cf. Skogan 1988). There may be no reason for residents of
 The promise of community policing 33

crime-ridden neighborhoods to think that community policing will turn

out to be anything but another broken promise; they are accustomed
to seeing programs come and go, without much effect (Sadd and Grinc
1994). They certainly will have to be trained in their new roles. Commu-
nity policing involves a new set of jargon as well as assumptions about
the new responsibilities that both police and citizens are to adopt. The
2000 survey of police departments by the federal government found that
“training citizens for community policing” was common in big cities; in
cities of more than 500,000, 70 percent reported doing so (Bureau of
Justice Statistics 2003).
In addition, community policing runs the risk of inequitable outcomes.
In an evaluation of one of the very first programs, in Houston, Texas, I
found that whites and middle-class residents received most of the ben-
efits of the program. They found it easy to cooperate with the police,
and shared with the police a common view of who the troublemak-
ers were in the community. Blue-collar blacks and Latinos remained
uninvolved, on the other hand, and they saw no visible change in their
lives (Skogan 1990). Finally, the investment that police make in com-
munity policing is always at risk. Nasty episodes of police misconduct
can undermine those efforts. When excessive force or killings by police
become a public issue, years of progress in police–community relations
can disappear. The same is true when there are revelations of widespread
corruption.
On the police side there may be resistance in the ranks. Public offi-
cials’ and community activists’ enthusiasm for neighborhood-oriented
policing encourages its detractors within the police to dismiss it as “just
politics,” or another passing civilian fad. Officers who get involved can
become known as “empty holster guys,” and what they do gets labeled
“social work.” Police officers prefer to stick to crime fighting. (For a case
study in New York City of how this happens, see Pate and Shtull 1994.)
My first survey of Chicago police, conducted before that city’s commu-
nity policing program began, found that two-thirds of them disavowed
any interest in addressing “non-crime problems” on their beat. More
than 70 percent of the 7500 police officers surveyed thought community
policing “would bring a greater burden on police to solve all community
problems,” and also “more unreasonable demands on police by com-
munity groups” (Skogan and Hartnett 1997). Police are often skeptical
about programs invented by civilians, who they are convinced cannot
possibly understand their job. They are particularly hostile to programs
that threaten to involve civilians in setting standards or evaluating their
performance, and they do not like civilians influencing their operational
priorities. Police can easily find ways to justify their aloofness from the
34 Wesley G. Skogan

community; as one officer told me, “You can’t be the friend of the people
and do your job.”
On the other hand, some studies point to positive changes in offi-
cers’ views once they become involved in community policing. Lurigio
and Rosenbaum (1994) summarized twelve studies of this, and found
many positive findings with respect to job satisfaction, perceptions of
improved relations with the community, and expectations about com-
munity involvement in problem solving. Skogan and Hartnett (1997)
found growing support for community policing among officers involved
in Chicago’s experimental police districts, in comparison to those who
continued to work in districts featuring policing as usual.

Problem solving
Community policing also involves a shift from reliance on reactive patrol
and investigations toward a problem-solving orientation. In brief (for it
is discussed in detail in other chapters of this book) problem-oriented
policing is an approach to developing crime reduction strategies. Prob-
lem solving involves training officers in methods of identifying and analyz-
ing problems. It highlights the importance of discovering the situations
that produce calls for police assistance, identifying the causes which lie
behind them, and designing tactics to deal with these causes. Problem
solving is a counterpoint to the traditional model of police work, which
usually entails responding sequentially to individual events as they are
phoned in by victims. Too often this style of policing is reduced to driv-
ing fast to crime scenes in order to fill out pieces of paper reporting
what happened. Problem solving, on the other hand, calls for examin-
ing patterns of incidents to reveal their causes and to help plan how to
deal with them proactively. This is facilitated by the computer analyses
of “hot spots” that concentrate large volumes of complaints and calls for
service. Problem-oriented policing also recognizes that the solutions to
those patterns may involve other agencies and may be “non-police” in
character; in traditional departments, this would be cause for ignoring
them. The best programs encourage officers to respond creatively to the
problems they encounter, or to refer them appropriately to other agencies
(Eck 2004).
Problem-solving policing can proceed without a commitment to com-
munity policing. A key difference between problem solving and com-
munity policing is that the latter stresses civic engagement in identi-
fying and prioritizing a broad range of neighborhood problems, while
the former frequently focuses on patterns of traditionally defined crimes
that are identified using police data systems. Problem-oriented policing
 The promise of community policing 35

sometimes involves community members or organizations in order to

address particular issues, but more often it is conducted solely by spe-
cialized units within the police department. On the other hand, commu-
nity policing involves neighborhood residents as an end in itself, and in
evaluation terms it is important to count this as a “process success.” The
problem with relying on the data that is already in police computers is that
when residents are involved they often press for a focus on issues that are
not well documented by department information systems, such as graf-
fiti, public drinking, and building abandonment. Effective programs must
have systems in place to respond to a broad range of problems, through
partnerships with other agencies. The 2000 survey found that in cities
of more than 250,000 residents, more than 50 percent of departments
reported they had formed problem-solving partnerships with community
groups and local agencies (Bureau of Justice Statistics 2003).
Is this easy to do? It is at least as hard as involving the community, for
bureaucracies are involved, and interagency cooperation can easily fail.
For a long list of familiar bureaucratic and political reasons, other city
and state agencies usually think that community policing is the police
department’s program, not theirs. They resist bending their own profes-
sional and budget-constrained priorities to accommodate police officers
who call on them for help. Making this kind of interorganizational coop-
eration work turns out to be one of the most difficult problems facing
innovative departments. When the chief of an East Coast city was new,
he told me that he could handle things in his department; his biggest
fear was that his mayor might not handle the city’s other agencies, and
that they would not provide the kind of support that community policing
requires. If community policing is the police department’s program, it
may fail. Community policing must be the city’s program.
It is also hard to involve police officers in problem solving. Cordner
and Biebel (2005) did an in-depth study of problem-solving practice in a
major American city. Although the department had been deeply commit-
ted to problem solving for more than fifteen years, they found that street
officers typically defined problems very narrowly (e.g., one address, or
one suspected repeat offender); their analysis of it consisted of making
personal observations from their car; they crafted solutions from their own
experience; and two-thirds of the time their proposed solution did not go
past arresting someone. The study concluded that, after fifteen years of
practice, this department’s “glass” was only “half full.” What observers
would classify as “full-scale” problem solving was rarely encountered.
Even the advocates of problem solving (you will hear from them in later
chapters) admit that it requires a great deal of training, close super-
vision, and relentless follow-up evaluation to make it work. However,
36 Wesley G. Skogan

one important organizational function that often gets shortchanged is

training. Training is expensive and officers have to be removed from the
line – or paid overtime – to attend. And few departments are adequately
staffed with supervisors who themselves were full-fledged problem solvers
(Eck 2004).
Community policing has also revived interest in systematically address-
ing the task of crime prevention. In the traditional model of policing,
crime prevention was deterrence based. To threaten arrest, police patrol
the streets looking for crimes (engaging in random and directed patrol),
they respond quickly to emergency crime calls from witnesses and vic-
tims, and detectives then take over the task of locating offenders. Con-
cerned residents, on the other hand, do not want the crime that drives
these efforts to happen in the first place. Their instinct is to press for true
prevention. Police-sponsored prevention projects are in place through-
out the country. Problem solving has brought crime prevention theories
to the table, leading police to tackle the routine activities of victims and
the crucial roles played by “place managers” such as landlords or shop-
keepers, and not just offenders (Eck and Wartell 1998; Braga, Weisburd,
Waring et al. 1999). When community policing came to Chicago, one
of the first actions of a new district commander was to convince a bank
to open an ATM machine in his police station, so residents had a safe
place to go to transact business. An emphasis on “target hardening” has
gotten police involved in conducting home security surveys and teaching
self-defense classes. But when communities talk about prevention they
mostly talk about their children, and ways of intervening earlier with
youths who seem on a trajectory toward serious offending. Much of the
work preventing the development of criminal careers lies with agencies
besides the police, including family courts, children’s protection agen-
cies, parents, peer networks, and schools. To their efforts the police add
involvement in athletic and after school programs, DARE (Drug Aware-
ness and Resistance Education) presentations in schools, special efforts
to reduce violence in families, and initiatives that focus attention on the
recruitment of youths into gangs.

Decentralization
Decentralization is an organizational strategy that is closely linked to the
implementation of community policing. Decentralization can be pursued
at two levels. Typically, more responsibility for identifying and responding
to chronic crime and disorder problems is delegated to mid-level com-
manders in charge of the geographical districts or precincts that make
up a city. Departments have had to experiment with how to structure
 The promise of community policing 37

and manage a decentralization plan that gives mid-level managers real

responsibility, and how to hold them accountable for measures of their
success. Here community policing intersects with another movement in
policing (and the subject of another pair of chapters in this book), the
emergence of a culture of systematic performance measurement and man-
agerial accountability.
The idea is to devolve authority and responsibility further down the
organizational hierarchy. Departments do this in order to encourage the
development of local solutions to locally defined problems, and to facili-
tate decisionmaking that responds rapidly to local conditions. The police
are not independent of the rest of society, where large organizations in
both the public and private sectors have learned that decentralization can
create flexibility in decisionmaking at the customer contact level. There
may be moves to flatten the structure of the organization by compressing
the rank structure, and to shed layers of bureaucracy within the police
organization to speed communication and decisionmaking. In Chicago,
most of the department’s elite units – including detectives, narcotics
investigators, special tactical teams, and even the organized crime unit –
are required to share information and more closely coordinate their work
with the geographical districts. The department’s management account-
ability process calls them on the carpet when they fail to serve as “support
units” for uniformed patrol officers (Skogan et al. 2003). To flatten the
organization, Chicago abolished the civil service rank of captain, leaving
the department with just three civil service ranks (Skogan and Hartnett
1997).
At the same time, more responsibility for identifying and responding
to community problems may be delegated to individual patrol officers
and their sergeants, who are in turn encouraged to take the initiative
in finding ways to deal with a broad range of problems specific to the
communities they serve. Structurally, community policing leads depart-
ments to assign officers to fixed geographical areas, and to keep them
there during the course of their day. This is known as adopting a “turf
orientation.” Decentralization is intended to encourage communication
between officers and neighborhood residents, and to build an awareness
of local problems among working officers. They are expected to work
more autonomously at investigating situations, resolving problems, and
educating the public. They are being asked to discover and set their own
goals, and sometimes to manage their work schedule. This is also the
level at which collaborative projects involving both police and residents
can emerge. By 1999, a national survey of police departments found
that assigning officers geographically was virtually the norm in cities over
250,000 (Bureau of Justice Statistics 2001).
38 Wesley G. Skogan

This pattern of dual decentralization is adopted not only so that police

can become more proactive and more preventive, but also so that they can
respond efficiently to problems of different magnitude and complexity.
Under the professional model, marching orders for the police traditionally
come from two sources: 911 calls from the public concerning individual
problems, and initiatives or programs originating at police headquarters
or even City Hall. Every experienced officer can tell stories of the crazy
things officers sometimes have to do because “downtown” announced a
citywide initiative that was irrelevant for their district. A Chicago com-
mander once described to me how he was punished (he lost a day’s pay)
because – as a district commander – he assigned two officers to identify-
ing abandoned cars and getting them towed, rather than the maximum
of one officer that the rule book mandated. He used this story to good
effect whenever officers complained in a meeting that the department was
getting away from its traditional practices because of community policing.
Decentralization, paired with a commitment to consultation and
engagement with local communities, also allows the police to respond
to local problems that are important to particular communities. Police
were not organized to respond to the organized groups and community
institutions that make up “civil society.” Now surveys of departments
indicate that, as part of a community policing initiative, virtually all larger
departments now consult local advisory boards representing specific
communities.
Is decentralization easy to pull off? It is at least as hard as problem solv-
ing, and politically risky to boot. For all of the adoption of specific pro-
grams, researchers who track trends in police organization are skeptical
that there has been much fundamental “flattening” of police hierarchies
– which is, after all, about their jobs (Greene 2004). Resistance to reform
does not just come from the bottom of the organization. Junior execu-
tives at police headquarters may resist authority being taken from them
and pushed to lower levels in the organization. Managers at this level are
in a position to act as a filter between the chief and operational units,
censoring the flow of decisions and information up and down the com-
mand hierarchy (for a case study of how this can undermine community
policing initiatives, see Capowich 2005). This is one reason why special
community policing units are often run from the chief’s office, or housed
in a special new bureau – this enables the department to get neighborhood
officers on the street while bypassing the barons who dominate key posi-
tions at headquarters. Too often they are command-and-control oriented
and feel most comfortable when everything is done by the book. Discus-
sions of community policing often feature management buzz words like
 The promise of community policing 39

“empowerment” and “trust,” and this makes them nervous because they
also worry about inefficiency and corruption.
And, of course, these concerns are real. One of the dilemmas of com-
munity policing is that calling for more operational and street-level dis-
cretion runs counter to another trend in policing, which is to tighten the
management screws tighter and create an increasingly rule-bound envi-
ronment in order to control police corruption and violence. Ironically,
many of the recent innovations discussed in this book go the other way;
they recognize, widen, and celebrate the operational independence of
individual officers. Community policing recognizes that problems vary
tremendously from place to place, and that their causes and solutions
are highly contextual. We expect police to use “good judgment” rather
than somehow enforce “the letter of the law.” Decentralizing, reducing
hierarchy, and granting officers more independence, and trusting in their
professionalism are the organizational reforms of choice today, not tight-
ening things up to constrain officer discretion. But police do misuse this
discretion, and they do take bribes.
It may be difficult to pull off decentralization to the turf level because
it takes too many people. Community policing is labor-intensive, and
may require more officers. Police managers and city leaders will have
to find the officers required to staff the program. Finding the money
to hire more officers to staff community policing assignments is hard,
so departments may try to downsize existing projects. This can bring
conflict with powerful unit commanders and allied politicians who sup-
port current arrangements. Police departments also face “the 911 prob-
lem.” Their commitment to respond to 911 calls as quickly as possible
dominates how resources are deployed in every department. Community
policing has encountered heavy political resistance when the perception
arose (encouraged to be sure by its opponents) that resources previously
devoted to responding to emergency calls were being diverted to this
“social experiment.”
Decentralization is also difficult to manage because evaluation of the
effectiveness of many community policing initiatives is difficult. The
management environment in policing today stresses “accountability for
results” (Weisburd, Mastrofski, McNally et al. 2003). Units are not
rewarded for their activities, however well meaning, but for declining
crime. However, the public often wants action on things that department
information systems do not account for at all. In decentralized depart-
ments, residents of different neighborhoods make different demands on
police operations. They value the time officers spend meeting with them,
and they like to see officers on foot rather than driving past on the way
40 Wesley G. Skogan

to a crime scene. As a result, both individual and unit performance is

harder to assess in community policing departments (Mastrofski 1998).

Prospects
An unanswered question about community policing is whether it can sur-
vive the withdrawal of federal financial support and attention. Under the
1994 Violent Crime and Law Enforcement Act, the federal government
spent billions of dollars to support community policing. Federal agen-
cies sponsored demonstration projects designed to spur innovation and
promote the effectiveness of community policing, and they promoted it
heavily through national conferences and publications. The Act specified
that one of the roles of these new officers should be “to foster problem
solving and interaction with communities by police officers.” Innova-
tions such as community policing highlight the importance of training
for officers, and the 1994 crime act also funded the creation of regional
community policing centers around the country. By 1999, 88 percent of
all new recruits and 85 percent of serving officers worked in departments
that were providing some community policing training (Bureau of Justice
Statistics 2001).
The issue is whether police departments will continue to staff their
community policing components. Federal financial support for commu-
nity policing certainly is on the wane. Now crime is down, a new team is
in the White House, and federal largesse toward local law enforcement
is being redirected to post-September 11 concerns. Even where commit-
ment to community policing is strong, maintaining an effective program
can be difficult in the face of competing demands for resources. For exam-
ple, between 1995 and 2003, the city of Cleveland received $34 million in
federal assistance for hiring police officers, but for 2004 that figure shrank
to $489,000, and the city expected to receive even less in 2005. To han-
dle the shortfall they cut 250 officers from the payroll and closed the
neighborhood mini-stations that were created as part of the city’s com-
munity policing effort (Butterfield 2004). There is also pressure from
the federal government to involve local police extensively in enforcing
immigration laws. This is being stoutly resisted by many chiefs of police,
who claim that it would be a great setback to their community involve-
ment and trust-building projects. We shall see if they can continue to
resist.
A second issue is whether community policing can survive accountabil-
ity management. This is another new thing in policing, and many of its
features push in the opposite direction. Community policing continues
to ask officers to think and act in new and unaccustomed ways, and many
 The promise of community policing 41

of its presumed benefits do not show up in police information systems.

To a significant extent, in this new management environment what gets
measured is what matters. Top managers decide what is a success, and
hold mid-level managers to their standards. The accountability process
is about harnessing the hierarchy to achieve top management’s objec-
tives, which are in turn driven by the data they have at hand, and those
data say little about community priorities. The thrust of New York City’s
Compstat and similar management initiatives all over the country is that
measured accomplishments get attention and unmeasured accomplish-
ments do not. As a result, there is a risk that the focus of departments
will shift away from community policing, back to the activities that bet-
ter fit a recentralizing management structure driven by data on recorded
crime.
Community policing also stresses the importance of developing the
general purpose skills of line officers through education and training,
and it frequently features talk about empowering rank-and-file employees
and encouraging them to act autonomously. It stresses that workers at the
very bottom of the organization are closest to the customer, and are to
use their best judgment about how to serve the neighborhoods where
they are assigned. However, these are at best low priorities for Compstat-
style accountability management. Community policing is an attack on
the traditional hierarchical structure of police departments. It calls for
the bottom-up definition of problems. Police researchers attribute many
of the problems of contemporary policing to the mismatch between the
formal hierarchical structure of police organizations and the true nature
of their work, which is extremely decentralized, not amenable to “cookie
cutter” solutions, dependent on the skills and motivation of the individual
officers handling it, and mostly driven externally by 911 calls rather than
management strategies. Perhaps the accountability process has ridden to
the rescue of the traditional hierarchical structure, trying again to impose
that hierarchy on work that does not fit its demands. Is the accountability
process the last refuge of the command-and-control mentality of the past,
and can community policing survive it?
The final question is whether community policing can live up to its
promises. Like many new programs, its adoption in many instances pre-
ceded careful evaluation of its consequences. The effectiveness of com-
munity policing has been the subject of some research, ranging from its
impact on crime to how openly it is embraced by the officers charged with
carrying it out. There has not been enough research to definitively address
the effectiveness question. As this chapter has documented, implement-
ing a serious community policing program is risky and hard, and depart-
ments can fail at it.
42 Wesley G. Skogan


Braga, A. A., Weisburd, D. L., Waring, E. J., Mazerolle, L. G., Spelman, W., and
Gajewski, F. (1999). Problem-oriented policing in violent crime places: A
randomized controlled experiment. Criminology, 37, 541–580.
Bureau of Justice Statistics. (2001). Community policing in local police departments
1997 and 1999. Washington, DC: Bureau of Justice Statistics, US Depart-
ment of Justice.
(2003). Local police departments 2000. Washington, DC: Bureau of Justice Statis-
tics, US Department of Justice.
Butterfield, F. (2004). As cities struggle, police get by with less. New York Times,
July 27.
Capowich, G. E. (2005). A case study of community policing implementation:
Contrasting success and failure. In K. R. Kerley (ed.), Policing and program
evaluation. Upper Saddle River, NJ: Prentice Hall.
Cordner, G. and Biebel, E. (2005). Problem-oriented policing in practice. Crim-
inology and Public Policy, 4, 155–180.
Eck, J. E. (2004). Why don’t problems get solved? In W. G. Skogan (ed.), Com-
munity policing: Can it work? Belmont, CA: Wadsworth.
Eck, J. E. and Wartell, J. (1998). Improving the management of rental properties
with drug problems: A randomized experiment. In L. G. Mazerolle and
J. Roehl (eds.), Civil remedies and crime prevention, 9 (pp. 161–183). Monsey,
NY: Criminal Justice Press.
Greene, J. R. (2004). Community policing and organization change. In Wesley
G. Skogan (ed.), Community policing: Can it work? (pp. 30–53). Belmont,
CA: Wadsworth.
Grinc, R. M. (1994). “Angles in marble”: Problems in stimulating com-
munity involvement in community policing. Crime and Delinquency, 40,
437–468.
Lurigio, A. and Rosenbaum, D. (1994). The impact of community policing
on police personnel. In D. P. Rosenbaum (ed.), The challenge of community
policing: Testing the promises. Thousand Oaks, CA: Sage Publications.
Mastrofski, S. D. (1998). Community policing and police organization structure.
In J. Brodeur (ed.), Community policing and the evaluation of police service
delivery (pp. 161–189). Thousand Oaks, CA: Sage Publications.
Moy, P. and Pfau, M. (2000). With malice toward all? The media and public confi-
dence in democratic institutions. Westport, CT: Praeger Publishing.
Pate, A. M. and Shtull, P. (1994). Community policing grows in Brooklyn: An
inside view of the New York Police Department’s model precinct. Crime and
Delinquency, 40, 384–410.
Sadd, S. and Grinc, R. (1994). Innovative neighborhood oriented policing: An
evaluation of community policing programs in eight cities. In D. P. Rosen-
baum (ed.), The challenge of community policing: Testing the promises (pp. 27–
52). Thousand Oaks, CA: Sage Publications.
Sampson, R. J., Raudenbush, S., and Earls, F. (1997). Neighborhoods and
violent crime: A multilevel study of collective efficacy. Science, 277, 918–
924.
 The promise of community policing 43

Skogan, W. G. (1988). Community organizations and crime. In M. Tonry and

N. Morris (eds.), Crime and justice: An annual review, 10. Chicago: University
of Chicago Press.
(1990). Disorder and decline: Crime and the spiral of decay in American cities. New
York: The Free Press.
(2004). Community policing: Common impediments to success. In L. Fridell
and M. A. Wycoff (eds.), Community policing: the past, present, and future.
Washington, DC: The Annie E. Casey Foundation and the Police Executive
Research Forum, 159–168.
Skogan, W. G. and Hartnett, S. (1997). Community policing, Chicago style. New
York: Oxford University Press.
Skogan, W. G. and Steiner, L. (2004). Community policing in Chicago, year ten.
Chicago: Illinois Criminal Justice Information Authority.
Skogan, W. G., Hartnett, S., DuBois, J., Comey, J., Kaiser, M., and Lovig, J.
(1999). On the beat: Police and community problem solving. Boulder, CO: West-
view Press.
Skogan, W. G., Steiner, L., Hartnett, S., DuBois, J., Bennis, J., Rottinghaus, B.,
Kim, S., Van, K., and Rosenbaum, D. (2003). Community policing in Chicago,
years eight and nine. Chicago: Illinois Criminal Justice Information Authority.
Weisburd, D., Mastrofski, S., McNally, A. M., Greenspan, R., and Willis, J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2, 421–456.
3 Critic
Community policing: a skeptical view

Stephen Mastrofski

About two decades ago a few American police leaders caught the wave of
community policing reform, and now just about everybody’s going surf-
ing. Early interest in this reform can be attributed to police anxieties about
skyrocketing crime and urban violence, unmet rising expectations from
the civil rights movement, and middle-class alienation from government
authority (Fogelson 1977: ch. 11). Like a “perfect storm” these forces
converged in the 1960s, stimulating intense criticism from blue ribbon
commissions and a daily drumbeat of negative press for police. Calls for
change were issued, some of them radical: community control of polic-
ing, deprofessionalization, and reassignment of some core police tasks to
other government agencies and the private sector. Alarmed and intent
on ending this crisis of legitimacy (LaFree 1998), progressive police and
scholars began to experiment with ways to make American police both
more effective and more democratic without losing many of the advances
made in policing in the previous half century. The resulting “community
policing” reforms were influenced by fashions in organization develop-
ment intended to make them less bureaucratic, more responsive to the
“customer,” and more results-oriented (Mastrofski 1998; Mastrofski and
Ritti 2000).
Unlike some reforms that narrow their focus over time, commu-
nity policing has remained multifaceted and diverse. Some departments
emphasize broken windows policing, others feature problem-oriented
policing and still others stress policing that establishes stronger police–
public partnerships (Mastrofski, Worden, and Snipes 1995: 540–541).
Much of the reform literature suggests that all of these are desirable and
compatible. Whether or not that is so, it is clear that different elements of
community policing appeal to different audiences, and it has led to fruit-
less debates over what community policing “really” means. In fact, it is
this ambiguity and flexibility that gives community policing its all-things-
to-all-people character and has contributed to its political viability over
two decades, embraced by public leaders across the political spectrum.
However, this diversity makes it difficult to conduct a rigorous inquiry

44
 Community policing 45

into the merits of community policing, a finding noted by the panel on

police policies and practices organized by the National Research Coun-
cil to assess, among other things, this reform (Committee to Review
Research 2004: 232–233). It concluded that community policing was
simply too amorphous a concept to submit to empirical evaluation and
recommended that researchers evaluate it by breaking it down into more
specific components.
The editors of this volume seem to have taken the NRC panel’s
recommendation; they have organized chapters around a number of
specific reforms that have in fact been widely construed as impor-
tant elements of community policing: problem-oriented policing, broken
windows policing, third-party policing, offender-focused policing, and
hot spots policing. Therefore, I will limit my discussion of community
policing to those efforts that seek to link the police more closely to the
community in “partnership” arrangements: joint activities to coproduce
services and desired outcomes, giving the community a greater say in
what the police do, or simply engaging with each other to produce a
greater sense of police–community compatibility.
I will fulfill my role as community policing critic by offering a skep-
tical view. The skeptic is a thoughtful doubter; doubting is a methodol-
ogy, not a conclusion. To the extent that community policing is a good
thing, it should withstand the careful scrutiny of thoughtful doubters. I
will ask four questions about community policing in the United States.
How has the community policing reform movement changed what the
public expects of its police? How has the community policing movement
changed the way that policing is organized and performed? How, if at
all, has community policing produced beneficial public outcomes – less
crime and disorder, more or better service, more social capital, and more
equitable service delivery? What impact has community policing had on
police legitimacy?

The public’s expectations of the police

Reform movements are built on promises that may stimulate expecta-
tions not previously held widely. They can fuel change because “Nothing
happens unless first a dream” (Sandburg 1970). Community policing
begins with a distinct advantage, for its advocates promise to deliver
what the public has long dreamed of but received only in middle-class
suburbs: client-oriented, “service-style” policing (Wilson 1968). So it
seems unlikely that community policing has changed what folks want,
but it may well have altered that to which a good many feel entitled, espe-
cially those segments of society that were heretofore least likely to enjoy
46 Stephen Mastrofski

it, the economically and socially disadvantaged. The majority of Amer-

icans are probably aware in a general way of community policing and
its promises (Harris Poll 1999), but most probably cannot articulate the
finer points of community policing programs. Nonetheless, it appeals to
powerful political and cultural themes of our time that have to do with
a blend of both communitarian and good-government ideals (Manning
1984; Crank 1994; Lyons 1999: 18–25).
There has been virtually no research on what citizens expect of their
police that would enable us to draw conclusions about the impact of com-
munity policing. Here we must rely upon a bit of indirect evidence and
a great deal of speculation. The promises of community policing reform
have steadily rained down on the American public over the last decade,
leaving us with a reasonable, if untested, assumption that they have had
at least some impact. Community policing has been a significant pres-
ence in many local news outlets around the nation, and in virtually all of
them the picture of community policing has been unremittingly positive
(Mastrofski and Ritti 1999; Chermak and Weiss 2002). The most fre-
quently mentioned theme in community policing stories is the police–
community partnership dimension (Mastrofski and Ritti 1999), followed
by bringing more resources to the local police, and producing tangi-
ble results. This provides some fuel for speculation about the “field of
dreams” that community policing advocates are building in the minds
of the American public. It has two basic elements: police responsiveness
and results. Community policing promises greater police responsiveness
that takes the form of accessibility, the police knowing and appreciating
what citizens want, and better prospects in the competition for police ser-
vices. Deeply embedded in responsiveness is also an expectation of civility
and caring, concerns that were front and center in the legitimacy crisis
of the 1960s (Reiss 1971). Under results I include safer, easier to live
in communities and a stronger “community” through shared values and
collective action. If asked to articulate what they expect from community
policing, citizens would give these themes dominance. They have been
nicely summarized as a yearning to “get back to Mayberry” (Lyons 1999:
163).
If the public has come to embrace community policing’s promises as
entitlements, what does that mean for the future of community policing?
Pre-existing expectations do significantly color how citizens evaluate the
policing they receive (Reisig and Chandek 2001), but as Shakespeare
noted, “Oft expectation fails and most oft there where most it promises.”
Community policing reformers have encouraged the public to expect so
much more, which makes it all the more imperative for police organiza-
tions to demonstrate results.
 Community policing 47

The organization and practice of policing

Community policing has been well marketed, but to what extent has it
transformed the nature of policing in a nation where local influences
predominate? I will consider three ways in which police organizations
might be transformed: the adoption of new programs and altered struc-
tures, changes in the philosophy and culture of the rank-and-file, and
changes in how the police practice their work at the street level.

Community policing programs and structures

We expect two kinds of programmatic and structural changes from
community policing. First, it should broaden police organization goals
because it should increase responsiveness to the much wider range of
services citizens presumably want from their police compared to the
much narrower confines imposed by the “professional” model. Sec-
ond, it should alter the way police are organized to accomplish their
goals.
If we were to measure changes in organization goals by observing police
departments’ mission statements and strategic plans, we would undoubt-
edly conclude that community policing has had a major transformative
effect. By 2000 more than 80 percent of large police departments and
more than 60 percent of small ones reported that they had incorporated
community policing values in their mission statements (Roth, Roehl, and
Johnson 2004: 20). However, mission statements and plans by them-
selves tell us little about the goals that the organization really enacts.
We learn more about the organization’s commitment to the community-
oriented mission by examining such things as how the organization rou-
tinely evaluates itself, how it evaluates and rewards its officers, and how
it prepares them to answer the demands of community policing. What
little we know about the transformation of police organizations in these
domains is not particularly encouraging. The vast majority of depart-
ments still expend most of their energies tracking traditional crime and
enforcement statistics, rather than developing performance systems that
track neighborhood quality of life and problem-solving (Weisel and Eck
1994: 66; Skogan and Hartnett 1997: 241; Greene 2004: 39). Perhaps
most telling, recruit training has not been substantially revised to promote
community policing.1
Despite reformers’ admonition that community policing cannot be
defined solely as a set of programs, program implementation has been
a common strategy for accomplishing its goals. Neighborhood Watch,
citizen police academies, citizen surveys, and the establishment of
48 Stephen Mastrofski

community policing units are some examples of programs that are often
appended to the existing organization (Roth et al. 2004: 8). In addition, a
department may adopt certain structures thought to facilitate police–
community partnerships, such as drawing beat boundaries to coincide
with neighborhood boundaries, permanent assignment of officers to
beats, and delegating decisionmaking to the lowest level to enable a more
fine-tuned degree of responsiveness to community preferences and needs
(Roth et al. 2004: 18–21). The adoption of such programs has been vari-
able by type of program, but nonetheless widespread across the American
landscape (Maguire and Mastrofski 2000; Roth, Ryan, Gaffigan et al.
2000; Maguire and Katz 2002). Most importantly, in a relatively short
time the number of police agencies adopting such programs has grown
significantly (Roth et al. 2004). However, these types of studies tend to
indicate that the partnership aspects of community policing are the most
weakly implemented (Maguire and Katz 2002; Maguire and King 2004;
Roth et al. 2004).
The above evidence is based exclusively on police agency self-reports
to the Office of Community Oriented Policing Services (COPS) and from
mail surveys. Questions about the validity of such data naturally arise. The
popularity of community policing creates a strong incentive to skew the
department’s self-presentation in favor of program adoption, especially
since 1994, when large federal grants from COPS became available to
departments willing to show a commitment to implementing community
policing (Maguire 2002: 52–53; Maguire and Katz 2002: 513). And even
if a program has been adopted, there is the question of “dosage,” how
extensive and intensive the program is over time and space (Maguire and
Mastrofski 2000; Roth et al. 2004: 4). These claims about community
policing adoption, measured from a distance, through the eyes of inter-
ested parties, and without much precision or independent validation,
should be interpreted as an indication of the desirability of community
partnership programs and structures, or the desirability of the appearance
of these programs.
We might have greater faith in indirect measures of structural change
in police organizations, debureaucratizing changes thought instrumental
to the accomplishment of community policing, but based on measures
arguably less susceptible to the reactivity of the desirability of reform
(Maguire 1997; Mastrofski 1998; Mastrofski and Ritti 2000; Greene
2004). The most comprehensive analysis found mixed results between
1987 and 1998 (Maguire, Shin, Zhao, and Hassell 2003). Large munic-
ipal police agencies moved significantly toward the community polic-
ing ideals of decentralization, lower administrative intensity, and greater
civilianization. Some aspects of spatial differentiation did increase (more
police stations and mini-stations), but the number of beats did not. On
 Community policing 49

the other hand, hierarchical flattening did not occur, nor did hierarchical
segmentation, and vertical differentiation continued to increase. Police
agencies also failed to become less formalized and less functionally differ-
entiated. This hardly constitutes evidence of a revolution in the structure
of policing.
A more valid approach to measuring community policing program
implementation is on-site observation, but here we lack a sufficiently
large, representative sample of organizations monitored over time. What
we have instead are a small collection of in-depth case studies and
cross-sectional comparisons of departments, most of which were selected
because they were thought to exemplify good community policing imple-
mentation. Some of these have claimed successes in implementing com-
munity policing (Roth et al. 2000: ch. 7), but a substantial number
have also documented failures and disappointments (Greene, Bergman,
and McLaughlin 1994; Sadd and Grinc 1994; Tien and Rich 1994;
Rosenbaum and Wilkinson 2004).
Skeptics should be prepared to accept that some departments have
made changes in programs and structures, moving them closer to the
ideals of community policing reformers. But there is reason to think that
many, and probably most, American police departments fall far short
of those few leaders. Some have understood this sort of phenomenon
from the perspective of “institutional” organization theory (Meyer and
Rowan 1977; DiMaggio and Powell 1983). One version of this theory
distinguishes early adopters of a reform from later adopters (Tolbert and
Zucker 1983). Early adopters are driven to undertake new structures
and programs to meet the demands of their technical environment (e.g.,
high or rising crime rates). These pioneers experiment and are willing to
take risks by innovating to improve their technical performance in such
domains as reducing crime and improving the quality of life. But for
industries such as policing, that do not use validated technologies that
are well proven to produce the desired results, the further diffusion of
these innovations among the vast majority of police organizations depends
primarily upon their becoming accepted as the “right” thing to do –
entirely independent of the quantity and quality of empirical evidence on
their ability to accomplish desired goals (Scott 1992). Once important
leaders in policing begin to adopt community policing as the “right” way
to do things, and once political leaders embrace it, rigorous scientific
evidence regarding its performance becomes irrelevant as organizations
scramble to adopt the programs and structures of the early adopters. This
may account for the success of programs such as DARE, which flourished
before any evidence was available on its ability to prevent illicit drug use
and which continues to flourish in the face of evidence that it provides no
prevention benefits (Rosenbaum and Hanson 1998; Gottfredson, Wilson,
50 Stephen Mastrofski

and Najaka 2002). In fact, it accounts for the phenomenon that strikes
some researchers as curious: that billions of dollars have been spent on
community policing reforms and millions on research on community
policing, but there is so little rigorous evidence on its effects on crime
and disorder (Weisburd and Eck 2004).
Some researchers have applied institutional theory to policing (see
Crank 2003 for a review), but there is relatively little evidence on the
early-versus-late-adopter diffusion model. Ritti and Mastrofski (2002)
found evidence consistent with that model in their examination of arti-
cles on community policing in professional police journals. Based on a
nationally representative mail survey, Moon (2004) examined the extent
to which community policing programs were adopted in the late 1990s,
comparing the influence of technical and institutional pressures on early
and late adopters. While not a complete confirmation of institutional
theory’s application to the diffusion of community policing, his research
strongly supported the greater power of institutional over technical influ-
ences as the driving force behind community policing adoption at this
stage.

Community policing philosophy and culture

“Community policing is a philosophy, not a program” (Roth et al. 2000:
183) is a popular phrase among academics and practitioners. From this I
infer that community policing is implemented only when police officers in
large numbers adopt an appropriate set of values; whether or not commu-
nity policing has been externalized into programs and structures, it has
not been implemented until police officers have internalized it. Depend-
ing on how one defines “philosophy,” this can be a very low standard or
a very high one. If a philosophy means what one thinks, feels, or avers
(Sparrow, Moore, and Kennedy 1990: ch. 5), then the implementation
standard is rather low. It is well established in social psychology and the
study of police that values or philosophies measured in these ways are
weak predictors of actual practice (Ajzen 1987; Worden 1989; Mastrof-
ski and Parks 1990; Snipes and Mastrofski 1990; Terrill and Mastrofski
2004). Nonetheless, police attitudes about community policing in gen-
eral, and specifically the partnership aspects of community policing are
by some accounts overwhelmingly positive (e.g., Weisburd, Greenspan,
Hamilton et al. 2001: 25–26). Such surveys ask officers to assess commu-
nity policing principles in the abstract, but it is the particulars of officers’
experiences in their departments that have the greatest meaning for how
they act, and here the level of rank-and-file skepticism is much higher
 Community policing 51

(Sadd and Grinc 1994; Zhao, He, and Lovrich 1999; Schafer 2001: ch. 9;
Rosenbaum and Wilkinson 2004).
If we take the philosophy-not-a-program (PNAP) perspective to mean
that community policing must not only be embraced, but that it must
guide practice, then we set a much higher standard. Some evidence
suggests that officers who claim to embrace community policing values
also behave differently from those who do not (Mastrofski et al. 1995;
Mastrofski, Snipes, Parks, and Maxwell 2000; Terrill, Paoline, and Man-
ning 2003), but we note that far more powerful are the structural and
programmatic elements (e.g., whether the officer has been given a com-
munity policing assignment). On the whole, a pro-community policing
attitude does not appear to be a powerful predictor of officer behavior.
“Probably the biggest obstacle facing anyone who would implement
a new strategy of policing is the difficulty of changing the ongoing cul-
ture of policing” (Moore 1992: 150). A skeptic should wonder about
the validity of this claim, a fellow traveler to PNAP, because it may have
the causal mechanism of behavioral change backwards. It is like blam-
ing the patient’s heart attack on his failing heart, rather than his diet,
exercise, and genes. In the same way, changes in the occupational police
culture may be less the cause of community policing practices than the
consequence of structural changes, focusing on occupational culture as the
change directs attention away from what might shape practice.
There is some evidence to suggest that a monolithic police subculture
opposed to the principles of community policing may have never existed
(Muir 1977; Worden 1995) and even if it once did, it is now splintered into
many different subgroups with varying degrees of support and opposition
to community policing (Cochran and Bromley 2003; Wood, Davis, and
Rouse 2004). Over the last three decades hiring practices have changed,
increasing the proportion of officers who are female, minority, and col-
lege educated (Committee to Review Research 2004: 79). These sorts
of people may be more receptive to community policing (Weisburd et al.
2001: 31; Cochran and Bromley 2003: 102), but most research also shows
that police socialization processes exert far more influence over officers’
behavior than hiring practices (Committee to Review Research 2004:
ch. 4).
What are the structural determinants of police culture, and to what
extent have they been transformed to produce the expected changes in
police practice? The principal management tool for shaping police prac-
tices is training, and, unfortunately, we have little rigorous evidence on the
impact of police training in general, much less that of community policing
training (Committee to Review Research 2004: 141–147). Of the handful
52 Stephen Mastrofski

of studies that look at the effects of training relevant to community polic-

ing, some indicate that it does yield more community policing practices
and about an equal number indicate that it has no effect. Critics of the
usual community policing training have argued that it wrong-headedly
seeks to change officers’ values and beliefs rather than giving them skills
they can use to promote values they already possess (Buerger 1998; Haar
2001). Perhaps a more fundamental problem with community policing
training is that many departments may use it as the sole or principal
mode of changing police culture, and when its message is not reinforced
by supervisors, managers, and policies that determine how performance
is measured, monitored, and rewarded, there can be little hope that any
fruit born of the training will but die on the vine. Indeed, police depart-
ments seem to invest more energy into converting the rank-and-file to
a community policing philosophy than they do motivating supervisors
and managers to promote community policing and giving them the skills
and resources needed to do it. Large-scale studies are not available to
document the scope of this problem, but this is a theme of case stud-
ies reporting disappointing results in the implementation of community
policing (Schafer 2001; Rosenbaum and Wilkinson 2004). Even the expe-
riences of Chicago suggest that getting the training right is no easy matter,
and getting the hierarchy capable and on board is challenging too (Sko-
gan and Hartnett 1997). Creating the proper structural environment for
community policing is then no slam dunk, and there is good reason to
suspect that large numbers of departments are not even getting close to
the basket.
Police reformers have long emphasized the role of the top executive
in shaping the culture and ultimately the practices of the police agency
(Goldstein 1990; Sparrow et al. 1990; Moore and Stephens 1991). I have
elsewhere expressed skepticism about the capacity of the police chief to
transform the organization in this sort of way (Mastrofski 2002), but we
lack objective and rigorous empirical tests of this proposition. We are not
without autobiographical accounts of leadership success (Bratton 1998;
Kerik 2001), albeit to create organizational cultures that tried to undo
the brand of community-partnership policing that is the focus of this
chapter, but these accounts are clearly not the product of disinterested,
rigorous evaluation.

Community policing at the street level

The greatest discretion in the delivery of police services rests with the
lowest-ranking officers, so the ultimate test of the reform’s impact will
be observed at the street level. This section is organized according to
different types of practices that community policing, as construed in this
 Community policing 53

chapter, is supposed to promote. I consider whether officers who receive

or show some form of community policing “dosage” (e.g., community
policing specialist assignment, training, positive attitude toward commu-
nity policing) to a greater extent than others engage in more of the desired
practice.
Engaging the community. One of the great attractions of community
policing is its promise to bring officers face-to-face with the public – to
learn more about people and their problems, solve them, or at least com-
fort them when they cannot be solved. Permanent beat assignments, foot
and bike patrol, mini-stations, and park-and-walk tactics are designed to
increase “face time” between officers and their clientele. The philoso-
phy of community policing highly values these interactions. Systematic
observations of community policing present disappointing results. The
data from a handful of departments suggest that the daily routine of both
patrol generalists and specialists does not involve heavy engagement with
the public, especially those in the neighborhoods to which officers are
assigned.2
Many expect that community policing will produce a decline in the
use of arrest and physical coercion. Some systematic field research shows
that the style of community policing embraced by management may influ-
ence the use of verbal and physical coercion (Terrill and Mastrofski 2004:
127), and another study suggests that officers who embrace community
policing make fewer arrests (Mastrofski et al. 1995: 552), but in the first
case the researchers were unable to illuminate the mechanisms by which
management achieved these results, 3 and in the second they were unable
to determine whether making fewer arrests was a good thing (Mastrofski
2004). A study of situations where officers were seeking citizen compli-
ance showed that officers positively disposed to community policing were
not statistically distinguishable from negatively disposed officers in their
resort to coercion (Mastrofski, Snipes, and Supina 1996: 291). The lim-
ited evidence suggests that community policing has not had profound
effects on the use of coercion.
Aside from improving the quantity and quality of police–citizen
encounters on the street, many departments employ programs to increase
police–citizen interaction in settings that focus on the discussion of pri-
orities, exchanging information and ideas, and planning for joint citizen–
police problem-solving ventures. Perhaps the most earnest and sustained
effort of this sort, and certainly the most studied, is Chicago’s (Skogan,
Steiner, Benitez et al. 2004). Chicago’s police-sponsored “beat meetings”
offer neighborhood residents an opportunity to influence police priori-
ties, participate in problem-solving, and evaluate and discuss the quality
of police service in the neighborhood. Researchers report that the extent
to which these things are achieved varies greatly, but that on the whole,
54 Stephen Mastrofski

public involvement has improved over the last decade. Large numbers
of residents show up, citizens actively discuss neighborhood problems,
police provide information, solutions are proposed, and over time this
has become as likely in disadvantaged neighborhoods as in those that
are better off. By 2002 over half of the observed beat meetings involved
equal police–citizen participation (Skogan et al. 2004: 19–23). However,
despite intensive efforts, the meetings have had much less success in
mobilizing citizens to engage in collective self-help behavior. It seems
that large numbers of those who go to the trouble to participate tend to
view the meetings as places to lobby for service delivery, not participate
in its coproduction.
Seattle’s neighborhood partnership efforts emerged from community
groups organizing on their own to influence police and policing, but an
in-depth study there yielded a less positive assessment than in Chicago
(Lyons 1999). Despite some early advances in advancing democratic
mechanisms for interaction, problem-solving, and accountability (both
among different citizen groups and between citizens and local govern-
ment), these were observed to atrophy, and instead Seattle’s community
groups served as a device to secure community conformance to police
structures and priorities.
Providing services. Community policing advocates expect that offi-
cers will be more inclined to serve the public in ways that appeal to
those who desire direct benefits. Community policing specialists do not
appear to be more inclined than generalists to engage in such activities
as: controlling problem citizens, being nice to citizens, spending time
on citizens’ problems (Mastrofski et al. 2000; Snipes 2002), or using
an informal, order maintenance solution (Novak, Hartman, Holsinger,
and Turner 1999). Some studies do show that training and officer atti-
tudes are associated with an increased proclivity to deliver services,
but even here the results are mixed (Committee to Review Research
2004: 142–145). Some (including a randomized experiment) show the
expected relationship, but others (also including a randomized experi-
ment) do not. One study found that officers having both training and
a pro-community policing attitude were more inclined to grant citi-
zens’ requests to control people who were bothering them (Mastrofski
et al. 2000), but another found that officers more positively oriented to
community policing spent less time on their encounters with citizens
(Snipes 2002). At best, the capacity of community policing to enhance
service-style practices is contingent on currently unspecified conditions,
making the prediction of improvements a river boat gamble.
Process-oriented policing. Many community policing reformers want
officers to attend to the processes of police–citizen interaction so that
citizens feel good about those interactions – regardless of what the police
 Community policing 55

do to or for the citizen. That is, the public should accept these actions as
legitimate. Legitimacy requires officers to show respect to citizens, listen
to their complaints and viewpoints, show concern for their well-being
through inquiry and demonstrations of understanding, and demonstrate
procedural fairness (Mastrofski 1999; Tyler 2004). While there is a grow-
ing body of evidence indicating that when police behave in these ways,
police legitimacy is increased and concomitantly citizen compliance and
law abidingness also increase (Sherman 1997b; Committee to Review
Research 2004; Tyler 2004), there is very little evidence about the extent
to which community policing efforts are responsible for officers engaging in
these actions. Of the two studies testing this, one found that officers who
regarded community policing positively were more likely to get citizens
to comply with their requests to stop or avoid misbehavior, but this anal-
ysis also showed that pro-community policing officers were generally not
more likely to resort to legitimacy-enhancing procedural methods (Mas-
trofski et al. 1996: 290; McCluskey, Mastrofski, and Parks 1999). The
researchers hypothesized that officers who embraced community policing
were perhaps more skilled at fitting the most effective compliance strategy
to the particular circumstances. Another study of adherence to constitu-
tional standards of search and seizure found that those officers who were
most strongly committed to community policing were also those who
committed the largest share of constitutional violations, perhaps because
they were more committed to ridding the community of troublemakers
than adhering to procedural requirements (Gould and Mastrofski 2004).
Thus, we cannot say with any confidence that community policing has
really contributed to a more procedurally sensitive style of policing in
America.

The outcomes of policing

Reformers claim that community policing will lessen crime, disorder,
and fear of crime. They also expect that it will render more and better
service to the public and increase the social capital in our communities.
Some expect that it will produce a more equitable distribution of ser-
vices – giving more to those who have less. And, ultimately, community
policing perhaps above all, is expected to produce more support for the
public police, a greater degree of legitimacy. What does the evidence
show?

Crime, disorder, and fear of crime

The National Academies have recently published a volume that includes a
review of research on the impact of community policing. I draw liberally
upon its findings here, and those of a somewhat modified version of
56 Stephen Mastrofski

its report published by the two researchers who chaired this part of
the National Academy of Science (NAS) review (Committee to Review
Research 2004: ch. 6; Weisburd and Eck 2004). Both of these reports con-
cluded that the available research shows that foot patrol, storefront offices,
newsletters, and community meetings do not reduce crime, although
some may influence perceptions of disorder, a pattern consistent with ear-
lier reviews of the literature (Rosenbaum 1988; Sherman 1997b). Com-
munity policing programs designed to increase police–community inter-
action or make police more visible and accessible were found to reduce
fear of crime and perceptions of disorder, but only one strategy, police
officers making routine, door-to-door contact with residents, was shown
to reduce crime. Finally, they noted a growing body of research that finds
with consistency that when police undertake acts perceived as enhanc-
ing procedural fairness, citizens are more likely to comply with police
requests and more likely to obey the law in the future (see McCluskey
2003; Tyler 2004).
The experts’ review of community policing’s effectiveness in reducing
crime, fear, and disorder seem equivocal at best and pessimistic at worst.
The National Academies committee concluded:

Some community policing strategies appear to reduce crime, disorder, or fear of

crime. Many others have not been found to be effective when evaluated. (Com-
mittee to Review Research 2004: 246)
Nonetheless, the research available suggests that when the police partner more
generally with the public, levels of citizen fear will decline. Moreover, when
the police are able to gain wider legitimacy among citizens and offenders,
nonexperimental evidence suggests that the likelihood of offending will be
reduced. (Committee to Review Research 2004: 250–251)

Weisburd and Eck (2004: 59) were somewhat more cautious:

Yet in reviewing existing studies, we could find no consistent research agenda that
would allow us to assess with strong confidence the effectiveness of community
policing. Given the importance of community policing, we were surprised that
more systematic study was not available.

The skeptic’s perspective on this body of research subscribes to the fol-

lowing points. First, the claim that community policing at least reduces
fear of crime and perceptions of disorder deserves closer scrutiny. Even
the experimental evaluations failed to control for the possibility that
reductions in people’s cognitions of crime and victimization risk were
due to a sort of Hawthorne effect. That is, it may be that what produced
these results was the perception that citizens were receiving more or bet-
ter attention; the control groups did not include programs that ensured
 Community policing 57

comparable levels of service of a different sort, nor did they attempt to

create a comparable sense of “specialness” enjoyed by the community
policing treatment groups.
Second, we should not place too much weight on the small number of
studies that have shown that door-to-door home visits reduce crime. Even
though these were based on experimental evaluations, the theory that
links an infrequent home visit to neighborhood crime reduction requires
considerable elaboration. How is it that so mild an intervention could
yield a significant reduction in crime? Did this cause neighbors to mobilize
more effectively to prevent crime, and if so, how does that square with
so many other studies that show that such citizen mobilizations show no
crime reduction effects? Did these visits render better information for
police to identify and arrest criminals? To take preventive action?
Third, nearly all community partnership programs suffer from a sim-
ilar set of theoretical and practical challenges that make success prob-
lematic. Their failure may be due to weak implementation, which I have
already indicated is likely widespread (see also Rosenbaum 1988). But it
seems just as likely that even where the programs have been thoroughly
and vigorously implemented, success will be haphazard at best. Stronger
police–community partnerships, at least as implemented, do not appear
to add much to a community’s capacity to control crime and disorder.
That is because police organizations (a) have not done much to develop
better relations with those who have the best information about crime
and disorder, and (b) police have not developed systems that can sift
through and process information that will really help them develop effec-
tive strategies to reduce crime and disorder. There can be little doubt
that effective policing depends upon a cooperative citizenry, but the value
added by community policing programs, which focus on citizens predis-
posed to work with the police, has been marginal. Even special programs
designed to make youths feel good about the police (DARE, police ath-
letic leagues), in the hope that they will later provide useful information
about crime and other community problems, are very weak interventions
when compared with the day-to-day experiences those youths and their
acquaintances have with the police. The public is more cooperative when
they perceive the police to be considerate and fair (Tyler and Huo 2002),
but this has not been a well-developed aspect of community policing pro-
grams (see next paragraph). Those most likely to have the best informa-
tion about crime are the very ones who also report the worst experiences
with the police. And it is not at all clear that police departments have
figured out how to use effectively the information that does flow to them
from the public. Crime analysis units still comprise a very small part of
police departments, and recent reports about their limitations are not
58 Stephen Mastrofski

very reassuring (O’Shea and Nicholls 2003), nor are reports that depart-
mental communications about community-based problem solving are too
poorly written to facilitate departmental monitoring (Skogan et al. 2004:
154).
Fourth, despite the undeniable appeal of a kinder, fairer police also
being more effective in reducing crime, the relevance of these studies to
community policing must be questioned. Community policing promises
to deliver this style of policing, and the question for us is whether com-
munity policing interventions have done so and can therefore be linked to
those desirable outcomes. As mentioned earlier, research has not linked
community policing to this tendency to engage in procedural fairness.
Further, we know nothing of the effects of community policing training
and other strategies to promote this style of policing. The evidence sug-
gests that when officers engage in procedural fairness, citizens are more
compliant and law-abiding, but the real trick appears to be in finding
ways to get officers to adopt those approaches, and here the very limited
evidence available on community policing is not too promising.

Building social capital

Some research offers a beacon of hope to community policing advocates.
Researchers are finding that collective efficacy (social cohesion among
neighbors and a willingness to act toward the common good) is associ-
ated with lower levels of violence and disorder, and that this may be causal
in nature (Sampson and Groves 1989; Sampson, Raudenbush, and Earls
1997; Sampson and Raudenbush 1999; Reisig and Parks 2004). Neigh-
borhoods are socially efficacious when members monitor the behavior
of children, intervene to prevent socially disruptive practices (e.g., tru-
ancy, prostitution, drug dealing), and to secure resources from public
and private institutions that can benefit the neighborhood. Through these
mechanisms the neighborhood’s capacity to prevent and react to threats
to safety and order are enhanced. Although there is much still to be
learned about the effects of social efficacy on crime and disorder, these
findings are encouraging. The key question is whether community polic-
ing promotes social efficacy. Unfortunately, the answer seems to be that
most community policing programs are not well suited to promote this
purpose.
The most obvious problem with the popular community partnership
programs is that they simply do not focus on developing the skills and
habits among the public (or the police) that build collective efficacy.
Neighborhood Watch does little to build social cohesion where it does
not already exist. DARE tries to produce individuals with a drug-free
 Community policing 59

lifestyle, but does little to develop the civic virtues of persons engaged
in collective enterprises for the common good. Citizen–police academies
teach a great deal about how police organizations operate, but that seems
useful mostly for getting citizens to fit in with the department’s organi-
zational structures and routines, not for taking the initiative in one’s own
neighborhood. The problem with most community policing partnership
programs is that they focus too much on how to conform to police expec-
tations rather than how members of the community can be given the will,
skill, and resources to themselves engage in behaviors that contribute
to less crime and a higher quality of life in the community. Whether
the police are even the best instrument or should take the lead to pro-
mote collective efficacy is debatable. While the evidence on community-
based efforts is not especially heartening (Sherman 1997a), there is some
encouraging evidence about certain school-based programs (Gottfredson
1997). Wherever such efforts are sponsored, they are hard work, espe-
cially in those neighborhoods that are most disadvantaged, because the
limitations of distrust, lack of resources, and inexperience in organizing
present great challenges.
To be sure, some departments attempt to address the challenge of
building collective efficacy by coming to grips with how to foster civic
habits in citizens and sharing decisionmaking power in police. The best
known of these is Chicago’s. But as already indicated, despite impres-
sive levels of participation in beat meetings (following a “re-engineering”
of the city’s original model), Chicago has not enjoyed much success in
mobilizing residents to act, and the police department has a high turnover
rate in officers attending the meetings, which undermines the establish-
ment of stronger police-community relations (Skogan et al. 2004: 154).
The picture of Chicago’s “district advisory committees,” operating at a
more strategic level, is far less positive, lacking a clear purpose and getting
more advice from police than they give to them. Thus, while Chicago’s
community policing may have made significant advances in developing
the bases for social capital, it has a long way to go.

More equitable distribution of policing’s benefits

Liberals are attracted to the prospect that community policing will deliver
more good things to society’s have-nots. Even if community policing were
not more effective, if it redistributed services more equitably, one could
argue that it had produced a net good for society. However, community
policing calls for an alteration in the mechanisms of service delivery,
shifting from one that relies on reacting to individual calls for service,
to one that reacts to priorities established by organized elements of the
60 Stephen Mastrofski

community – neighborhood organizations, businesses, churches, victim

groups, and the like. As previously noted, some research shows that when
society’s disadvantaged are well organized for collective action in these
ways, greater benefits do come their way, but the problem remains: the
disadvantaged are less likely to organize effectively for collective action
(Skogan 1994: 179). The question then is whether community policing
invests in society’s have-nots a greater capacity to engage the police and
lobby for the benefits they can bestow. If it does not, then by virtue of its
distributional biases, giving more to those who can organize effectively, it
may actually exacerbate service quantity and quality differentials between
the haves and have-nots.
Skogan (2004; and Skogan et al. 2004: 26–37) provides the most
focused research on this question, confirming that those most active in
neighborhood beat meetings and district advisory committees in Chicago
are indeed those who are better off, especially in Latino areas. In Chicago
there is only a weak correspondence between what concerns the residen-
tial population of neighborhoods and those active in beat meetings, and
activists tend to be more positive about the quality of police services (Sko-
gan 2004: 65–67). However, there is some evidence that where neigh-
borhood organizations work most energetically with the police, they will
receive more service and that the city’s response does tend to correspond
to the priorities of both residents in general and especially activists. Yet
this research does not provide a comparison of whether Chicago’s reor-
ganization of police service delivery to respond to neighborhood orga-
nization priorities works as well for the disadvantaged as it does for the
advantaged, much less whether the disadvantaged are “gaining ground”
compared to the more advantaged neighborhoods. Ultimately, we do not
know whether Chicago’s disadvantaged are better off under the com-
munity policing arrangements for mobilizing police responses than what
preceded them.4
The research seems to tell us that where police-community partner-
ships flourish, there is a greater possibility of improving the quality of
life, but that community policing’s benefits will least likely be delivered
to society’s most afflicted. Researchers have not yet demonstrated that
community policing helps them gain ground, but it does suggest that
in a community policing environment, collective action is an important
element for doing so. The challenge is in finding how to infuse the dis-
advantaged with the will and resources to organize effectively, and how
to develop a community will to make the necessary investments to make
that happen. If there is a magic wand to do this, community policing has
not yet found how to activate its software.
 Community policing 61

The success story: nothing exceeds like success

When Americans want ever so badly to have a success, there are always
many willing to seize upon an emblematic example, a “poster child” that
inspires us to see what can be done, and from that infer that it is being
done. Public officials, ever sensitive to the state of police legitimacy, surely
welcome these success stories and understandably embrace them as good
omens. Following in the wake of the crises of legitimacy that helped to ring
down the curtain on the “professional” era of police reform, community
policing success stories have proven to be a powerful force for sustain-
ing the reform. They fuel the engines of “institutional isomorphism” in
policing – as evidenced by the rapidity with which contemporary reforms
are adopted. Obversely, skepticism is not all that palatable when people
seek to nourish their hopes from such success stories. There are several
such community policing success stories around the nation, and perhaps
the most chronicled of these is Chicago’s. I will focus on it, because it
has the added advantage of being the object of over a decade’s intense
empirical evaluation.
The large body of studies published on Chicago’s community policing
effort comes from a consortium of researchers led by Wesley Skogan. As
he has written the companion chapter to mine in this volume, I will not
offer a detailed review of this careful, impressive body of work, but from
it I will draw my own conclusions. Does community policing in Chicago
work? Not like a Swiss clock, but it is a worthy effort that seems to have
rendered many significant improvements over what preceded it (Skogan
et al. 2004). That seems to me cause for encouragement, but my skeptic
responsibilities lead me to take a second lesson from the Chicago story,
which is that effective implementation of community policing is not a walk
in the park; it is hard work that requires identifying and correcting mis-
takes and a sustained commitment at many levels inside and outside the
organization (DuBois and Hartnett 2002; Skogan et al. 2004). Repeat-
edly, Chicago’s evaluators note that there are many obstacles to full and
effective implementation, they are not always overcome, and often the
successes are mixed. Is this the model for the nation? I think not. Like
running a sub-four-minute mile, we know it can be done, but we also
know that few have the will and resources to do it. Early adopters may
have that degree of commitment, but later adopters often rely upon the
notoriety of the early adopters’ success, and fail to attend to the chal-
lenges that must be overcome to secure that success in their case (Tolbert
and Zucker 1983; Ritti and Mastrofski 2002). In fact, it seems likely that
most police departments and their communities are unwilling or unable
62 Stephen Mastrofski

to make the sorts of commitments it takes to go beyond the fairly superfi-

cial transformations that come from adopting canned programs that are
pale replicas of the “real deal.” Even as early as 1993, only half of the
top executive respondents to a national survey indicated that community
policing would require major changes in policies, goals, mission state-
ments, and training, and only 27 percent indicated that it would require
extensive reorganization (Wycoff 1994). Thus, it seems unlikely that large
numbers were committed to the kind of effort that Chicago’s experience
suggests is necessary to show substantial gains.

Enhanced police legitimacy

At first blush, the answer to the question, “Did community policing
increase police legitimacy?” appears to be a no-brainer. Its rampant popu-
larity among police agencies and elected officials, and the virtual absence
of negative publicity point to a hearty, “Yes!” But if we apply the same
standard of evidence to this issue as to others, the answer becomes less
obvious.
We do not have anything resembling an experimental or quasi-
experimental design by which we might judge the impact of the reform on
legitimacy, so we need to rely on less rigorous methods. One approach is to
measure legitimacy using public opinion surveys of the police. If commu-
nity policing were enhancing police legitimacy, we should note that as its
implementation increased in popularity over time, the public should ren-
der more positive views of the police, and we should expect to see a pro-
nounced acceleration in positive evaluations in the years following 1994,
when the availability of COPS funds added incentives for departments
to implement or at least claim to implement community policing pro-
grams. National public opinion surveys offer no evidence of a nation-wide
trend of this sort (Gallagher, Maguire, Mastrofski, and Reisig 2001). A
variety of indicators (confidence, respect, satisfaction, fairness, honesty,
brutality) suggest that during the era of community policing, the pub-
lic’s evaluations of police have remained stable, fluctuated considerably
according to no distinct pattern, or have declined. Cross-sectional com-
parisons of public opinion about departments that vary in their degree of
community policing implementation or symbolic commitment to com-
munity policing could provide another indicator. Unfortunately, no such
study has been conducted. A 1998 national survey of citizens’ satisfac-
tion with the police serving their neighborhood compared twelve cities,
finding that cities ranged from 78–97 percent reporting that they were
satisfied. However, eight of the twelve varied little, only 84–89 percent.
 Community policing 63

Among those cities at the high and low levels of satisfaction there were
no obvious patterns related to community policing.
An alternative view is that American police do not need to heighten
the public’s support beyond normal levels but are keen to ensure that
it is sustained at an acceptable level, one that provides a stable environ-
ment in which organizations can operate without the need for radical
change. In this regard, the above survey evidence suggests that commu-
nity policing may have served admirably, because support for the police
as an American institution has remained remarkably stable over the last
two decades. This argument is subsumed under the claim that com-
munity policing is but the most recent manifestation of a long series of
reforms designed to sustain the legitimacy of police (militarization, legal-
ization, and professionalization), all circumlocutions that conceal or palli-
ate the coercive nature of police work about which society is so ambivalent
(Bittner 1970; Klockars 1988).
The two legs on which the legitimacy of community policing stands
are responsiveness to the public and delivering results, and in both cases
American society has not demanded rigorous scientific evidence to find
value in this reform. While strong and consistent scientific evidence of its
performance has not been available, the reform has prospered, at least if
one judges by the eagerness with which communities around the nation
have signed up for COPS grants (Maguire and Mastrofski 2000; Roth
et al. 2000; Worrall and Zhao 2003), police leaders have employed its
rhetoric and adopted its programs. Indeed, the very adoption of commu-
nity policing programs seems to have provided a sufficiently strong signal
to community leaders and the public at large that their police are doing
the “right” thing (Committee to Review Research 2004: 308–12), and
in this sense, community policing must be considered successful. This
state of affairs is predicted by institutional organization theory (Scott
1992), and is likely to remain that way unless a different environment for
legitimacy evolves, such as the “evidence-based” policing environment
advocated by Sherman (1998).

The future of community policing: terrorist-oriented

policing
Despite the contribution that community policing has made to the legit-
imacy of the American police, it is on hard times. The COPS Office,
financial engine of the American reform movement, has had its bud-
get slashed, and it is struggling to remain viable by demonstrating the
utility of community policing to what one police chief recently called
64 Stephen Mastrofski

“terrorist-oriented policing” (Kerlikowske 2004). Meanwhile, the

Department of Homeland Security is distributing billions of dollars to
state and local police agencies to join the federal government in the war
against terror. Here community policing demonstrates its worth to the
extent that it encourages citizens to be forthcoming with information
useful to counterterrorist measures. This conjures up the vision of an
eager public of amateur informants, voluntarily feeding information to
the police, who carefully sift through it to guide their efforts to under-
mine terrorist threats.
The skeptic understandably wonders how realistic this terrorist-
oriented, community policing vision is. Neighborhood Watch and similar
crime prevention efforts have failed to demonstrate substantial benefits
in reducing “regular” crime, so these canned, off-the shelf approaches
seem even less likely to be effective in developing reliable sources on
terrorism in those neighborhoods where the most valuable information
may reside – immigrant communities, and especially, Islamic-American
communities (Lyons 2002). Here the challenges for eliciting voluntary
terrorist-oriented information are great. To the extent that local police
respond to the federal government’s desire to be an intelligence conduit,
how are local police to distinguish themselves from the federal enforce-
ment practices that make immigrant and Islamic citizens and residents
so fearful (dragnets, interrogations, secret trials and incarcerations, and
deportation of undocumented aliens)? And how are local police to man-
age the tension between placating the fears of a public that supports profil-
ing Islamic persons for terrorism, while demonstrating to those targeted
groups an equitable sensitivity to the civil rights of all citizens (Davies
2002)?
Community policing advocates are rushing to assure Congress that
community policing is essential to the success of this effort. Some have
argued that this form of intelligence brokering has long been the hallmark
of community policing (Ericson and Haggerty 1997), but the enlistment
of community policing in the domestic war against terror marks a distinct
change in the bases for justifying the reform. This transformation further
subordinates the community to the police in this partnership, relegat-
ing police responsiveness and accountability to the public to the hinter-
lands of the police mission. The partnerships that matter most to police
under terrorist-oriented policing are those that promote the more effi-
cient exchange of information with federal law enforcement, intelligence,
and various disaster response agencies at all levels of government (Com-
mittee to Review Research 2004: 209–214). And many are concerned
that the terrorist-oriented mission and close association with military
and intelligence agencies will accelerate the militarization and isolation
 Community policing 65

of state and local police forces, rather than bringing them closer to the
police-community partnerships that served as the source of legitimacy
for community policing in the past. Can this be the end of community
policing? To the extent that the war against terror shapes community
policing, it will end it as we have known it.
Most troubling in the way that community policing is being bent to fed-
eral priorities in the war against terror is the limited role envisioned for
the community – primarily one of serving the state as its eyes and ears.
Virtually absent from the discussion is the great need for community-
based self-help responses to disaster, especially when state responses fall
short, as they most assuredly will (Clarke 2003). The current obsession
with establishing government command and control overplays the state’s
capacity to do all that must be done, and it overlooks the potential of cit-
izen grassroots organizations both to reduce vulnerability and to respond
to crisis. I would prefer to see more emphasis on a form of commu-
nity policing that sought a role for the government in strengthening the
capacity of citizens to engage in collectively efficacious action in crisis
conditions.

Conclusion
Community policing’s advocates have promised a great deal, and they
may have thereby contributed to a rising sense of entitlement to better
service. But after a few decades of reform, those looking for evidence of
a “quiet revolution,” a “new blue line,” or a “paradigm shift,” should
be disappointed by the state of affairs. The quantity and quality of the
evidence is itself disappointing, but considering that which is available,
the skeptic must conclude that the glass of community policing’s benefits
is closer to empty than full. Its implementation has not transformed the
structure and operations of American policing so much as it has altered its
rhetoric. While there may be a few community policing exemplars among
the nation’s police departments, there is little in the national policing
landscape to assure us that structures and practices have been strikingly
altered. The evidence of its effectiveness resides primarily in its capac-
ity to make the public less fearful of crime, while doing little to reduce
crime itself. While collective action promises to provide real benefits to
the nation’s neighborhoods and communities, community policing advo-
cates are still struggling to find the elixir that will transform quiescent
and alienated citizens into virtuous civic actors. While perhaps shifting
the bases of police and government resource mobilization toward those
who engage in collective action, community policing has not necessarily
been effective in spreading these benefits to those who need them most,
66 Stephen Mastrofski

society’s disadvantaged. Its capacity to serve effectively as a bulwark in

the defense of the homeland against terrorism remains at present only
wishful thinking.
To say that community policing’s performance has been disappointing
does not lead to the conclusion that it was the wrong path, but it is one
that has for many meandered into a stagnant swamp of tepid programs
and amorphous “philosophies,” too often run by and for the police more
than the community. Its greatest weakness, to adapt Gertrude Stein’s
famous line, is that there is often so very little there. Most police depart-
ments treat it as an add-on rather than something that requires a radical
transformation of how the police and public organize to do policing. Such
a transformation would require that the police first engage a community
that can serve as an effective partner, and once that is accomplished, the
police will inevitably require sharing far more power with the community
than they have thus far been willing to do – the political equivalent of
the loving, but oft-contentious partnerships portrayed in film by Spencer
Tracy and Katherine Hepburn. This is risk-taking of a major sort, because
our police have long struggled, often with good reason, to build effective
buffers from community pressure (Reiss 1992), and it is not at all clear
that enough citizens are willing to commit the necessary degree of effort,
especially in the most afflicted areas. These partnerships, even those most
earnestly pursued, have been constructed, often with the acquiescence of
citizens, to prevent truly potent political organizing to evolve (Skogan
2004: 59). A cynic might even suggest that this is their intended purpose.
Whether that is so, or not, they represent a bland version of the ideals
of a democratically invigorated policing in America. Without a piquant
dish of community policing, we lack a true test of its potential to flavor
our communities with beneficial outcomes. American police and their
communities would be better served spending less time celebrating com-
munity policing and more time figuring out how to do it in a meaningful
way.

   
The author thanks David Klinger, Edward Maguire, Roger Parks, and David
Weisburd for their helpful suggestions in the preparation of this manuscript.
1. Respondents to a 1999 national survey of municipal police agencies of ten
or more full-time officers showed that 29 percent provided their recruits with
sixteen or more hours of training on community policing concepts and prin-
ciples, while 54 percent provided that much training on traffic accident inves-
tigation, and 67 percent provided that much training in physical tactics and
martial arts. Only 59 percent provided any training on organizing community
groups, and the vast majority of these provided less than eight hours of such
 Community policing 67

training. These figures have not been published heretofore. They are taken
from “Community Policing in America: National Survey of Police Execu-
tives and Agencies,” which is described in Mastrofski, Parks, and Wilson
(2003).
2. The Project on Policing Neighborhoods found that general patrol officers
spent only about a quarter of their time in face-to-face public encounters and
community policing specialists spent less than 20 percent (Parks, Mastrofski,
DeJong, and Gray 1999: 499–500; see also Mastrofski, Parks, Reiss et al.,
1998). On average, patrol generalists spent less than a handful of minutes at
community meetings, while specialists averaged only somewhat more (see also
Frank, Brandl, and Watkins 1997: 721, 724). Despite the best management
intentions to keep officers in their permanent beat assignments, officers seem
to have difficulty remaining in their assigned beats for extended periods (Parks,
Mastrofski, Reiss et al. 1998: 2–18). Finally, community policing special-
ists appear inclined to spend more of their time with citizens who are more
respectable and less likely to present them with elevated emotions (Parks et al.
1999).
3. Officers’ attitudes toward developing partnerships with the community had
no significant bearing on the inclination to use verbal or physical force with
suspects in the two departments studied. The effects of training and commu-
nity policing specialist assignment did not show consistent effects in the two
departments.
4. However, research on Indianapolis and St. Petersburg does suggest that police–
citizen partnerships did have the desired effects on levels of perceived safety,
regardless of the degree of structured disadvantage in the neighborhood, but
modest differences did remain across socioeconomic categories at the individ-
ual level (Reisig and Parks 2004).



Ajzen, I. (1987). Attitudes, traits, and actions: Dispositional prediction of behav-

ior in personality and social psychology. Advances in Experimental Social
Psychology, 20, 1–63.
Bittner, E. (1970). The functions of police in modern society. Washington, DC:
National Institute of Mental Health.
Bratton, W. (1998). Turnaround: How America’s top cop reversed the crime epidemic.
New York: Random House.
Buerger, M. (1998). Police training as a pentecost: Using tools singularly ill-suited
to the purpose of reform. Police Quarterly, 1, 27–63.
Chermak, S. and Weiss, A. (2002). Identifying strategies to market police in the
news. Final report to the National Institute of Justice. Retrieved from:
http://www.ncjrs.ors/pdffiles1/nij/grants/194130.pdf.
Clarke, L. (ed.). (2003). Terrorism and disaster: New threats, new ideas. Research
in Social Problems and Public Policy, 11. Oxford: Elsevier.
Cochran, J. K. and Bromley, M. L. (2003). The myth(?) of the police subculture.
Policing: An International Journal of Police Strategies and Management, 26, 88–
117.
68 Stephen Mastrofski

Committee to Review Research on Police Policy and Practices. (2004). Fair-

ness and effectiveness in policing: The evidence. Washington, DC: National
Academies Press.
Crank, J. P. (1994). Watchman and community: Myth and institutionalization in
policing. Law and Society Review, 28, 325–351.
(2003). Institutional theory of police: A review of the state of the art. Policing:
An International Journal of Police Strategies and Management, 26, 186–207.
Davies, S. L. (2002). Profiling terror. Paper delivered at the 16th International
Conference of the International Society for the Reform of Criminal Law,
Charleston, SC. Retrieved from: http://www.isrcl.org/Papers/Davies.pdf.
DiMaggio, P. J., and Powell, W. W. (1983). The iron cage revisited: Institutional
isomorphism and collective rationality in organizational fields. American Soci-
ological Review, 48, 147–160.
DuBois, J. and Hartnett, S. M. (2002). Making the community side of community
policing work. In Dennis J. Stevens (ed.), Policing and community partnerships
(pp. 1–16). New York: Prentice Hall.
Ericson, R. V. and Haggerty, K. D. (1997). Policing the risk society. Toronto,
Canada: University of Toronto Press.
Fogelson, R. M. (1977). Big city police. Cambridge, MA: Harvard University
Press.
Frank, J., Brandl, S. G., and Watkins, R. C. (1997). The content of community
policing: A comparison of the daily activities of community and “beat” offi-
cers. Policing: An International Journal of Police Strategy and Management, 20,
716–728.
Gallagher, C., Maguire, E. R., Mastrofski, S. D., and Reisig, M. D. (2001).
The public image of the police. Final report to the International Association
of Chiefs of Police. Retrieved from: http://www.theiacp.org/profassist/ethics/
public image.htm.
Goldstein, H. (1990). Problem-oriented policing. New York: McGraw-Hill.
Gottfredson, D. C. (1997). School-based crime prevention. In L. W. Sherman,
D. Gottfredson, D. MacKenzie, J. Eck, P. Reuter, and S. Bushway (eds.),
Preventing crime: What works, what doesn’t, what’s promising? (pp. 5–74). Wash-
ington, DC: National Institute of Justice.
Gottfredson, D. C., Wilson, D. B., and Najaka, S. S. (2002). School-based crime
prevention. In L. W. Sherman, D. P. Farrington., B. C. Welsh., and D. L.
MacKenzie (eds.), Evidence-based crime prevention. London: Routledge.
Gould, J. B. and Mastrofski, S. D. (2004). Suspect searches: Assessing police
behavior under the US Constitution. Criminology and Public Policy, 3 (3),
315–362.
Greene, J. R. (2004). Community policing and organization change. In Wesley
G. Skogan (ed.), Community policing: Can it work? (pp. 30–53). Belmont,
CA: Wadsworth.
Greene, J. R., Bergman, W. T., and McLaughlin, E. J. (1994). Implementing
community policing: Cultural and structural change in police organizations.
In D. Rosenbaum (ed.), The challenge of community policing: Testing the promises
(pp. 92–109). Thousand Oaks, CA: Sage Publications.
Haar, R. N. (2001). The making of a community policing officer: The impact of
basic training and occupational socialization on police recruits. Police Quar-
terly, 4, 402–433.
 Community policing 69

Harris Poll. (1999). Study 828460. Retrieved from: ftp://ftp.irss.unc.edu//pub/

search results/POLL..7Jun2004.09.38.53.txt
Kerik, B. B. (2001). The lost son: A life in pursuit of justice. New York: Regan Books.
Kerlikowske, R. G. (2004). The end of community policing: Remembering the
lessons learned. FBI Bulletin, April 6–10.
Klockars, C. B. (1988). The rhetoric of community policing. In J. R. Greene and
S. D. Mastrofski (eds.), Community policing: Rhetoric or reality? (pp. 239–258).
New York: Praeger.
LaFree, G. (1998). Losing legitimacy: Street crime and the decline of institutions in
America. Boulder, CO: Westview Press.
Lyons, W. (1999). The politics of community policing: Rearranging the power to punish.
Ann Arbor, MI: University of Michigan Press.
(2002). Partnerships, information and public safety: Community policing in a
time of terror. Policing: An International Journal of Police Strategies and Man-
agement, 25, 530–542.
Maguire, E. R. (1997). Structural change in large municipal police organizations
during the community policing era. Justice Quarterly, 14, 547–576.
(2002). Multiwave establishment surveys of police organizations. Special Issue,
Justice Research and Policy, 4, 39–60.
Maguire, E. R. and Katz, C. M. (2002). Community policing, loose coupling,
and sensemaking in American police agencies. Justice Quarterly, 19, 501–
534.
Maguire, E. R. and King, W. R. (2004). Trends in the policing industry. The
Annals of the American Academy of Political and Social Science, 593, 15–
41.
Maguire, E. R. and Mastrofski, S. D. (2000). Patterns of community policing in
the United States. Police Quarterly, 3, 4–45.
Maguire, E. R., Shin, Y., Zhao, J., and Hassell, K. D. (2003). Structural change
in large police agencies during the 1990s. Policing: An International Journal
of Police Strategies and Management, 26, 251–275.
Manning, P. K. (1984). Community policing. American Journal of Police, 3, 205–
227.
Mastrofski, S. D. (1998). Community policing and police organization structure.
In J. Brodeur (ed.), Community policing and the evaluation of police service
delivery (pp. 161–189). Thousand Oaks, CA: Sage Publications.
(1999). Policing for people. Ideas in American Policing Series. Washington, DC:
Police Foundation.
(2002). The romance of police leadership. In E. Waring, D. Weisburd, and
L. W. Sherman (eds.), Theoretical advances in criminology: Crime and social
organization (pp. 153–196). New Brunswick: Transaction Publishers.
(2004). Controlling street-level police discretion. The Annals of the American
Academy of Political and Social Sciences, 593, 100–118.
Mastrofski, S. and Parks, R. (1990). Improving observational studies of police.
Criminology, 28, 475–496.
Mastrofski, S. D. and Ritti, R. R. (1999). Patterns of community policing: A view
from newspapers in the United States. COPS Working Paper #2. East Lansing,
Michigan State University.
(2000). Making sense of community policing: A theoretical perspective. Police
Practice and Research Journal, 1, 183–210.
70 Stephen Mastrofski

Mastrofski, S. D., Parks, R. B., and Wilson, D. B. (2003). Influences on the adoption
of community policing in the United States. Report to the National Institute of
Justice. Manassas, VA: George Mason University.
Mastrofski, S. D., Snipes, J. B., and Supina, A. (1996). Compliance on demand:
The public’s response to specific police requests. Journal of Research in Crime
and Delinquency, 33, 269–305.
Mastrofski, S. D., Worden, R. E., and Snipes, J. B. (1995). Law enforcement in
a time of community policing. Criminology, 33, 539–563.
Mastrofski, S. D., Snipes, J. B., Parks, R. B., and Maxwell, C. D. (2000). The
helping hand of the law: Police control of citizens on request. Criminology,
38, 307–342.
Mastrofski, S. D., Parks, R. B., Reiss, A. J., Jr., Worden, R. E., DeJong, C., Snipes,
J. B., and Terrill, W. (1998). Systematic observation of public police: Applying
field research methods to policy issues. Washington, DC: National Institute of
Justice.
McCluskey, J. D. (2003). Police requests for compliance: Coercive and procedurally
just tactics. New York: LFB Scholarly Publishing, Inc.
McCluskey, J. D., Mastrofski, S. D., and Parks, R. B. (1999). To acquiesce or
rebel: Predicting citizen compliance with police requests. Police Quarterly, 2,
389–416.
Meyer, J. W. and Rowan, B. (1977). Institutionalized organizations: Formal struc-
ture as myth and ceremony. American Journal of Sociology, 83, 340–363.
Moon, Y. (2004). Police organizations’ response to institutionalized pressure
for community policing. Ph.D. Dissertation. Bloomington, IN: Indiana
University.
Moore, M. H. (1992). Problem-solving and community policing. In M. Tonry
and N. Morris (eds.), Modern policing (pp. 99–158). Chicago: University of
Chicago Press.
Moore, M. H. and Stephens, D. (1991). Police organization and management:
Towards a new managerial orthodoxy. Washington, DC: Police Executive
Research Forum.
Muir, W. K., Jr. (1977). Police: Streetcorner politicians. Chicago: University of
Chicago Press.
Novak, K. J., Hartman, J. L., Holsinger, A. M., and Turner, M. G. (1999).
The effects of aggressive policing of disorder on serious crime. Policing: An
International Journal of Police Strategies and Management, 22, 171–190.
O’Shea, T. C. and Nicholls, K. (2003). Crime analysis in America: Findings and
recommendations. Washington, DC: Office of Community Oriented Policing
Services.
Parks, R. B., Mastrofski, S. D., DeJong, C., and Gray, M. K. (1999). How officers
spend their time with the community. Justice Quarterly, 16, 483–518.
Parks, R. B., Mastrofski, S. D., Reiss, A. J., Jr., Worden, R. E., Terrill, W. C.,
DeJong, C., Stroshine, M., and Shepard, R. (1998). St. Petersburg project on
policing neighborhoods: A study of the police and the community. Report to the
National Institute of Justice. Bloomington, IN: Indiana University.
Reisig, M. D. and Chandek, M. S. (2001). The effects of expectancy discon-
firmation on outcome satisfaction in police–citizen encounters. Policing: An
International Journal of Police Strategies and Management, 24, 88–99.
 Community policing 71

Reisig, M. D. and Parks, R. B. (2004). Community policing and quality of life.

In W. G. Skogan (ed.), Community policing: Can it work? (pp. 207–227).
Belmont, CA: Wadsworth.
Reiss, A. J., Jr. (1971). The police and the public. New Haven: Yale University Press.
(1992). Police organization in the twentieth century. In M. Tonry and N. Morris
(eds.), Modern policing (pp. 51–98). Chicago: University of Chicago Press.
Ritti, R. R. and Mastrofski, S. D. (2002). The institutionalization of community
policing: A study of the presentation of the concept in two law enforcement jour-
nals. Final report to the National Institute of Justice. Manassas, VA: George
Mason University.
Rosenbaum, D. P. (1988). Community crime prevention: A review and synthesis
of the literature. Justice Quarterly, 5, 323–395.
Rosenbaum, D. P. and Hanson, G. S. (1998). Assessing the effects of school-
based drug education: A six-year multi-level analysis of project D.A.R.E.
Journal of Research in Crime and Delinquency, 35 (4), 381–412.
Rosenbaum, D. P. and Wilkinson, D. L. (2004). Can police adapt? Tracking the
effects of organizational reform over six years. In W. Skogan (ed.), Community
policing: Can it work? (pp. 79–108). Belmont, CA: Wadsworth.
Roth, J. A., Roehl, J., and Johnson, C. C. (2004). Trends in community policing.
In W. Skogan (ed.), Community policing: Can it work? (pp. 3–29). Belmont,
CA: Wadsworth.
Roth, J. A., Ryan, J. F., Gaffigan, S. J., Koper, C. S., Moore, M. H., Roehl, J.,
Johnson, C. C., Moore, G. E., White, R. M., Buerger, M. E., Langston,
E. A., and Thacher, D. (2000). National evaluation of the COPS program –
Title I of the 1994 Crime Act. Washington, DC: National Institute of Justice.
Sadd, S. and Grinc, R. (1994). Innovative neighborhood oriented policing: An
evaluation of community policing programs in eight cities. In D. P. Rosen-
baum (ed.), The challenge of community policing: Testing the promises (pp. 27–
52). Thousand Oaks, CA: Sage Publications.
Sampson, R. J. and Groves, W. B. (1989). Community structure and crime:
Testing social-disorganization theory. American Journal of Sociology, 94, 774–
802.
Sampson, R. J. and Raudenbush, S. W. (1999). Systematic social observation
of public spaces: A new look at disorder in urban neighborhoods. American
Journal of Sociology, 105, 603–651.
Sampson, R. J., Raudenbush, S. W., and Earls, F. (1997). Neighborhoods and
violent crime: A multilevel study of collective efficacy. Science, 277, 918–924.
Sandburg, C. (1970). Washington monument by night. Stanza 4, The complete
poems of Carl Sandburg, rev. and expanded ed. (p. 282). New York: Harcourt
Press.
Schafer, J. A. (2001). Community policing: The challenges of successful organizational
change. New York: LFB Scholarly Publishing LLC.
Scott, W. R. (1992). Organizations: Rational, natural, and open systems (3rd ed).
Englewood Cliffs, NJ: Prentice Hall.
Sherman, L. W. (1997a). Communities and crime prevention. In L. W. Sherman,
D. Gottfredson, D. MacKenzie, J. Eck, P. Reuter, and S. Bushway (eds.),
Preventing crime: What works, what doesn’t, what’s promising? (pp. 3–1–3–49).
Washington, DC: National Institute of Justice.
72 Stephen Mastrofski

(1997b). Policing for crime prevention. In L. W. Sherman, D. Gottfredson,

D. MacKenzie, J. Eck, P. Reuter, and S. Bushway (eds.), Preventing crime:
What works, what doesn’t, what’s promising (pp. 8–1–8–58). Washington, DC:
National Institute of Justice.
(1998). Evidence-based policing. Ideas in American policing series. Washington,
DC: Police Foundation.
Skogan, W. G. (1994). The impact of community policing on neighborhood
residents. In D. Rosenbaum (ed.), The challenge of community policing: Testing
the promises (pp. 167–81). Thousand Oaks, CA: Sage Publications.
(2004). Representing the community in community policing. In W. G. Sko-
gan (ed.), Community policing: Can it work? (pp. 57–75). Belmont, CA:
Wadsworth.
Skogan, W. G. and Hartnett, S. M. (1997). Community policing, Chicago style.
New York: Oxford University Press.
Skogan, W. G., Steiner, L., Benitez, C., Bennis, J., Borchers, S., DuBois, J.,
Gondocs, R., Hartnett, S., Kim, S. Y., and Rosenbaum, S. (2004). Commu-
nity policing in Chicago, year ten: An evaluation of Chicago’s alternative policing
strategy. Springfield, IL: Illinois Criminal Justice Information Authority.
Snipes, J. B. (2002). Police response to citizen requests for assistance: An assess-
ment of deservedness, workload, social status, and officer predisposition per-
spectives. Ph.D. Dissertation. University of Albany, State University of New
York.
Snipes, J. B. and Mastrofski, S. D. (1990). An empirical test of Muir’s typology
of police officers. American Journal of Criminal Justice, 14, 268–296.
Sparrow, M. K., Moore, M. H., and Kennedy, D. M. (1990). Beyond 911: A new
era for policing. New York: Basic Books.
Terrill, W. and Mastrofski, S. D. (2004). Working the street: Does community
policing matter? In W. G. Skogan (ed.), Community policing: Can it work?
(pp. 109–135). Belmont, CA: Wadsworth.
Terrill, W., Paoline, E. A., III, and Manning, P. K. (2003). Police culture and
coercion. Criminology, 41, 1003–1034.
Tien, J. M. and Rich, T. F. (1994). The Hartford COMPASS Program: Experi-
ences with a weed and seed-related program. In D. Rosenbaum (ed.), The
challenge of community policing: Testing the promises (pp. 192–206). Thousand
Oaks, CA: Sage Publications.
Tolbert, P. S. and Zucker, L. G. (1983). Institutional sources of change in the
formal structure of organizations: The diffusion of civil service reform, 1880–
1935. Administrative Science Quarterly, 28, 22–39.
Tyler, T. R. (2004). Enhancing police legitimacy. Annals of the American Academy
of Political and Social Science, 593, 84–99.
Tyler, T. R. and Huo, Y. J. (2002). Trust in the law: Encouraging public cooperation
with the police and courts. New York: Russell Sage Foundation.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder,
and fear? Annals of the American Academy of Political and Social Science, 593,
42–65.
Weisburd, D., Greenspan, R., Hamilton, E. E., Bryant, K. A., and Williams, H.
(2001). The abuse of police authority: A national study of police officers’ attitudes.
Washington, DC: Police Foundation.
 Community policing 73

Weisel, D. L. and Eck, J. E. (1994). Toward a practical approach to organizational

change: Community policing initiatives in six cities. In D. P. Rosenbaum
(ed.), The challenge of community policing: Testing the promises (pp. 53–72).
Thousand Oaks, CA: Sage Publications.
Wilson, J. Q. (1968). Varieties of police behavior. Cambridge, MA: Harvard Uni-
versity Press.
Wood, R. L., Davis, M., and Rouse, A. (2004). Diving into quicksand: Program
implementation and police subcultures. In W. G. Skogan (ed.), Community
policing: Can it work? (pp. 136–161). Belmont, CA: Wadsworth.
Worden, R. E. (1989). Situational and attitudinal explanations of police behavior:
A theoretical reappraisal and empirical assessment. Law and Society Review,
23, 667–711.
(1995). Police officers’ belief systems: A framework for analysis. American Jour-
nal of Police, 14, 49–81.
Worrall, J. L. and Zhao, J. (2003). The role of the COPS Office in community
policing. Policing: An International Journal of Police Strategies and Management,
26, 64–87.
Wycoff, M. A. (1994). Community policing strategies: Draft final report. Washington,
DC: Police Foundation.
Zhao, J., He, N., and Lovrich, N. P. (1999). Value change among police officers
at a time of organizational reform: A follow-up study using Rokeach values.
Policing: An International Journal of Police Strategies and Management, 22, 152–
170.
Part II

Broken windows policing

4 Advocate
Of “broken windows,” criminology, and
criminal justice

William H. Sousa and George L. Kelling

Introduction: a few up-front discourses

Despite attacks from the criminological, legal, and academic left, “broken
windows” theory is a robust policy option in criminal justice practice and
crime prevention. It has not only fueled the community policing move-
ment, it has also informed the evolution of community courts, commu-
nity prosecution, and community probation and parole. The Midtown
Community Court in Manhattan, to give just one example, emphasizes
that broken windows is integral to its philosophy and practice. Moreover,
the ideas embodied in broken windows have moved beyond criminal jus-
tice and criminology to areas like public health, education, parks, and
business improvement districts (BIDs).
The original article (Wilson and Kelling 1982), published in the
Atlantic, has had surprising “legs.” Although exact figures are not avail-
able, circulation staff of the Atlantic have told both James Q. Wilson and
Kelling (one of the authors of this paper), that “Broken Windows” has
been reproduced more than any other article in Atlantic’s history. More-
over, familiarity with broken windows is widespread internationally: Fix-
ing Broken Windows, published by Kelling and Catherine M. Coles in
1996, has been translated into Spanish, Polish, and Japanese. The vast
publicity, of course, associated with both the restoration of order in New
York’s subways during the early 1990s and the crime reduction in the
city itself in the mid-1990s contributed to the popularization of broken
windows, especially since both then Mayor Rudolph Giuliani and Police
Commissioner William Bratton repeatedly identified it as a key part of
their policing strategy.

Of metaphors
As background, the term “broken windows” is a metaphor. Briefly, it
argues that just as a broken window left untended is a sign that nobody

77
78 William H. Sousa and George L. Kelling

cares and invites more broken windows, so disorderly behavior left

untended is a sign that nobody cares and leads to fear of crime, more
serious crime and, ultimately, urban decay. (“Broken windows policing”
refers to a police emphasis on disorderly behavior and minor offenses,
often referred to as “quality of life” offenses like prostitution, public uri-
nation, and aggressive panhandling.)
Its expression as a metaphor partially explains the rapid spread of ideas
embodied in broken windows. A metaphor, as defined in the Oxford
English Dictionary, is “(T)he figure of speech in which a name or descrip-
tive term is transferred to some object different from, but analogous to,
that to which it is properly applicable” (1989: 676). Its origins are from
the Greek, “to transfer,” “to carry,” or “to bear.” Breaking it down, the
broken windows metaphor transfers the “common wisdom” that a minor
happening like a broken window can lead to increased damage if not
taken care of, to the presumed consequences of uncivil and petty crimi-
nal behaviors: fear, serious crime, and urban decay.
The strength of a good metaphor is that it puts forward complex and
nuanced ideas in simple and original ways that are easily communicated
and readily recalled. When fresh and vivid, metaphors shock readers into
attention. Criminal justice and criminology are riddled with metaphors –
“white-collar” crime, criminal justice “system,” “wars” against crime and
drugs, “blind justice,” and the “thin blue line” are just a few examples.
Metaphors, however, cut both ways. As the poet Robert Frost has noted:
“All metaphor breaks down somewhere. That is the beauty of it. It is
touch and go with the metaphor, and until you have lived with it long
enough you don’t know where it is going. You don’t know how much
you can get out of it and when it will cease to yield. It is a very living
thing.”1
Using metaphors, as a consequence, is risky. Because they simplify,
metaphors distort as well as reveal. They mask complexity; they call
attention to some aspects of an issue and ignore others; they age; they
“break down somewhere” as Frost puts it; and soon, “everybody knows
what they mean,” regardless of whether everybody does or does not.
As a result, metaphors also easily lend themselves both to misstate-
ment or misrepresentation, either out of ignorance or to serve some
purpose.
Complicating this issue for a metaphor like broken windows, is that
the ideas in broken windows have policy implications and have come to
be practiced: that is, the broken windows metaphor is expressed not just
in words, but in day-to-day action by agencies – most often by public
police, but by other sectors as well. The extent to which these prac-
tices adhere to the spirit, philosophy, and intent of the original broken
 “Broken windows,” criminology, criminal justice 79

windows argument is, of course, open to debate. We have seen many

applications of what is called a broken windows approach that we have
found worrisome. We have also seen and participated in applications of
broken windows of which we are proud: the New York City subway, to
give just one example.

Of broken windows: what are the ideas of broken windows?

Although one can find many of the core ideas of broken windows in
earlier works by James Q. Wilson and Kelling (Wilson 1968; Kelling, Pate,
Ferrara et al. 1981) – as well as many other authors (see, for example,
Jacobs 1961; Glazer 1979) – the most important presentation was the
Atlantic article. What are these core ideas?
1. Disorder and fear of crime are strongly linked; (29–30)
2. Police (in the examples given, foot patrol officers) negotiate rules of
the street. “Street people” are involved in the negotiation of those
rules; (30)
3. Different neighborhoods have different rules; (30)
4. Untended disorder leads to breakdown of community controls; (31)
5. Areas where community controls break down are vulnerable to
criminal invasion; (32)
6. “The essence of the police role in maintaining order is to reinforce the
informal control mechanisms of the community itself.” (34)
7. Problems arise not so much from individual disorderly persons as
from the congregation of large numbers of disorderly persons; (35)
and,
8. Different neighborhoods have different capacities to manage disorder.
(36)
Additionally, the article raises some of the complexities associated with
order maintenance. They include:
1. To what extent can order maintenance be shaped by the rules of neigh-
borhoods rather than criminal law? (34)
2. How do we ensure equity in the enforcement of ordinances so “that
police do not become the agents of neighborhood bigotry”? (35)
3. How is the balance maintained between individual rights and commu-
nity interests? (36)
4. How do we ensure that community controls do not turn into
neighborhood vigilantism? (36)
In 1996, these ideas and issues were again discussed in considerable
detail in Fixing Broken Windows by Kelling and Catherine M. Coles (1996,
with a Foreword by James Q. Wilson). Fixing not only restated these ideas,
it discussed in detail many of the complexities and issues raised by the
80 William H. Sousa and George L. Kelling

ideas of broken windows and their implementation in many communities.

For example, Chapter 2, “The growth of disorder,” is a detailed discus-
sion of the historical and legal issues involved in defining disorder and
balancing individual rights with community interests. We will not bother
here to provide more details about Fixing except to make two points:
first, Fixing explicitly located order maintenance within the context of
community policing and the emerging community prosecution move-
ment. (It will be remembered that community policing was in inchoate
stages in 1981 when the original article was written.) Second, it heavily
emphasized the differential strengths of neighborhoods – the important
consequence of this was that order maintenance policies and activities
were highly discretionary, from administrative policymaking to officers
on the street.
An important question here is: do we (Wilson, Kelling, Coles, and,
more recently, Sousa (see Kelling and Sousa 2001; Sousa 2003)) own
these ideas? Obviously not, in two senses: first, many authors and pro-
grams emphasized police order maintenance long before the original arti-
cle was written. Second, the ideas in broken windows are now “out there”
and readers, academicians, and policymakers are free to make of them
what they wish. Ideas have a life of their own and such is as it should
be. The fact that some broken windows programs are a far cry from any-
thing any of us ever had in mind is simply what happens in the policy
arena. On the other hand, we do have a special claim when critics attack
our written work – especially the original article and Fixing. Critics, at
least academic critics, of broken windows are obligated to “second-order
agreement”: that is, the obligation to reproduce the ideas under question
faithfully, if not enthusiastically. This is not only a matter of good schol-
arship, it is also a matter of professional ethics. Alas, this has not always
happened.
Indeed, many of the academic and legal critiques have not only
distorted broken windows, but they have done so with considerable
zeal and passion. Among other charges, broken windows gives rise
to “wars” on the poor, racism, and police brutality. For one author,
Wilson and Kelling are “aversive racists” (Stewart 1998). Another argues
that Wilson’s and Kelling’s main policy recommendation to police is that
they should “kick ass” (Bowling 1999). To give just one other example,
one of our Rutgers colleagues (not a criminal justice faculty member)
while leading a campus demonstration held up Fixing Broken Windows,
indicating that it was then Mayor Rudolph Giuliani’s blueprint for polic-
ing New York City and, as such, responsible for police killings of citizens.
The question is: Why such misrepresentation and passion?
 “Broken windows,” criminology, criminal justice 81

Of the special ire of criminologists

To answer the question posed above some background is needed. The
dominant criminological and criminal justice paradigm of the past
half-century is that formulated by President Lyndon Johnson’s Presiden-
tial Commission on Law Enforcement and the Administration of Justice.
Its 1967 publication, The Challenge of Crime in a Free Society, endorsed
the “system” model of criminal justice and gave rise to, and framed, crim-
inal justice education to this day (President’s Commission 1967).2 The
underlying assumption that shapes the entire report and its policy and
educational consequences is that crime is caused by structural features
of society: racism, poverty, and social injustice – the “root causes” of
crime.
The assumed causal links among poverty, racism, and crime are woven
throughout the President’s Commission reports. Moreover, the report is
laced with recommendations that deal with such broad societal problems:
schools should be improved; youth should be prepared for employment;
barriers to employment posed by discrimination should be eliminated;
housing and recreational facilities should be improved; minimum family
income should be provided – many, if not all, highly desirable social
policies with which we have no quarrel.
The criminal justice “system” in this model is largely reactive. Police
may patrol neighborhoods, but they do so in a largely non-intrusive
fashion: in cars, remaining “in-service” – that is driving around in a ran-
dom fashion – to ensure that they are available for calls. As the “front
end” of the system, their primary responsibility is to respond to serious
crime through enhanced communication systems. In this view, minor
offenses are either formally decriminalized or virtually decriminalized as
a matter of priorities and policies. Finally, the report is basically silent on
the role of citizens and the community (except to support police), takes
no notice of private security, and disregards the private sector. Crime
control is achieved through broad social/political action to redress the
structural inequities in society and by the activities of a public criminal
justice system that processes offenders.
Crime prevention in this model is equated with grand ideas of social
change: basic societal problems will have to be resolved. As the report
indicates:
Warring on poverty, inadequate housing, and unemployment is warring against
crime. A civil rights law is a law against crime. Money for schools is money
against crime. Medical, psychiatric, and family-counseling services are services
against crime. More broadly and most importantly every effort to improve life
82 William H. Sousa and George L. Kelling

in the “inner city” is an effort against crime. A community’s most enduring

protection against crime is to right the wrongs and cure the illnesses that tempt
men to harm their neighbors. (President’s Commission 1967: 6)

In this view then, we are left with two policy options: change society
and/or process cases. Liberals and conservatives (they have their own
macro approach – restore the family and its values) alike largely accept
this framework. Thus we get extensive debates about sentencing, capital
punishment, the exclusionary rule, mandatory sentencing, “three strikes
you’re out,” prison construction, and so forth, but virtually the entire
debate is within the bounds of the paradigm first put forward by President
Johnson’s Crime Commission.
As such, the criminal justice system paradigm was integral to the 1960s
Great Society.3 Government could solve the problems of poverty, racism,
and social injustice. The Great Society was not just a set of programs,
linked as it was to civil rights, it was a moral cause to which social scientists
were intensely committed and in which they were deeply involved. Crime
prevention was to be a by-product of solving society’s major problems. In
the opinion of many championing this view, disloyalty to the root causes
theory is evidence that one disregards the problems of poverty, racism,
and social injustice. The liberal/left fear is that if disorder, fear, and crime
are uncoupled from root causes, society’s motivation to manage its ills will
be reduced. Whether intended or not, crime prevention is held hostage
to the pursuit of extremely broad social goals, some of which are only
attainable over decades, at best.
So then to return to the question, what explains the special atten-
tion that broken windows has gotten from critics? First, broken windows
defies root cause orthodoxy: that is, to prevent crime one must alleviate
these social ills. It also questions corollary issues and policies: decrimi-
nalization, deinstitutionalization, “victimless” crime, and the views that
only individuals and not neighborhoods can be victims and that indi-
vidual rights almost always trump community interests. Second, broken
windows and crime control success in New York City came out of the
political right (e.g., James Q. Wilson and Mayor Rudolph Giuliani). This
was a hard dose for liberal social scientists to swallow. Hence, the heated
debates about why crime declined in New York City.4 Finally, because of
New York City, broken windows has become widely known, both in pro-
fessional circles and in the popular media. It has become so well known
that often when cited in the popular media it is neither attributed nor
defined. Consequently, a press release or title that “refutes” broken win-
dows, or implies something like it, is more likely to gather attention than
it would otherwise. In other words, some authors are “piggybacking” on
 “Broken windows,” criminology, criminal justice 83

broken windows. We have no quarrel with this – we are in a marketplace

of ideas; we would just caution such authors to represent broken windows
and its implications accurately.

Of research into broken windows

Research on broken windows can be loosely divided into two categories.
The first includes studies that examine the theoretical underpinnings
of the hypothesis, such as the link between disorder and fear or the associ-
ation between incivilities and serious crime. The second includes research
that evaluates policies that are derived from or otherwise influenced by
broken windows, such as quality-of-life programs or order maintenance
enforcement practices. We briefly discuss these two categories.

Broken windows research: the disorder–fear and disorder–crime

connections Broken windows argues that disorderly conditions and
behaviors are linked both to citizen fear and to serious crime. Few crim-
inologists have concern with this disorder-fear portion of the hypothesis,
and a fair amount of empirical research – some of which goes as far back
as research conducted by the President’s Commission during the 1960s –
demonstrates an association between incivilities and fear (see Skogan and
Maxfield 1981 and Ross and Jang 2000 for two more recent examples).
Some debate continues regarding measurement concerns, causal order,
and individual versus ecological level influences on fear. LaGrange, Fer-
raro, and Supancic (1992), for instance, suggest that while incivility is
related to fear, the effect is mediated through perceptions of risk. More
recently, Taylor (2001) concludes that the incivility–fear connection is
stronger at the individual level (one’s perception of incivilities in the
neighborhood has a greater impact than the actual amount of incivili-
ties in the neighborhood) and that the connection is weak when exam-
ined longitudinally (incivilities influence later changes in fear, but not as
strongly or consistently as other factors). Overall, however, most research
to date agrees that disorder is, at least in some way, positively associated
with fear.
Unlike the disorder–fear hypothesis, the disorder-serious crime con-
nection is much more controversial – and far less studied. Skogan (1990)
was the first to find support for the link empirically in Disorder and
Decline. Using primarily survey data from forty neighborhoods in six
cities (Chicago, Atlanta, San Francisco, Philadelphia, Houston, and
Newark (NJ)), Skogan found a highly significant disorder–crime con-
nection while taking into account other factors such as poverty, insta-
bility, and race. Harcourt (1998) has since challenged these findings,
84 William H. Sousa and George L. Kelling

claiming among other concerns that Skogan’s inclusion of several neigh-

borhoods with particularly strong disorder–crime connections (from the
city of Newark) manipulated the overall results. After reproducing the
analyses and removing these neighborhoods from Skogan’s dataset, Har-
court finds that the relationship between disorder and serious crime
disappears.5 Harcourt’s results, however, have also been questioned (see
Xu, Fiedler, and Flaming 2005). Harcourt removed several neighbor-
hoods from the analysis to produce the no-association result, but the
removal of several different neighborhoods from the dataset may have
strengthened the disorder-crime connection found by Skogan (Eck and
Maguire 2000; see also Katz, Webb, and Schaefer 2001). Harcourt’s
manipulation of the data, therefore, does less to disprove Skogan’s results
and more to point out a limitation to the original dataset: its sensitivity
to outliers (Eck and Maguire 2000).
Often cited as the most convincing evidence against the disorder-crime
association (and to the broken windows hypothesis overall) is Sampson
and Raudenbush’s (1999) assessment of the relationship between “collec-
tive efficacy,”6 disorder, and serious crime. The authors use a variety of
methods for their investigation, including systematic social observations
designed to capture disorderly behaviors and conditions on the streets
of Chicago. They challenge the connection between disorder and seri-
ous crime by suggesting that while disorder is moderately correlated with
predatory crime, once antecedent neighborhood constructs (such as col-
lective efficacy) are considered, the direct relationship between the two
all but disappears. Sampson and Raudenbush conclude that the level of
collective efficacy is a strong predictor of both disorder and predatory
crime and that the relationship between incivilities and crime is spu-
rious except for officially measured robbery. According to the authors,
these results “contradict the strong version of the broken windows thesis”
(1999: 637).7
We have several difficulties with the Sampson and Raudenbush study
that involve the authors’ methodological decisions (such as their failure
to observe night-time activities) and their interpretation of the data (such
as their casual dismissal of the robbery finding) (see Kelling 2001), but
others have challenged their analyses as well (see Xu et al. 2005). Jang and
Johnson (2001), for example, argue that Sampson and Raudenbush have
not tested the broken windows theory at all because they misinterpreted
the original thesis and therefore mis-specified their analyses. As Jang and
Johnson point out, broken windows postulates that disorder indirectly
leads to crime via weakened community and neighborhood controls
(stated somewhat differently, Wilson and Kelling (1982) argue that disor-
der, left unchecked by community and neighborhood controls, will lead to more
 “Broken windows,” criminology, criminal justice 85

serious crime). Sampson and Raudenbush, however, assume the thesis

proposes that disorder is directly associated with crime, and so test a model
in which disorder mediates the effects of neighborhood characteristics
(including collective efficacy) on crime rather than neighborhood char-
acteristics mediating the effects of disorder (Jang and Johnson 2001).
Taking into account this misinterpretation, Jang and Johnson estimate
that Sampson and Raudenbush’s assessment actually provides positive
rather than negative support for broken windows.8
In any event, the debate over the link between disorder and crime
remains contentious and research on the topic has produced mixed
results. Interestingly enough, Taylor’s (2001) examination of the
incivility–crime connection seems to verify the inconsistency of previous
research. His longitudinal assessment of Baltimore neighborhoods pro-
vides qualified support for the idea that “grime” leads to crime. He finds,
however, that while disorder influences some later changes in criminal
activity (as well as changes in neighborhood decline and fear), the results
differ across indicators (types of disorder) and across outcomes (types
of crime). Additionally, Taylor finds that other indicators, such as initial
neighborhood status, are more consistent predictors of later crime.9

Broken windows research: policy evaluation A primary policy

implication derived from the original Atlantic article is that if police and
communities are able to manage minor disorders, the result can be a
reduction in criminal activity. As such, activities that can be classified
as broken windows policing,10 which emphasize the assertive enforce-
ment of minor offenses, continue to be implemented in communities
across the country. Subsequent evaluations of broken windows policing
activities and their impact on crime typically consider whether measures
of minor offense enforcement are significantly related with measures of
serious crime reduction.
The New York City Police Department provides perhaps the most
obvious example of a macro policy of order maintenance, as it is well
known that officers were asked to be more assertive in the management of
minor offenses (i.e., typically those offenses that were virtually ignored in
the past).11 That an increase in minor offense enforcement accompanied
a reduction in serious crime in New York helped to spark a continuing
debate: did the strategy contribute to the reduction, or were other factors
involved?
Numerous factors have been offered as potential causes for crime
reduction in New York: changes in demographic and economic trends,
shifts in drug use patterns, statistical regression to the mean, changes
in the cultural values of at-risk populations, and many others.12 Few
86 William H. Sousa and George L. Kelling

studies, however, have examined the extent to which a general order

maintenance strategy has contributed to the crime drop relative to other
factors. Previously, we found that NYPD order maintenance activities in
the 1990s (proxied by arrests for misdemeanor offenses) had a significant
effect on violence reduction in New York net of economic, demographic,
and drug-use variables (Kelling and Sousa 2001). Similarly, Corman and
Mocan (2002), using the same proxy for broken windows policing, found
that misdemeanor arrests had a significant impact on robbery and motor
vehicle theft in New York during the 1990s after controlling for economic
and other criminal justice factors (they do not find a significant impact
on murder, assault, or burglary, however). Further support for a general
strategy of order maintenance is also provided by Worrall (2002), though
not using data from New York. Worrall’s analysis of county-level data from
California demonstrates that an increase in arrests for misdemeanors is
associated with a reduction in felony property offenses independent of
demographic, economic, and deterrence variables.
While these investigations of general broken windows strategies offer
policy insight, the data available for such analyses are often less than ideal.
The limitations of macro level data – often put forward by the authors
themselves – therefore prohibit conclusive statements from these analyses
alone. Several field evaluations, however, have examined the effectiveness
of the strategy as implemented in focused, place-specific initiatives. Green
(1996), for instance, examined Oakland’s Specialized Multi-Agency
Response Team (SMART) program – a problem-solving, place-oriented
strategy that emphasized police coordination with city agencies to enforce
drug nuisance abatement ordinances and other civil laws. Green’s analy-
sis demonstrates that these tactics helped to decrease disorder and drug
problems without significant displacement. Braga, Weisburd, Waring
et al. (1999), in another example, designed a field experiment in Jer-
sey City, NJ, to assess the impact of problem-oriented policing strategies
focused on social and physical disorders in violent places. They concluded
that these strategies were associated with decreases in observed disorders,
citizen calls for service, and criminal incidents.
Other field studies, though perhaps less detailed than those of Green
(1996) and Braga et al. (1999), are less supportive of broken windows
policies. Novak, Hartman, Holsinger, and Turner (1999) analyzed
a police enforcement effort designed to reduce specific disorders –
primarily alcohol and traffic-related offenses – in a community in a
Midwestern city (although it was not designed to necessarily impact seri-
ous crime). The authors determined that the effort was not associated
with a decrease in either robbery or burglary (the two serious crimes they
analyzed) at the target site, although they acknowledge the result may be
 “Broken windows,” criminology, criminal justice 87

due to the duration (one month) and dosage level of the intervention.
More recently, Katz, Webb, and Schaefer (2001) evaluated the impact
of a police quality-of-life program in Chandler, AZ, designed to reduce
social and physical disorders with the intent of decreasing serious crime.
The authors found that, in general, the program had an impact on public
morals disorders (such as prostitution and public drinking) and physical
disorders in the target area. However, the impact on serious crime was
minimal and changes in criminal activity varied by section of the target
area.

Discussion
Whether regarding theory or policy, empirical research on broken win-
dows has produced mixed results. Some academics, attorneys, and crim-
inologists, however, have used the “mixed” results to mount offensives
against the thesis. Their argument goes something like the following:

Studies are inconsistent when it comes to broken windows. Some find the nec-
essary link between disorder and crime, some do not, and some find it only in
certain places and/or for certain types of criminal activity. Because there are sta-
tistically better predictors of crime – such as neighborhood collective efficacy,
neighborhood stability, etc. – policies should concentrate on improvements in
those areas rather than on “fixing broken windows.” Policies based on these bet-
ter predictors can be more effective and are less morally objectionable than the
management of minor offenses.

We argue three points in the remainder of this paper: (1) broken windows
may have merit beyond the link between disorder and crime; (2) claims
that broken windows is morally objectionable, to date, are based on little
actual knowledge of order maintenance in practice; (3) despite criticisms,
broken windows offers a viable policy option within communities.

Broken windows and strong causal reasoning The concepts of dis-

order and serious crime each capture extremely complex sets of activities –
the fact that research is inconsistent concerning the link between the two
is of little wonder. Indeed, it is difficult for us to argue that all instances
of serious crime are the result of social and/or physical incivilities. Our
guess is that as investigations continue into the relationship between
disorder and crime, research will find stronger or weaker associations
as both concepts are disaggregated into their numerous components.
Current research foreshadows this conclusion. In several studies, for
example, robbery has been linked to disorder where other serious crimes
have not (Skogan 1990; Harcourt 1998; Sampson and Raudenbush 1999;
88 William H. Sousa and George L. Kelling

but see Taylor 2001). We also suspect that even different types of robbery
are more or less associated with disorderly conditions and/or behaviors.13
In any event, debates will likely continue as to the strength of the causal
connections between disorder and crime and the policy relevance behind
these connections. But as Thacher (2004) indicates, both proponents
and opponents of “broken windows” have become preoccupied with
the search for strong causal relationships between disorder and criminal
activity – a type of connection that is rarely (if ever) clearly understood in
criminology despite the best efforts of objective social science. Thacher
suggests that by basing the merits of order maintenance on the results of
causal connection studies, criminology avoids a more important moral
question: Is the management of minor offenses justified regardless of its
indirect effects on crime? In other words, is the direct effect of order
maintenance on public order a legitimate public policy goal?
In his analysis, Thacher argues that at least some types of order mainte-
nance policing practices are important as ends unto themselves – regard-
less of their impact on serious crime – because “they address important
instances of accumulative harms and offenses” (2004: 101). Indeed, some
police order maintenance strategies are implemented with no original
intent to reduce serious crime but instead with the goal of restoring pub-
lic order. Restoration of order in the New York City subway provides an
example. While evidence indicates that reductions in both disorder and
serious crime (i.e., robberies) were linked to police order maintenance
efforts (especially against fare-beating and aggressive begging), the polic-
ing effort was initially implemented as an attempt to bring control to an
environment that had grown chaotic (Kelling and Coles 1996). Similarly,
the disorder-reduction program described by Novak et al. (1999) did not
have a substantial impact on serious crime, but it was only intended
to reduce specific incivilities (primarily alcohol and traffic violations) in
response to community complaints about these minor offenses (unfor-
tunately, the authors were unable to assess the intervention’s impact on
these disorders). Even when an order maintenance intervention that is
designed to reduce serious crime fails to do so, this does not necessar-
ily mean the intervention is without merit. For instance, although Katz
et al. (2001) concluded that a disorder reduction program did not have
the intended impact on serious crime, they suggest that the intervention
may still have been worthwhile because it had a significant effect on both
social and physical disorders in the target area.

Broken windows and morally complex policing Of course, even if

we build an argument that public order is a legitimate goal for public pol-
icy and that order maintenance policing can directly benefit public order,
 “Broken windows,” criminology, criminal justice 89

this still leaves the question of whether such a policing strategy is morally
appropriate (Thacher 2004). It is true that broken windows is morally
complex. In the original Atlantic article – as we outlined above – Wilson
and Kelling were greatly concerned with the ambiguities, complexities,
and controversies concerning order maintenance. Likewise Kelling and
Coles discussed the legality and constitutionality of order maintenance
throughout Chapter 2 – the longest chapter in Fixing Broken Windows.
Some critics of broken windows, however, have literally ignored what was
written in the original article and in Fixing and have instead argued that
order maintenance is morally reprehensible. Among other concerns, they
claim that order maintenance policies encourage heavy-handed, “zero-
tolerance” police tactics, or that they criminalize relatively innocuous
behaviors deemed acceptable in communities, or that they dispropor-
tionately affect citizens living in poor and minority neighborhoods.
The difficulty with critics’ arguments, however, is that their assertions
are based on little actual knowledge of order maintenance as implemented
by police managers or as performed by line officers. Thacher (2004)
makes this point strongly. Certainly many critics claim to “know” about
broken windows policing, but their understanding of it appears to come
from dramatized media accounts and either deliberate or careless dis-
tortions of the broken windows metaphor.14 In fact, few have actually
examined order maintenance in practice. Most criminologists/lawyers,
for example, who have attacked the NYPD’s practices, claimed that bro-
ken windows as practiced was morally reprehensible, and dismissed New
York’s crime reductions as due to structural variables, have spent little or
no time “on the ground” either in neighborhoods or with police.
We recently analyzed broken windows policing in New York City
(Kelling and Sousa 2001; Sousa 2003). One of us (Sousa) spent con-
siderable time riding with NYPD and recording observations of officers
as they performed various tasks including those that can be considered
order maintenance. The observations suggest that order maintenance,
at least as performed by NYPD, can best be described as officers paying
attention to minor offenses that were essentially ignored in the past. Some-
times “paying attention” to minor offenses involved formal action – such
as arrest or citation – but more often than not it involved no official action
at all. While officers did not ignore disorderly behavior, they were much
more likely to informally warn, educate, scold, or verbally reprimand cit-
izens who violated minor offenses. Contrary to the claims of critics, we
concluded that officers were mindful of the moral complexities behind
their activities, considered the contexts and circumstances surrounding
incivilities and minor offenses before taking action, and exercised careful
discretion while performing order maintenance tasks.
90 William H. Sousa and George L. Kelling

All may not agree with our assessment of order maintenance in New
York and all are free to disagree with our interpretations of the observa-
tions. The point, however, is that these observations were made with the
intent to develop a more thorough understanding of broken windows in
practice – at least in New York. We share with Thacher (2004) the view
that the merits of broken windows should be evaluated less on causal
connection studies (which are unlikely to produce definitively conclusive
advice for policy) and more on detailed descriptions of order maintenance
as it is practiced. Only the accumulation of more detailed investigations is
likely to shed light on the ethical considerations of applied order mainte-
nance and its impact on disorder in communities. Critics rightly point out
that broken windows policies are morally complex, but until they begin
to develop a more substantive understanding of that which they criticize,
their claims that order maintenance policies are objectionable are noth-
ing more than assertions based on questionable media accounts, dubious
suspicions, and often politically driven speculations. In the end, research
may find that order maintenance, as interpreted by some police depart-
ments or implemented by some police officers, is morally questionable,
but for critics to condemn the practice without sufficient knowledge of it
or on the basis of media representations is professionally irresponsible.

Broken windows and policy options Finally in this section, we wish

to briefly address two points made by some critics of broken windows.
The first is that other policy options, such as problem-oriented polic-
ing, situational crime prevention, and community crime prevention, are
available in lieu of broken windows policies. The second is that because
there are better indicators of crime and crime reduction – for example,
“collective efficacy,” neighborhood stability, etc. – focus should be on
policy improvements in those areas rather than on policies derived from
broken windows.
Regarding the first point, we want to acknowledge that research con-
tinues to show that crime control efforts resulting from problem-solving,
situational crime prevention, and community crime prevention demon-
strate potential at reducing crime and restoring order in communities.15
We do not, nor have we ever, suggested that order maintenance polic-
ing should be implemented instead of these efforts. Quite the contrary,
we believe that order maintenance should represent a policy option in
support of police and community efforts to be implemented as problem-
analysis and problem-solving dictates. In fact, close examination reveals
that this is the reality even in an organization that has a reputation
as an “order maintenance” department: we found numerous examples
of successful NYPD problem-solving efforts driven by the Compstat
 “Broken windows,” criminology, criminal justice 91

process – some of which had virtually no order maintenance quality to

them at all (Kelling and Sousa 2001; Sousa 2003).
Regarding the second point, we wish to point out that many of the
“better” statistical predictors of crime essentially offer little in the way
of policy options. No one will argue, for example, against the desire to
improve neighborhood “collective efficacy,” but notions such as this are
nebulous in both concept and practice. Easing unemployment, poverty,
and racial tensions are highly desirable goals to be sure, but the meth-
ods by which these goals are to be attained are far from certain. Even
if practitioners possessed the knowledge, skill, and resources to effect
change in these areas, it is unclear whether, when, and to what extent
these changes would later impact crime. The “better” statistical predic-
tors of crime may be intellectually informative, but they represent vague
concepts and/or unachievable goals to the practitioner who is tasked with
implementing realistic crime control policies. Broken windows policies,
in contrast, are practical options. They can be implemented as part of a
larger problem-solving agenda, can be employed in a timely fashion, and
can offer the potential for timely results.

Conclusion: of crime, criminology, and

criminal justice
The contrast between the response to broken windows by policymakers
and practitioners on the one hand, and a good portion of criminologists,
on the other, is stark. In the world of policy and practice broken win-
dows has become, for the most part, integral to the conventional wisdom
of community justice – whether it be policing, prosecution, probation
and parole, or community courts. The same largely holds true for the
BID movement and the neighborhood community anticrime movement.
Indeed many of our policymaking and practitioner colleagues find the
responses of many criminologists to broken windows mystifying – for
them a sign that criminologists are simply out of touch with the real
world. They have a point. In fact, they have several points.
First, the root cause ideology has locked criminology and criminal jus-
tice into a practical dead end (it probably is a theoretical dead end as well,
but that is another story). Until the past ten or fifteen years or so, crimi-
nology and criminal justice has had little to offer to crime prevention than
political advocacy/action to achieve a liberal/left version of a “just soci-
ety.” If one thinks of the “big ideas” in criminal justice that have enriched
the recent past, the story is telling. Community policing had its origins in
the work of the American Bar Foundation, the early Police Foundation
(1970–80), and policing itself. Problem-solving, too, had its origins in the
92 William H. Sousa and George L. Kelling

work of the American Bar Foundation, but as refined and articulated

by Herman Goldstein, a public administrator. Situational crime preven-
tion was originated and largely formulated by a psychologist, Ronald
Clarke. Pulling levers – the ideas that led to the dramatic drop in gang
killings in Boston – was largely the product of line police in the Boston
Police Department and shaped by public policy scholar, David Kennedy.
Compstat – the administrative mechanism that addresses problems of
information sharing and accountability in policing – was a product of the
private sector and an entrepreneurial police chief, William Bratton and
some of his closest colleagues in the NYPD (Simons 1995). Broken win-
dows grew out of work by a political scientist, James Q. Wilson, and a social
worker turned police researcher, Kelling. It can be argued that the notion
of hot spots, rooted in the work of Glenn Pierce (see Pierce, Spaar, and
Briggs 1984), was quickly enhanced and broadly promulgated through
the work of contemporary criminologists (see Sherman 1989; Sherman,
Gartin, and Buerger 1989; Weisburd, Maher, and Sherman 1993), but
the initial work, both theoretical and practical, was developed outside of
traditional criminological/criminal justice circles. With few exceptions, it
is difficult to come up with recent criminal justice innovations that have
their origins in criminology.
Second, and this is closely linked to the above, criminology, with its
special interest in why people commit crimes, has co-opted schools of
criminal justice. We have no quarrel with the academic study of why peo-
ple commit crimes, it is a legitimate and important inquiry; however,
we believe it should be properly lodged within sociology (within which
criminology is a specialty). Why criminal justice has been co-opted by
criminology is complex but includes the newness of criminal justice as a
field, the status and tenure structures of universities, the ascendancy of
“Great Society” theoreticians in crime control thinking, and the domi-
nance of root causes ideology during the post-President’s Commission
era. Nonetheless, the idea that university units dedicated to crime con-
trol – after all, it was the crime problem that spurred their origin under
the 1970s Law Enforcement Assistance Administration – should be ded-
icated to criminological pursuits, rather than crime control, was a turn
of events that mired such units into relatively fruitless pursuits – at least
from a public policy point of view.
Finally, criminology and criminal justice have confused scientific stan-
dards of evidence with the evidence that policymakers and practitioners
require in the real world. To be sure, policymakers would love to live in
a world where they were 95 per cent certain that implementing particu-
lar policies or practices would have the desired outcome. As a matter of
fact, they do not and will not. Policymakers live in a world in which they
 “Broken windows,” criminology, criminal justice 93

have to make decisions – many of them, life and death – in which they
are confronted with mixes of problems and programs that do not lend
themselves to clean experiments, bad data, and often conflicting and/or
uncertain research findings. In such a world, 70 or 80 percent certainty
would be a happy thing. Broken windows looks pretty good in this world:
if properly done, it will most probably be approved of by neighborhood
residents, it will probably also reduce their fear of crime, and it looks like
it will reduce some street crimes. Not a bad bet for policymakers and
practitioners.

   
We wish to thank Michael Wagers, Anthony Braga, and David Weisburd for
their helpful comments on earlier drafts of this paper. Thanks also to Huan
Gao of Rutgers University and Jodi Olson of UNLV for their assistance.
1. This quote is taken from Frost’s “Education by poetry.” It can be found on
page 41 of Selected Prose of Robert Frost, ed. Hyde Cox and Edward Connery
Lathem (Holt, Rinehart, and Winston; New York).
2. The idea of a criminal justice “system” was first promulgated by the American
Bar Foundation during the 1950s. Many of the staff persons of the Bar Foun-
dation, e.g., Lloyd Ohlin, Frank Remington, and Herman Goldstein, were
important contributors to the President’s Commission report.
3. Lloyd Ohlin, e.g., was not only the author (with Richard Cloward) of Delin-
quency and Opportunity: A Theory of Delinquent Gangs (1960), one of the key
works giving rise to the War on Poverty and the Great Society, he was also a
member of the President’s Commission on Law Enforcement and the Admin-
istration of Justice.
4. It is interesting that neither the work of Ronald Clarke in situational crime
prevention nor Marcus Felson in routine activities has engendered the hostility
that broken windows has: both are as equally dismissive of root cause and
motivational theories of crime control.
5. After reproducing Skogan’s study, but before removing the Newark neigh-
borhoods from the analysis, Harcourt acknowledges a statistically significant
connection between disorder and robbery. When he eliminates the Newark
neighborhoods, this disorder–robbery connection disappears.
6. Collective efficacy is defined as “the linkage of cohesion and mutual trust
with shared expectations for intervening in support of neighborhood social
control” (Sampson and Raudenbush 1999: 612–613). For purposes of their
study, the collective efficacy measure was created by combining two mea-
sures from survey data: shared expectations for informal social control (rep-
resented by five survey items asking respondents to report the likelihood that
their neighbors would take action given certain scenarios) and social cohesion/
trust (represented by five survey questions asking residents to report on the
trustworthiness, helpfulness, and collegiality of their neighbors).
7. What “the strong version of the broken windows thesis” is remains a mystery
to us, but so be it.
94 William H. Sousa and George L. Kelling

8. Jang and Johnson (2001) themselves find support for broken windows, indi-
cating in their analysis that neighborhood disorder is significantly related to
illicit drug use among adolescents. However, individual “religiosity” (one’s
commitment to religion) and social networks weaken the effect of disorder
on drug use.
9. Some research on the geographic distribution of crime also confirms the
complex nature of the minor offense/serious offense relationship. Weisburd
et al. (1993), for example, suggest that calls for service for minor offenses
(public morals, drunks) correlate more strongly with certain serious offenses
(i.e., robberies) than with others in crime “hot spots.”
10. We use the terms “broken windows,” “quality-of-life,” and “order main-
tenance” interchangeably to describe a policing style that emphasizes the
management of minor offenses. For reasons that are evident in this paper, we
do not consider “zero tolerance” to be synonymous with these terms.
11. In some respects, the idea that “everyone knows” that NYPD is an “order
maintenance” department is unfortunate. While it is true that the manage-
ment of minor offenses is an important strategy in New York, the strategy is
most prominent when used to support focused problem-solving activities that
are often driven by the Compstat process (Kelling and Sousa 2001; Sousa
2003).
12. For a review of the potential causes of crime reduction in New York and
elsewhere during the 1990s, see generally Karmen (1999) and Blumstein and
Wallman (2000). See also Karmen (2000) and Kelling and Sousa (2001).
13. From our perspective, a finding that disorder is linked to robbery in these
studies is both interesting and important for policy. Others, however, in their
apparent zeal to disprove broken windows, gloss over the robbery finding.
Because only robbery is related to disorder, so the argument goes, the thesis
is inherently flawed (never mind that robbery is a “bellwether” crime and
general gauge of violence in many communities). We believe that those who
hold the disorder–crime connection up to such lofty standards suffer from
a similar affliction as those who believe in what Hirschi and Selvin refer to
as the first false criterion of causality: “Insofar as a relation between two
variables is not perfect, the relation is not causal” (1966: 256). Among other
reasons, Hirschi and Selvin argue that this criterion is false because perfect
relations are virtually unknown in criminology. We agree, and thus those who
will settle for little less than a perfect relationship between disorder and crime
as “proof” of broken windows are not likely to find this proof in past, present,
or future research.
14. An example of this comes from Greene (1999) who criticizes New York’s
order maintenance policing as brutal compared to other departments such
as San Diego: “[the comparison] between New York City and San Diego
offers compelling evidence that cooperative police-community problem solv-
ing can provide effective crime control through more efficient and humane
methods” (1999: 185). Comparing New York to San Diego is troublesome at
best, but even if such a comparison were possible, much of Greene’s evidence
for New York’s less “humane” methods is based on little first-hand knowledge
and comes instead from questionable and unreliable sources. For example,
she cites politician Mark Green’s opinions as authoritative on the subject of
 “Broken windows,” criminology, criminal justice 95

New York’s order maintenance policing, but fails to mention Green’s trans-
parent agenda as one of Giuliani’s chief political rivals of the 1990s. Addi-
tionally, while she correctly points out that complaints against the police
increased when order maintenance was introduced in New York (evidence
that she claims supports the New York brutality position), Greene fails to
place this point in its proper context. First, complaints against the police
did increase from 1992 to 1995, but the number of officers also increased
by nearly 10,000 during the same time period. Second, order maintenance
policing necessarily requires more frequent contacts between officers and
citizens – often in situations where the citizen is suspected of some sort of
legal violation. Considering the number of officers added, combined with
the increased frequency of contacts between police and citizens, one might
be surprised if the number of complaints against police did not increase.
Third, while the number of complaints against the police increased until
1995, the number decreased throughout the rest of the 1990s despite the fact
that assertive enforcement of minor offenses continued. Greene’s analysis
exemplifies a general lack of knowledge of broken windows policing – partic-
ularly in New York – by those who claim a competency of it.
15. For a review of this research, see Eck and Maguire (2000) and Weisburd and
Eck (2004).


Blumstein, A. and Wallman, J. (eds.). (2000). The crime drop in America.
Cambridge: Cambridge University Press.
Bowling, B. (1999). The rise and fall of New York murder: zero tolerance or
crack’s decline? British Journal of Criminology, 39 (4), 531–554.
Braga, A. A., Weisburd, D. L., Waring, E. J., Mazerolle, L. G., Spelman, W.,
and Gajewski, F. (1999). Problem-oriented policing in violent crime places:
a randomized controlled experiment. Criminology, 37 (3), 541–580.
Cloward, R. A. and Ohlin, L. E. (1960). Delinquency and opportunity: A theory of
delinquent gangs. New York: The Free Press.
Corman, H. and Mocan, N. (2002). Carrots, sticks and broken windows. NBER
Working Paper 9061. Cambridge, MA: National Bureau of Economic
Research.
Eck, J. E. and Maguire, E. R. (2000). Have changes in policing reduced violent
crime? An assessment of the evidence. In A. Blumstein and J. Wallman (eds.),
The crime drop in America. Cambridge: Cambridge University Press.
Glazer, N. (1979). On subway graffiti in New York. The Public Interest, 54
(Winter), 3–11.
Green, L. (1996). Policing places with drug problems. Thousand Oaks, CA: Sage
Publications.
Greene, J. A. (1999). Zero tolerance: a case study of police policies and practices
in New York City. Crime and Delinquency, 45 (2), 171–187.
Harcourt, B. E. (1998). Reflecting on the subject: a critique of the social influ-
ence conception of deterrence, the broken windows theory, and order-
maintenance policing New York style. Michigan Law Review, 97, 291–389.
96 William H. Sousa and George L. Kelling

Hirschi, T. and Selvin, H. C. (1966). False criteria of causality in delinquency

research. Social Problems, 13, 254–268.
Jacobs, J. (1961). The death and life of great American cities. New York: Vintage
Books.
Jang, S. J. and Johnson, B. R. (2001). Neighborhood disorder, individual reli-
giosity, and adolescent use of illicit drugs: a test of multilevel hypotheses.
Criminology, 39 (1), 109–144.
Karmen, A. (ed.). (1999). Crime and justice in New York City. New York: Primis
Custom Publishing.
(2000). New York murder mystery: The true story behind the crime crash of the
1990s. New York: New York University Press.
Katz, C. M., Webb, V. J., and Schaefer, D. R. (2001). An assessment of the
impact of quality-of-life policing on crime and disorder. Justice Quarterly, 18
(4), 825–876.
Kelling, G. L. (2001). Broken windows and the culture wars: a response to
selected critiques. In R. Matthews and J. Pitts (eds.), Crime, disorder, and
community safety. London: Routledge.
Kelling, G. L. and Coles, C. M. (1996). Fixing broken windows: Restoring order
and reducing crime in our communities. New York: The Free Press.
Kelling, G. L. and Sousa, W. H. (2001). Do police matter? An analysis of the impact
of New York City’s police reforms. Civic Report No. 22. New York: Manhattan
Institute.
Kelling, G. L., Pate, A., Ferrara, A., Utne, M., and Brown, C. E. (1981). Newark
foot patrol experiment. Washington, DC: Police Foundation.
LaGrange, R. L., Ferraro, K. F., and Supancic, M. (1992). Perceived risk and
fear of crime: role of social and physical incivilities. Journal of Research in
Crime and Delinquency, 29 (3), 311–334.
Novak, K. J., Hartman, J. L., Holsinger, A. M., and Turner, M. G. (1999).
The effects of aggressive policing of disorder on serious crime. Policing: An
International Journal of Police Strategies and Management, 22 (2), 171–190.
Oxford English Dictionary (2nd ed.). (1989). Oxford: Clarendon Press.
Pierce, G. L., Spaar, S. A., and Briggs, L. R. (1984). The character of police
work: Implications for the delivery of services. Boston: Center for Applied Social
Research, Northeastern University.
President’s Commission on Law Enforcement and the Administration of Justice
(1967). The challenge of crime in a free society. Washington, DC: US Govern-
ment Printing Office.
Ross, C. E. and Jang, S. J. (2000). Neighborhood disorder, fear, and mistrust: the
buffering role of social ties with neighbors. American Journal of Community
Psychology, 28 (4), 401–420.
Sampson, R. J. and Raudenbush, S. W. (1999). Systematic social observation
of public spaces: a new look at disorder in urban neighborhoods. American
Journal of Sociology, 105 (3), 603–651.
Sherman, L.W. (1989). Repeat calls for service: policing the “hot spots.” In
D. J. Kenney (ed.), Police and policing: Contemporary issues. New York: Praeger
Publishers.
 “Broken windows,” criminology, criminal justice 97

Sherman, L. W., Gartin, P. P., and Buerger, M. E. (1989). Hot spots of predatory
crime: routine activities and the criminology of place. Criminology, 27(1), 27–
55.
Simons, R. (1995). Control in an age of empowerment. Harvard Business Review
(March–April), 80–88.
Skogan, W. G. (1990). Disorder and decline: Crime and the spiral of decay in American
neighborhoods. New York: The Free Press.
Skogan, W. G. and Maxfield, M. G. (1981). Coping with crime: Individual and
neighborhood reactions. Beverly Hills, CA: Sage Publications.
Sousa, W. (2003). Crime reduction in New York City: The impact of the NYPD.
Unpublished doctoral dissertation, Rutgers University, Newark, NJ.
Stewart, G. (1998). Black codes and broken windows: the legacy of racial hege-
mony in anti-gang civil injunctions. Yale Law Journal, 107 (7), 2249–2279.
Taylor, R. B. (2001). Breaking away from broken windows. Boulder, CO: Westview
Press.
Thacher, D. (2004). Order maintenance reconsidered: moving beyond strong
causal reasoning. The Journal of Criminal Law and Criminology, 94 (2), 101–
133.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder,
and fear? The Annals, 593, 42–65.
Weisburd, D., Maher, L. and Sherman, L. (1993). Contrasting crime general and
crime specific theory: the case of hot spots of crime. Advances in Criminolog-
ical Theory, 4, 45–70.
Wilson, J. Q. (1968). Varieties of police behavior. Cambridge, MA: Harvard
University Press.
Wilson, J. Q. and Kelling, G. L. (1982). Broken windows: the police and neigh-
borhood safety. The Atlantic Monthly, March, 29–38.
Worrall, J. L. (2002). Does “broken windows” law enforcement reduce serious crime?
CICG Research Brief. Sacramento, CA: California Institute for County
Government.
Xu, Y., Fiedler, M. L., and Flaming, K. H. (2005). Discovering the impact
of community policing: the broken windows thesis, collective efficacy, and
citizens’ judgement. Journal of Research in Crime and Delinquency, 42 (2),
147–186.
5 Critic
Incivilities reduction policing, zero tolerance,
and the retreat from coproduction: weak
foundations and strong pressures

Ralph B. Taylor

A 2002 New Yorker cartoon depicts two grizzled prisoners whiling away
the day on their bunks. The one on the bottom bunk, presumably in reply
to a question from the inmate in the top bunk, explains, “There might
have been some carelessness on my part, but it was mostly just good police
work.” The inmate on the top bunk seems startled by the admission.
The question to consider here is whether broken windows or incivil-
ity reduction policing is good police work. Broken windows policing is
conceptually grounded on the incivilities thesis. The incivilities thesis,
although it comes in several different guises, suggests that: physical dete-
rioration and disorderly social conduct each contribute independently to
fear, neighborhood decline, and crime; by implication, incivility reducing
initiatives will contribute to neighborhood stability and safety, and lower
fear. To the extent that this logic model is inaccurate, inadequate, or
potentially misleading, incivilities reduction as a set of policing strategies
may fail to deliver. This chapter will summarize the conceptual limita-
tions of that thesis, and the empirical limitations of the supporting work.
It will then broaden the discussion context in two ways: first, to provide
an alternate historical outline of where broken windows policing came
from and, second, to outline the elements of a police–citizen coproduced
process of public safety. Given that context, it sketches the specific chal-
lenges facing successful coproduction over time in an urban residential
context. Some current practices justified on the basis of the incivilities
thesis, such as zero tolerance policing, are probably exacerbating the very
problems earlier versions of these policing strategies sought to alleviate.

Incivilities thesis1

Evolution
As the data from the first National Crime Victimization Surveys appeared
in the early 1970s, researchers started realizing many more residents were

98
 Incivilities reduction policing 99

fearful than were victimized (Garofalo and Laub 1978). “Fear of crime”
was more than “fear” of “crime.” In addition to being worried about
crime per se, people also were more frequently worried about untidy or
disintegrating physical conditions they saw around them, as well as rowdy,
decadent, or just misbehaving adults or teens (Wilson 1975). This version
of the thesis is psychological and cross-sectional.
The best-known elaboration was contributed by Wilson and Kelling.
They suggested, based in part on then-recent foot patrol evaluations
in places like Newark, that unrepaired physical damage suggested an
increasing and increasingly threatening level of “disorder” to local res-
idents (Wilson and Kelling 1982). Over time, they proposed, residents
and other street regulars who watched over the locale would retreat inside
from their monitoring posts, leaving the street vulnerable to increasing
domination by unruly teens and, later, invading serious offenders drawn
in by decreasing guardianship. Wilson and Kelling made the thesis social
psychological or group based, and suggested how consequences unfolded
over time.
Skogan further ecologized the idea, suggesting that deteriorated phys-
ical conditions, and an unruly social climate, could contribute indepen-
dently to community decline, in three different ways: crime rates should
increase faster there, residents would outmigrate faster, leading to struc-
tural decline, and residents’ fear or concern should go up faster (Skogan
1990).

Empirical evidence
This section briefly highlights what we know about the empirical support
for the thesis (for more details, see Taylor 2001: 179–239).
The cross-sectional version of the incivilities thesis receives strong sup-
port at the individual level, especially if the incivilities indicators used are
survey based. Those who see more problems in their locale are also more
fearful, see local crime rates as higher, and are less sanguine about their
neighborhood’s future. These connections persist after controlling for
demographics.
The cross-sectional version at the ecological level receives bivariate
support. Incivilities connect with crime and with fear. Questions arise,
however, when we attempt to partial out basic neighborhood social
structure from incivilities per se because the two connect so strongly,
especially if assessment-based indicators of incivilities are used (Taylor,
Shumaker, and Gottfredson 1985; Taylor 1999). Some earlier cross-
sectional, neighborhood-level analyses thought to support the connection
100 Ralph B. Taylor

(Skogan 1990) have been extensively criticized for analytic mis-steps

(Harcourt 2001: 76).
Two neighborhood-level studies examined the cross-sectional connec-
tions of incivilities and crime using data from the 1990s. In Chicago,
results showed the effects of social disorders, captured through assess-
ments, on robbery but not on homicide (Sampson and Raudenbush
1999). Re-analyses of data from five cities found connections between
perceived social incivilities and robbery and assault rates, but the strength
of the connection was much weakened after controlling for basic neigh-
borhood structure – race, stability, and socioeconomic status (Taylor
1999). In addition, the strength of the connection varied by the crime
in question and by city. Both these studies support the idea that incivil-
ities and crime “share similar theoretical features and are consequently
explained by the same constructs at the neighborhood level” (Sampson
and Raudenbush 1999: 636). They label the antecedents “concentra-
tion of disadvantage and lowered collective efficacy” (ibid.: 636). Given
the potential reciprocal relationship between incivilities and collective
efficacy, it may be premature to conclude collective efficacy is an inde-
pendent contributor to community crime rates. Nonetheless, the most
important point is that “results such as these contradict the strong ver-
sion of the broken windows thesis” (Sampson and Raudenbush 1999:
636).
The cross-sectional, ecological impacts of incivilities on responses to
crime, such as fear, are still not fully understood, but collective efficacy
does appear to be involved in these connections. Neighborhood-level
analyses of several dozen Baltimore neighborhoods showed a bivalent
impact of incivilities. Assessed incivilities directly elevated fear and behav-
ioral restriction, while at the same time they depressed fear and restriction
because they stimulated local involvement and attachment (Taylor 1996).
At this juncture we cannot say with confidence that, at the ecological level,
incivilities connect cross-sectionally with higher fear, as anticipated by the
model; we do not yet have an analysis from one city, or from several cities,
that looks at this connection while simultaneously considering not only
structure, but also social networks, social climate, and collective efficacy.
When we turn to longitudinal work, we have only one neighborhood-
level study with significant time elapsed between initial and follow-
up measures (Taylor 2001). Work with several dozen neighborhoods
in Baltimore showed that initial incivilities demonstrated some pre-
dicted lagged effects on crime, fear, and one pathway of neighborhood
decline. Nonetheless, these impacts were not well patterned across mul-
tiple outcomes, depended on the type of incivility indicator used, and
paled in comparison to the structural impacts of initial racial and status
 Incivilities reduction policing 101

composition. Another longitudinal study at the streetblock level, using a

shorter change time frame of one year, also finds weaker than expected
lagged incivilities effects on responses to crime (Robinson, Lawton,
Taylor, and Perkins 2003). In short, initial incivilities are not strongly
and consistently determinative of later neighborhood crime or structure,
or of later shifts in residents’ responses to crime.
Such murky empirical outcomes naturally lead us back to questions
about whether incivility reduction as a policing strategy will have a long-
term payoff. It may be that “crime fighting” is far more important than
“grime fighting” for improving future neighborhood quality (Taylor 2001:
372).
One final comment deserves mention before leaving questions of the
causal impact of neighborhood incivilities. Some have recently suggested
we should not worry whether order maintenance policing, based on the
incivilities thesis, does or does not have the long-term causal impacts it
is presumed to have (Thacher 2004). According to this view, the polic-
ing activities themselves are inherently worthwhile; they have “intrinsic
merit” (412) as policy tools and deserve closer examination irrespective
of potential impacts.
Such a proposal seems to ignore questions of policy selection. There
are many different forms locale-based policing can take. Order mainte-
nance policing is just one of those. It would seem more useful to find the
police strategies that are the most effective for the long-term outcomes
envisioned, and use those impact assessments to guide strategy selection,
rather than to invest energy in describing strategies that may be less than
optimal.
Empirical evidence speaks not only to questions of hypothesized
impacts, but also to presumed measurement structures. If incivilities
indeed reflect underlying conditions of disorder, different indicators,
from different methods, should correlate strongly with one another. They
don’t always (Taylor 1999; Taylor 2001: 117–121), especially when we
examine change rather than cross-sectional indicators. It is not unusual
for neighborhoods changing in one direction on perceived incivilities to
be changing in the opposite direction on assessed deterioration.
To expand on this last point, longitudinal versions of the incivilities
thesis fail to describe the temporal sequence of described processes.
How long does it take for increased social or physical incivilities to
“inspire” additional serious crime? We currently have in place both the
microecological principles (Taylor 1997), and the statistical procedures
(Weisburd, Bushway, Lum, and Yang 2004) for understanding how
different types of streetblocks change in different ways over time. We
need a series of studies, with continuous longitudinal data available for
102 Ralph B. Taylor

independent, mediating, and outcome variables, to help us further specify

the temporal component of the theory.

Evaluating police work

Considerable debate has ensued about the impacts of New York’s bro-
ken windows policing implemented under Commissioner Bratton, and
the extent to which it was responsible for dramatic crime reductions
in the 1990s (Bratton 1998; Silverman 1999; Blumstein and Wallman
2000). Those arguments are too complex to summarize here. One piece
of work cited in favor of the police strategies, however, uses New York
City precinct and borough-level data to examine changes in crime over
time, and links those changes to increased misdemeanor arrests by police
(Kelling and Sousa 2001). Unfortunately, the statistical analysis in that
work has numerous limitations.2 Further, at that time in New York City
the police were doing many things. In addition to broken windows polic-
ing they were implementing the Compstat process for organizational
review and management of ongoing crime and reduction efforts. It is sim-
ply not possible to separate out the impacts of these different program
components.
To resolve physical problems in a neighborhood related to housing or
street conditions, police need to work alongside other agencies or key
local actors like landlords or small business managers. Although often
referred to as third-party policing rather than broken windows policing,
the focus is often on physical or social incivilities, and lower crime and/or
lower drug sales are an expected outcome. Several studies suggest these
strategies work, although perhaps not as effectively as hoped (Green 1996;
Mazerolle, Kadleck, and Roehl 1998). With such strategies, moreover,
because they are usually multi-agency, working with housing inspectors
and courts, it is often not clear what the contributions are of the police
themselves.

Theoretical concerns
Space limitations preclude more than a brief summary of the major the-
oretical concerns raised by the incivilities thesis, and of the programs put
into place based in part on its tenets (for more detail see Taylor 2001:
95–122; Harcourt 2001: 123–216). Some of the major points of this
discussion are as follows.
What disorder where? As Wilson and Kelling (1982) originally antici-
pated, it is difficult to define one type of disorder deserving police atten-
tion in all locations. The focus of order maintenance policing efforts
 Incivilities reduction policing 103

always will be context-sensitive. Presumably the focus that does emerge

will be driven in large part by the officers’ communications with citizen
and business leaders in the community. Discretion always has been key to
police conduct (Kelling 1999). To say police will focus on different non-
crime problems in one location vs. another represents nothing new. It
does, however, create challenges for police supervisors, and it does place
a premium on police being closely enough connected with local citizen
and business leaders so they have a sense of what is most disturbing to
the community.
Collapse of the harm principle. Harcourt (2001) suggests that elevat-
ing minor misdemeanors to arrest-worthy behaviors, making the merely
annoying dangerous because of its anticipated long-term harmful impacts
on the community, results in a collapse of the harm principle. Stated
differently, previously accepted orderings differentiating more vs. less
serious criminal or deviant behavior get collapsed, creating confusion
rather than changing normative views toward minor misdemeanors. A
May 11, 2001 Associated Press story reported a fifth grader handcuffed
and taken from Oldsmar Elementary School in Florida for drawing
pictures of guns.
Increasing inequality. In addition, we know that those offenses where
officers have the most discretion are often policed in a racially biased
manner. Arrest patterns for minor misdemeanors are likely to introduce
even greater levels of racial inequality into criminal justice processing
(Tonry 1995). Increasing police activity around misdemeanors seems
likely to increase overall racial inequality in this “front end” portion of
criminal justice processing.
Subject creation. Focusing policing efforts on disorderly people widens
the social gap between the haves and the have-nots, making the latter
even less deserving, more criminal, and more dangerous than they are.
It creates an “uncritical dichotomy between disorderly people and law
abiders” (Harcourt 2001: 7). Such a binary schema is not only inaccurate,
but also encourages further social divisiveness.

Where does broken windows policing come from?

George Kelling and others have placed the development of broken win-
dows policing in a specific historical context (Kelling and Coles 1996;
Kelling 1999). That perspective sees police in the 1970s and 1980s as
hamstrung, unable to take care of the minor infractions they saw because
of concerns about citizen complaints and lack of court follow-through. I
propose here an alternate view, one that starts with the urban disorders
of the 1960s, and the various types of policing innovations that followed
104 Ralph B. Taylor

as a subsequent reform response. This amounts to further expanding

the above descriptive context by placing the theoretical evolution of the
broken windows thesis in a broader set of evolving police–citizen copro-
duction strategies. If my mis en scene is correct, then for broken windows
policing to work we must first attend to the essentials of the coproduction
process, and, second, understand the special longitudinal challenges to
that process posed by the urban residential fabric.
Although it is ancient history to anyone less than 50 in 2000, urban dis-
orders rocked dozens of major US cities in the 1960s. Looting and citizen
riots in many cities, and police riots in some cities were later put under
the analytical microscope by high-profile commissions or sociologists
(Fogelson 1968). One of the most widely quoted reports from the time
by the Kerner Commission concluded that policing practices had con-
tributed in part to some of the outbreaks (Kerner 1968).3 The Commis-
sion also pinpointed increasing segregation and structural inequality as
facilitating factors as well, although these were perhaps less remediable.
Consequent concerns about police practices inspired not only federal
initiatives seeking to further professionalize the police; they also encour-
aged widespread thinking about ways in which police department struc-
tures and practices could reduce the distrust and antagonism between
police officers and citizens, especially citizens of lower-income urban
communities of color.
Structural changes emerging from such thinking included the rise of
police-organized community relations councils, often staffed by a com-
munity relations sergeant; additional police–citizen review commissions
with some oversight over some range of police matters; and, later, citizen
police academies. Whether these changes were just “window dressing” or
more substantive, and why the degree of citizen review or oversight varied
across locales are interesting questions, but not ones to be pursued given
the focus here.
More relevant to the focus are a host of strategies first appearing in
the very early 1970s. These can be grouped roughly into two classes –
those emphasizing citizens’ and citizen groups’ roles, and those empha-
sizing how patrolling police interact with citizens. Strategies emphasizing
citizens’ roles included a broad range of community crime prevention
activities: Neighborhood Watch, Citizens on Patrol, Operation ID, secu-
rity surveys, and Operation Whistlestop, to name a few (Rosenbaum
1988). Although the origins of some of these go back to Winchester,
England and the 1200s, the key feature as these emerged in the 1970s
was a neighborhood-based coordinating group working with a designated
police officer. The latter was often a crime prevention specialist or a
community relations officer. Much has been written about whether these
 Incivilities reduction policing 105

programs were successful or not, for what crimes, under what types of
conditions (Rosenbaum 1986; 1987; 1988).
A second class of strategies also emerging in the early and mid-
1970s addressed patrolling officers’ relations with citizenry more gen-
erally. Many departments sought to develop more “stable” relationships
between patrolling officers, local citizens, and local small business per-
sonnel. Patrolling officers were assigned to particular neighborhoods
for an extended period of time with team policing and geographic- or
neighborhood-based policing. The hope was that over time officers would
develop an understanding of how local residents and business lead-
ers viewed crime and related problems in those locations (Greene and
Pelfrey 1997). To deepen this understanding, in the late 1970s researchers
like Robert Trojanowicz suggested flattening the tires on the patrol cars,
giving the police Nikes, and getting them out of their cars. Foot patrolling
officers also were tasked to work cooperatively with local business and
citizen leaders and identify the specific crimes, crime locations, and
crime-related problems deserving attention. Police–citizen contact, coor-
dination, and coproduction were seen as critical conditions for success
(Greene and Taylor 1988). This model, in contrast to the typical commu-
nity crime prevention model, links numerous police officers simultane-
ously with numerous local citizens and local business personnel without
an intervening neighborhood organization.
Progressing into the 1980s and 1990s, these models morphed into
related labels including third-party policing (Mazerolle, Kadleck, and
Roehl 1997; Buerger and Mazerolle 1998), problem-oriented polic-
ing (Spelman and Eck 1987; Eck and Spelman 1989; Goldstein 1990;
Mazerolle, Ready, Terrill, and Waring 2000), and, most difficult to
define, community policing (Cordner 1997). Despite widespread dis-
agreement about what constitutes community policing, and what its goals
are (Greene and Mastrofski 1988), most would agree on the follow-
ing underlying principles (Skogan and Hartnett 1997): “organizational
decentralization” (6) permitting more autonomy to those officers most
directly involved with the locals, and easier police–public communication;
“a commitment to . . . problem-oriented policing” (7); “responsiveness
to citizen input” (8); and building neighborhood capacity to prevent
crime or solve crime-related problems, i.e., to help neighborhood groups
become coproducers of safety (8).
Within this broader evolution, the development of both the prac-
tice and theory behind incivilities reduction policing proves intriguing
as well. On the practice side, the initial policing focus was a variant
of problem-oriented policing (Wilson and Kelling 1982). Following a
problem-oriented SARA approach (Scanning, Analysis, Response and
106 Ralph B. Taylor

Assessment) (Spelman and Eck 1987), police would identify trouble-

some conditions in a locale, and work with citizenry and other agen-
cies to resolve those. Oakland’s “Beat Health” program was an excellent
example of such an approach (Mazerolle, Kadleck, and Roehl 1998).
In short, broken windows policing clearly started out as a coproduction
model. Citizens were involved in problem identification, and other agen-
cies were involved in problem resolution. It was not presumed the police
would board up the abandoned houses themselves. To better understand
this coproduction component, the following section outlines its essential
elements.

Coproduction
Police rely on citizens and other agencies in numerous ways. Most simply,
citizens must report incidents to the police if the police are not right there
to see the crime taking place. So fundamentally, producing public safety
is a coproduction process, wherein police and citizens, and other organi-
zations working with the police all contribute to the outcome (Ostrom,
Parks, Whitaker, and Percy 1979). Ostrom (1996: 1073) defines copro-
duction as: “the process through which inputs used to provide a good or
service are contributed by individuals who are not ‘in’ the same organi-
zation.” Further: “Coproduction implies that citizens can play an active
role in producing public goods and services of consequence to them”
(Ostrom 1996: 1073).
Various factors make coproduction more or less effective than a service
provision model for achieving the intended outcome. Service recipients –
in this case neighborhood residents, neighborhood leaders, and local busi-
ness personnel – need to be active participants in the process (Ostrom
1996: 1079). In addition “both parties must be legally entitled to take
decisions, giving them both some room for manoeuvre . . . [and] partic-
ipants need to build credible commitments to one another (e.g., through
contractual obligations, based on trust or by enhancing social capital)”
(Jeffery and Vira 2001: 9).
Reputations, trust, and reciprocity play pivotal roles in increasing
cooperation between members of the partnership (Ostrom 1998). A
coproductive relationship between citizen leaders and police may cre-
ate policing which is more responsive and more favorably viewed, and
a citizenry more willing to report to the police (Ostrom and Whitaker
1973).
Clearly, significant structural impediments beyond the control of
police departments limit the possibilities of effective police–citizen copro-
duction. Most importantly, police, in contrast to other public sector
 Incivilities reduction policing 107

agencies, are charged with maintaining social control, and administering

“law” (Black 1980). Reciprocity cannot emerge given such a condition,
although mutual responsiveness can. In addition, the citizenry in loca-
tions where effective coproduction is most needed – low income, urban
communities of color – are exactly the same places where distrust between
the police and citizens is most profound (Weitzer 1999; Weitzer and Tuch
1999). The trust requisite for maximally effective coproduction will take
much longer to grow in these locations.
The above limitations aside, in key ways police innovations emerg-
ing since the 1970s contained elements of a coproduction model. They
attempted to increase citizen–police trust, to facilitate stable relationships
between police and local leaders and citizens, and, by tuning police to
local concerns, increase locality-based police responsiveness.

Zero tolerance policing, and back where we started

Unfortunately, some current police practices relying on the incivilities
thesis as justification have moved completely away from a coproduction
idea. The popular zero tolerance policing is a case in point. When police
focus just on that subset of social incivilities – disorderly or drunk peo-
ple, rowdy groups of teens, panhandlers or street vagrants – and seek to
remove them from the street, either through aggressive policing, or fines,
or even arrests, we have zero tolerance policing.
These behaviors are aggressively targeted by police in the belief that
suppressing these street activities will reduce the occurrence in those
places of more serious crimes. The primary focus is on social incivili-
ties, with less attention given to physical conditions. For example, in his
first year on the job Baltimore City Police Commissioner Kevin P. Clark
increased misdemeanor citations about 500 percent (Davis 2004).
The movement toward zero tolerance strategies and away from copro-
duction has taken place gradually over the last twenty years (McArdle and
Erzin 2001). This changing emphasis within the incivilities thesis may be
driven in part by wider societal changes (Garland 2002; Ismaili 2003:
262). Putting the policing shifts in a broader context, we may be see-
ing a historic shift in American and Canadian policing back to the social
control model of the late nineteenth century, and away from the crime
control model which dominated in the second quarter of the twentieth
century (Monkkonen 1981; Boritch and Hagan 1987).
Kelling (1999: 3) has recognized that incivilities models are used to
support zero tolerance strategies, but he argues against such a connection:
“it is an equation that I have never made.”
108 Ralph B. Taylor

Nevertheless, zero tolerance policing policies are widely accepted

because they are legitimated, in the minds of many, by the incivilities the-
sis. Over time the police and the public expect these strategies to reduce
serious crime rates. Police themselves believe this (see Kelling and Sousa
2001, ethnographic observation 2A).
There is some irony here. Post-urban disorder policing innovations
spanning at least three decades have sought to defuse police–community
antagonisms by moving closer to coproduction models. Scrutiny of
police–community relations following the disorders of the 1960s led
to a variety of coproduction models. With community crime preven-
tion initiatives, police were to coordinate with citizenry through local
citizen-led groups, and the members of those groups would serve as the
“eyes and ears” of the police. In return, the police would keep those
groups informed about local crime patterns. Geographic and team polic-
ing were department-based organizational strategies intended to increase
trust of the police among local citizens, citizen leaders, and business
personnel. In the mid-1980s departments sought closer connections
with communities through community policing partnerships; they also
sought to increase their effectiveness through third-party policing and
problem-oriented policing initiatives. These latter innovations required
police to coordinate much more closely than they had in the past with
other local agencies. Such coordination was in recognition of the limits
of police powers, coupled with the hope that solving problems related
to crime could make residents feel better and might even reduce crime
itself.
The perceived importance of problems related to crime such as social
or physical incivilities, has increased steadily through the last two if not
three decades, and this shift links to broader trends (Garland 2002).
Although many of these incivilities require the police to coordinate with
other agencies such as housing or licencing and inspections, there is a
subset of social incivilities against which the police can act directly: people
being disorderly.
Thus, this series of innovations has come full circle. Zero tolerance
policies, ostensibly in the minds of some justified by the incivilities the-
sis, endorse and entrench the aggressive policing thought originally to
contribute to civil disorders in the first place. Despite the ongoing debate
about the sources – from the individual police department up to the
changing social fabric – inspiring this evolution; despite scholars’ refu-
tation of the incivilities thesis/zero tolerance connection; and despite the
empirical weaknesses and theoretical vagaries of the incivilities thesis,
zero tolerance policing is currently popular and thought to be justified by
that thesis.
 Incivilities reduction policing 109

Returning policing innovations to a coproduction model

What is needed at this juncture is to return policing innovations more
generally to a coproduction model. What are the impediments to such a
reintegration?
Major challenges to effective police–community coproduction over time
emerge from the longitudinal texture of the urban residential fabric.
These stumbling blocks, and the empirical evidence pointing them out,
are described in more detail elsewhere (Taylor 2001: 303–357). In brief,
the main points are these. Using qualitative interviews with over sixty
Baltimore neighborhood leaders in the mid-1990s, and examining
changes in official neighborhood boundaries sanctioned by the local
planning department between 1979 and 1995, we gauged the types of
changes in neighborhood names and boundaries taking place, and how
those created challenges for coproduction. For a range of sociological rea-
sons, neighborhood boundaries and names can change over time (Hunter
1971; 1974). We found that both type and extent of change were often
linked to local racial differences, socioeconomic status, or political agen-
das from beyond the neighborhood itself (Logan and Molotch 1987).
We suggested that those changes create substantial difficulties for police–
citizen coproduction of public safety.
We saw that neighborhoods where stable police partnerships were most
needed were those where significant changes were most likely to occur
over a short time in both boundaries and representing organizations. In
such locales police working with community leaders were in danger of
collaborating with decreasingly representative community groups, or of
being confused about the relevant spatial domain, or both.
Ideally, community–police partnerships should be organized at the
neighborhood level. To do so would increase citizens’ buy-in and stake
in the outcome. Partnerships organized around police beats or districts
(e.g., Skogan and Hartnett 1997) end up working with a cluster of local
leaders in each beat, whose constituencies only partially overlap the police
organizing unit. In these situations it is not surprising that local leaders’
commitments often rapidly wane.
If police are to move to a neighborhood-based framework for organizing
partnerships, some sort of stability of neighborhood would seem to be
required. Ongoing neighborhood changes decrease police willingness to
move to that organizing framework, and create difficulties in establishing
working relationships between police and citizens. If police are to engage
in problem reduction, and to be effective and respond in ways that best
benefit the overall community, they need to understand a single spatial
arena and the key players. It takes time working in one locale with one set
110 Ralph B. Taylor

of stakeholders to develop such understanding. Understanding, trust, and

some element of reciprocity are key ingredients for effective coproduction
(Ostrom 1998).
Given the volume and diversity of neighborhood boundary and name
changes observed, and the understandable urban dynamics driving these
changes (Hunter 1974), it appears that successful police-community
coproducing relationships could be developed only in urban or subur-
ban locations with strong overarching neighborhood governance struc-
tures. Many cities have such structures (Hallman 1984; Ferman 1996).
To attempt to stabilize such partnerships without that broader, organizing
and legitimizing political structure, effectively dooms these partnerships.
It limits them to being no more than anemic relationships with no
real commitment either from police or from citizen leaders. Accord-
ing to the coproduction model, it is extremely difficult if not impossi-
ble to coproduce community safety in such settings without that broader
infrastructure.

Summary
The incivilities thesis and police work focusing on incivility reduction
emerged from a preceding tradition of policing innovations, starting in
the early 1970s, in response to urban riots of the 1960s. This emerg-
ing tradition sought to create successful police–community partnerships.
The incivilities thesis has evolved, moving from initial psychological,
cross-sectional formulations to ecological and longitudinal formulations.
Empirical support documenting impacts of incivilities over time on crime,
responses to crime, and neighborhood change, has been weaker and less
consistent than hoped. Empirical investigations documenting salutary
impacts of disorder reduction police work remain few and their rigor
contested. Solid evidence of successful incivilities reduction nested within
third-party policing has emerged, but the specific contribution of police
work itself is not clear. A currently widely practiced version of disorder
reduction policing is zero tolerance policing. It is thought by many to be
justified by the incivilities thesis despite the latter’s originator denying this
connection, and despite the weak support for the thesis. Zero tolerance
repudiates the police–community coproduction model on which the inci-
vilities thesis, and earlier police–community innovations, were grounded.
It is exacerbating the very problems these earlier policing innovations
sought to reduce. To move back to a coproduction model of policing
we will need to solve the problem of changing neighborhood identities,
organizations, and boundaries.
 Incivilities reduction policing 111

   
The author thanks the two editors and Ron Davis for helpful comments on
earlier drafts of the chapter. Address correspondence to RBT, Department of
Criminal Justice, Gladfelter Hall, Temple University, 1115 West Berks Street,
Philadelphia, PA 19122 ([email protected]).
1. More details on this material appear in Taylor (1999; 2001).
2. More specifically: the analysis confounds boroughs and precincts in one level
of analysis rather than separating them out; there are no controls for spatial
autocorrelation effects; and controls for basic ecological fabric appear to be
underspecified, with no variables for racial composition or stability.
3. Of course, this is not the only decade of the twentieth century in which police
activity has been linked to urban riots.


Black, D. (1980). The behavior of law. New York: Academic Press.
Blumstein, A. and Wallman, J. (eds.). (2000). The crime drop in America.
Cambridge: Cambridge University Press.
Boritch, H. and Hagan, J. (1987). Crime and the changing forms of class control:
Policing public order in Toronto the Good, 1859–1955. Social Forces, 66,
307–335.
Bratton, W. (1998). Turnaround. New York: Random House.
Buerger, M. and Mazerolle, L. G. (1998). Third-party policing: A theoretical
analysis of an emerging trend. Justice Quarterly, 15, 301–328.
Cordner, G. (1997). Community policing. In R. G. Dunham and G. P. Alpert
(eds.), Critical issues in policing: Contemporary readings. Prospect Heights, IL:
Waveland Press.
Davis, R. (2004). Quality of life crime splits police, prosecutors. Baltimore Sun,
July 12, A1.
Eck, J. and Spelman, W. (1989). A problem-oriented approach to police ser-
vice delivery. In D. J. Kenney (ed.), Police and policing: Contemporary issues
(pp. 95–103). New York: Praeger.
Ferman, B. (1996). Challenging the growth machine: Neighborhood politics in Chicago
and Pittsburgh. Lawrence: University of Kansas Press.
Fogelson, R. M. (1968). From resentment to confrontation: The police, the
negroes, and the outbreak of the nineteen-sixties riots. Political Science Quar-
terly, 83, 217–247.
Garland, D. (2002). The culture of control: Crime and social order in contemporary
society. Chicago: University of Chicago Press.
Garofalo, J. and Laub, J. (1978). The fear of crime: Broadening our perspective.
Victimology, 3, 242–253.
Goldstein, H. (1990). Problem-oriented policing. Philadelphia: Temple University
Press.
Green, L. (1996). Policing places with drug problems. Thousand Oaks, CA: Sage
Publications.
Greene, J. R. and Mastrofski, S. D. (eds.). (1988). Community policing: Rhetoric
and reality. New York: Praeger.
112 Ralph B. Taylor

Greene, J. R. and Pelfrey, W. V. (1997). Shifting the balance of power between

police and community. In R. G. Dunham and G. P. Alpert (eds.), Criti-
cal issues in policing: contemporary readings. Prospect Heights, IL: Waveland
Press.
Greene, J. R. and Taylor, R. B. (1988). Community-based policing and foot
patrol: Issues of theory and evaluation. In J. R. Greene and S. D. Mastrof-
ski (eds.), Community policing: Rhetoric or reality (pp. 195–224). New York:
Praeger.
Hallman, H. W. (1984). Neighborhoods: Their place in urban life. Beverly Hills, CA:
Sage Publications.
Harcourt, B. E. (2001). Illusion of order: The false promise of broken windows policing.
Cambridge, MA: Harvard University Press.
Hunter, A. (1971). The ecology of Chicago: Persistence and change, 1930–1960.
American Journal of Sociology, 77, 425–443.
(1974). Symbolic communities. Chicago: University of Chicago Press.
Ismaili, K. (2003). Explaining the cultural and symbolic resonance of zero tol-
erance in contemporary criminal justice. Contemporary Justice Review, 6,
255–264.
Jeffery, R. and Vira, B. (2001). Introduction. In R. Jeffery and B. Vira
(eds.), Conflict and cooperation in participatory natural resource management
(pp. 1–16). New York: Palgrave MacMillan.
Kelling, G. L. and Coles, C. M. (1996). Fixing broken windows: Restoring order
and reducing crime in our communities. New York: Free Press.
Kelling, G. L. (1999). Broken windows and police discretion. Research Report
NCJ17859. Washington, DC: National Institute of Justice.
Kelling, G. L. and Sousa, W. H., Jr. (2001). Do police matter?: An analysis of the
impact of New York City’s police reforms. Manhattan Institute Civic Report
No. 22.
Kerner, O. (1968). Report of the National Advisory Commission on Civil Disorders.
New York: Bantam.
Logan, J. R. and Molotch, H. (1987). Urban fortunes. Berkeley, CA: University
of California Press.
Mazerolle, L. G., Kadleck, C., and Roehl, J. (1997). Controlling drug
and disorder problems: The role of place managers. Criminology, 36,
371–404.
Mazerolle, L. G., Ready, J., Terrill, W., and Waring, E. (2000). Problem oriented
policing in public housing: The Jersey City evaluation. Justice Quarterly, 17,
129–158.
Mazerolle, L. G., Kadleck, C., and Roehl, J. (1998). Controlling drug and dis-
order problems: The role of place managers. Criminology, 36, 371–404.
McArdle, A. and Erzin, T. (2001). Zero tolerance: Quality of life and the new police
brutality in New York City. New York: New York University Press.
Monkkonen, E. H. (1981). Police in urban America, 1860–1920. Cambridge:
Cambridge University Press.
Ostrom, E. (1996). Crossing the great divide: Coproduction, synergy and devel-
opment. World Development, 24, 1073–1088.
 Incivilities reduction policing 113

(1998). A behavioral approach to the rational choice theory of collective action:

Presidential address, American Political Science Association, 1997. American
Political Science Review, 92, 1–22.
Ostrom, E. and Whitaker, G. (1973). Does local community control of police
make a difference?: Some preliminary findings. American Journal of Political
Science, 17, 48–76.
Ostrom, E., Parks, R. B., Whitaker, G. P., and Percy, S. L. (1979). The public
service production process: A framework for analysing police services. In
R. Baker and F. A. Meyer, Jr. (eds.), Evaluating alternative law enforcement
policies (pp. 65–73). Lexington: DC Heath.
Robinson, J., Lawton, B., Taylor, R. B., and Perkins, D. D. (2003). Longitudinal
impacts of incivilities: A multilevel analysis of reactions to crime and block
satisfaction. Journal of Quantitative Criminology, 19.
Rosenbaum, D. P. (1986). Community crime prevention: Does it work?, 22. Beverly
Hills: Sage Publications.
(1987). The theory and research behind neighborhood watch: Is it a sound fear
and crime reduction strategy? Crime and Delinquency, 33, 103–134.
(1988). Community crime prevention: A review and synthesis of the literature.
Justice Quarterly, 5, 323–395.
Sampson, R. J. and Raudenbush, S. W. (1999). Systematic social observation
of public spaces: A new look at disorder in urban neighborhoods. American
Journal of Sociology, 105, 603–651.
Silverman, E. (1999). NYPD battles crime: Innovative strategies in policing. Boston:
Northeastern University Press.
Skogan, W. (1990). Disorder and decline: Crime and the spiral of decay in American
cities. New York: Free Press.
Skogan, W. and Hartnett, S. (1997). Community policing, Chicago style. New York:
Oxford University Press.
Spelman, W. and Eck, J. E. (1987). Problem-oriented policing. Washington, DC:
National Institute of Justice.
Taylor, R. B. (1996). Neighborhood responses to disorder and local attachments:
The systemic model of attachment, and neighborhood use value. Sociological
Forum, 11, 41–74.
(1997). Social order and disorder of streetblocks and neighborhoods: Ecology,
microecology and the systemic model of social disorganization. Journal of
Research in Crime and Delinquency, 33, 113–155.
(1999). The incivilities thesis: Theory, measurement and policy. In R. L. Lang-
worthy (ed.), Measuring what matters (pp. 65–88). Washington, DC: National
Institute of Justice / Office of Community Oriented Policing Services.
(2001). Breaking away from broken windows: Evidence from Baltimore neighbor-
hoods and the nationwide fight against crime, grime, fear and decline. New York:
Westview Press.
Taylor, R. B., Shumaker, S. A., and Gottfredson, S. D. (1985). Neighborhood-
level links between physical features and local sentiments: Deterioration, fear
of crime, and confidence. Journal of Architectural Planning and Research, 2,
261–275.
114 Ralph B. Taylor

Thacher, D. (2004). Order maintenance reconsidered: Moving beyond strong

causal reasoning. Journal of Criminal Law and Criminology, 94, 381–414.
Tonry, M. H. (1995). Malign neglect: Race, crime and punishment in America. New
York: Oxford University Press.
Weisburd, D., Bushway, S., Lum, C., and Yang, S. M. (2004). Trajectories of
crime at places: A longitudinal study of street segments in the city of Seattle.
Criminology, 42, 283–321.
Weitzer, R. (1999). Citizens’ perceptions of police misconduct: Race and neigh-
borhood context. Justice Quarterly, 16, 819–846.
Weitzer, R. and Tuch, S. A. (1999). Race, class, and perceptions of discrimination
by the police. Crime and Delinquency, 45, 494–507.
Wilson, J. Q. (1975). Thinking about crime. New York: Basic.
Wilson, J. Q. and Kelling, G. L. (1982). Broken windows. Atlantic Monthly, 211,
29–38.
Part III

Problem-oriented policing
6 Advocate
Science, values, and problem-oriented
policing: why problem-oriented policing?

John E. Eck

Problem-oriented policing is so logical it is surprising it needs justifi-

cation. Problem-oriented policing is based on the premises that (a) the
public demands much of police; (b) the causes of these demands are
often complex; (c) the police serve the public better when they make sys-
tematic inquiries into these complexities; (d) knowledge helps build new
approaches to police services; and (e) learning from successful and unsuc-
cessful innovations makes police more effective in handling the demands
of the public. The validity of these premises can be seen if we pause to look
at their opposites: (a) the public demands little of police; (b) the causes
of these demands are simple; (c) systematic inquiry into the demands is
of little use; (d) knowledge will not help build new approaches to police
services; and (e) there is little to be learned from successes or failures,
so examining them will not improve police effectiveness. None of these
stand up to close scrutiny.
So why is problem-oriented policing often misinterpreted? One rea-
son is that problem-oriented policing fundamentally redefines policing.
It restates the police mission by creating a new unit of analysis for evaluat-
ing police actions: the “problem.” It shifts policing to a scientific approach
to preventing crime and away from the routine application of the law. And
it replaces the notion of the police as gatekeepers to the criminal justice
system with the idea that police are central to many networks that affect
public well-being.
This chapter describes the evolution of problem-oriented policing. I
begin at its origins and then describe its progress. Then I explain why it
is difficult to evaluate. Finally, I describe why there is no viable alternative
to problem-oriented policing.

What is problem-oriented policing?

Shot while playing basketball, a teenager lay bleeding to death just steps away
from a hospital as emergency room workers refused to treat him, saying it
was against policy to go outside. Hospital officials rescinded the policy later
today. (Associated Press 1998)

117
118 John E. Eck

Goldstein created problem-oriented policing to address police discretion.

Police officers are given wide latitude to handle incidents. This gives
them the ability to tailor decisions to the specific needs of situations,
but it allows officers to take inappropriate actions. Unguided discretion
is at the heart of most of the maladies facing policing – for example,
over and improper use of force, ineffective crime reduction procedures,
corruption, and discriminatory practices. Goldstein claimed that discre-
tion could be guided if more were known about the types of problems
police handle. Unfortunately, police administrators focus disproportion-
ately on what the police do rather than on what the police are supposed to
accomplish. This is the “means-over-ends syndrome” (Goldstein 1979).
Goldstein suggested that if police focused more on what they were sup-
posed to accomplish – addressing problems – officers could be provided
with meaningful guidance in how to use their discretion, thus reducing
the chances of inappropriate actions.
The means-over-ends syndrome can be seen best in police use of the
law. Law enforcement is a tool, according to Goldstein (1979), not a goal.
Police and citizens often confuse applying the law with reduction in trou-
ble circumstances. This occurs when policymakers use the application of
the law as the measure of success. In the mid-1990s I heard a new police
chief speak to a group of citizens in my county. After his speech he took
questions. One resident asked why police were arresting homeless men
in the old downtown. The chief answered, “Were they drunk in public?
If they were, they were violating the law and should be arrested.” The
chief did not try to show how these arrests addressed the legitimate con-
cerns the public might have with homeless men – from unsightliness, to
aggressive panhandling, to the health of the homeless men themselves.
The chief implied that enforcing the law trumped the concerns of the
public.
What are the ends that the police should attend to? The police role
is to address the diverse array of troublesome circumstances brought to
it by the public. Goldstein called these circumstances “problems.” The
criminal law is one of many tools the police could apply. How should
police select the right tools for a particular problem? The answer is to
examine the problem, learn why it continues, and select tools that fit the
problem. Success should be measured by problem reduction – fewer, less
serious, and less harmful problems.
Notice that problem-oriented policing contains three principles. The
empirical principle states that the public demands that the police handle
a diverse range of problems. The public is not wedded to a particular
means to reduce problems, but it does demand that they be addressed.
This is widely accepted and will not be examined further.
 Science, values, problem-oriented policing 119

The normative principle claims that police are supposed to reduce prob-
lems rather than simply respond to incidents and apply the relevant crim-
inal law. One can agree that the public demands a great deal of the police,
but it does not follow automatically that the police should deliver all that
is demanded of them. I will come back to this important point later.
The scientific principle asserts that police should take a scientific
approach to problems. Police should apply analytical approaches and
interventions based on sound theory and evidence, just as decisions of
doctors are supposed to be based on medical science. This has implica-
tions for how one scientifically assesses problem-oriented policing. I will
come back to this important point as well.

How has problem-oriented policing evolved?

These principles have remained constant, but problem-oriented policing
has evolved in four ways: its position within police agencies; its theoretical
underpinnings; its definition of “problem”; and in the amount of technical
advice.

Location, scale, and quality

Problem-oriented policing was originally to be located in a headquarters
unit. This unit would examine problems with the goal of providing advice
as to what officers could do to improve their handling of these situations.
A succession of pilot tests in Madison, Wisconsin (Goldstein and Sus-
milch 1982b), London (Hoare, Stewart, and Purcell 1984), Baltimore
County, Maryland (Cordner 1986), and Newport News, Virginia (Eck
and Spelman 1987) moved the location of analysis from headquarters to
police officers assigned to operational units.
Decentralization of problem-solving had consequences. It increased
the number of problems handled by increasing the officers who addressed
them. This reduced the scale of problem analysis from problem classes
to individual manifestations of problem classes – from how to handle
street prostitution, for example, to how to handle street prostitution on a
particular street corner. Decentralization and reduction in scale truncated
inquiry into problems, from the extensive analysis originally envisioned,
to limited probing by busy street cops. These changes reduced the quality
of the average problem-solving effort.
Though the average quality declined, the range of quality probably
widened. At the upper levels some important inquiries have been under-
taken resulting in reductions of serious problems, for example, Boston’s
youth homicide reduction effort (Kennedy, Braga, and Piehl 1997; Braga,
120 John E. Eck

Kennedy, Waring, and Piehl 2001), Liverpool’s reduction of glass bot-

tle injuries (Hester and Rice 2001), and the California Highway Patrol’s
project to reduce farm worker highway deaths (Helmick 2002). At the
bottom level, what is called “problem solving” may simply be targeted
enforcement or referrals to other agencies.
Recently, there has been renewed interest in centralized problem
inquiries to improve quality. Advances in crime analysis have propelled
these units to the center of problem-oriented policing, both as support for
line officers engaged in problem solving, and as leaders of major inquiries
(Boba 2003; Clarke and Eck 2003).

Theoretical underpinnings
Problem-oriented policing requires a theory of problems. By 1987, the
link between problem solving and routine activity theory was appar-
ent (Eck and Spelman 1987). By the early 1990s, routine activity the-
ory became incorporated within problem-solving training through the
“problem-analysis triangle.” The triangle is now a standard part of prob-
lem analysis (Office of Community Oriented Policing Services 1998;
Read and Tilley 2000; Braga 2002; Clarke and Eck 2003). Today,
problem-oriented policing uses a variety of concepts, tools, and proce-
dures from environmental criminology: repeat victim and place analy-
sis, examining property ownership and street configurations, situational
crime prevention and crime mapping, to name but a few (Braga 2002;
Clarke and Eck 2003).

Problem definition
Problem-oriented policing is unique in creating an entirely new concept
defining the goals of policing. Despite its centrality, “problem” was left
undefined at first. Instead, Goldstein provided examples. The term was
used as shorthand for the wide array of issues the public calls upon the
police to handle.
A more precise definition was needed to operationalize problem-
oriented policing. In 1998, the Office of Community Oriented Policing
Services (COPS) of the US Justice Department defined a problem as
“two or more incidents similar in one or more ways that is of concern to
the police and a problem for the community” (Office of Community Ori-
ented Policing Services 1998: 4). Recently, a problem has been defined
as a reoccurring set of similar events, harmful to members of the commu-
nity, that members of the public expect the local police to address. This
definition is summarized by the acronym, CHEERS, for Community,
 Science, values, problem-oriented policing 121

Harm, Expectations, Events, Recurring, and Similarity (Clarke and Eck

2003).
There are many different types of problems. To describe how problems
are related to each other, Ronald V. Clarke and I developed a classifica-
tion system that identifies sixty-six forms of problems on the basis of the
behaviors involved and the environments within which these behaviors
take place (Eck and Clarke 2003).

Improvements in guidance
The transition of problem-oriented policing from theory to practice
necessitated the development of technical advice. The SARA process
(Scanning, Analysis, Response, Assessment), a first effort in this direction
(Eck and Spelman 1987), has become almost synonymous with problem-
oriented policing (Scott 2000). Later, Rana Sampson spearheaded the
dissemination of a problem analysis framework known as the crime or
problem triangle (Office of Community Oriented Policing Services 1998:
9). The triangle was expanded further to include people who can exercise
control over situations (Eck 2003).
There has been a dramatic increase in the number of publications pro-
viding problem-solving guidance. These include general-purpose guides
(Office of Community Oriented Policing Services 1998; Bynum 2001),
as well as Anthony Braga’s (2002) in-depth text on problem solving.
There is now a small library of manuals on technical areas, such as sur-
veys (Eck and LaVigne 1993; Weisel 1999), evaluations (Eck 2002a), and
mapping (Harries 1999; Eck, Chainey, and Cameron 2005). Recently,
the Jill Dando Institute for Crime Sciences published a guide to prob-
lem solving for British crime analysts (Clarke and Eck 2003).1 And in
2004 the Center for Problem-Oriented Policing implemented a website
with an on-line problem-solving simulation and a theory-based protocol
(InPART) to facilitate individual and group problem solving (Center for
Problem-Oriented Policing 2004). Much of this guidance is rooted in
environmental criminology.
The COPS office has also funded a series of problem-specific guides.
Each guide uses research and practices on a problem type (burglaries of
single family homes (Weisel 2002), for example, or rave parties (Scott
2002)) to suggest methods for handling that type of problem.
A quarter of a century since it was first articulated important changes
have taken place in the location of problem-oriented policing, the theory
of problem-oriented policing, the concept of “problem,” and in problem-
solving guidance. Except for organizational location, these changes have
been toward increasing specificity. This stands in stark contrast to other
122 John E. Eck

police reforms, such as community policing, broken windows theory,

and Compstat, which have become less defined over time (Eck and
Rosenbaum 1994; Harcourt 2001; Weisburd, Mastrofski, McNally,
Greenspan et al. 2003).

Should police adopt a problem-oriented approach?

Any social innovation should be judged on the basis of its utility. Though
there is substantial evidence for its effectiveness (Committee to Review
Research 2004; Weisburd and Eck 2004), there is no simple test for
the utility of problem-oriented policing. Unlike a program designed to
address a narrow set of objectives (such as DARE or arrests in misde-
meanor domestic violence calls), any particular evaluation can examine
only a piece of a problem-oriented strategy. Developments in problem-
oriented policing make the strategy a moving target. In addition to these
standard limits on empirical evaluations there are limits particular to
problem-oriented policing.

Problem-oriented policing as a values statement

Policing can be about many things. So the decision to assert that the police
are about addressing problems is a choice among alternative values.
One could claim that the police should only provide emergency ser-
vices. The prevention of crime, disorder, and other public concerns
should be left to individuals, community groups, private markets, other
professions, and other government entities. One could claim that this is
today’s standard form of policing and that this is how policing should be
carried out.
Alternatively, one could claim that policing should be about fighting
crime – both handling its aftermath and reducing future occurrences.
Everything else is a sideshow created by historical accident, inept public
administration, and other correctable factors. Police have shed many for-
mer police functions and so should shed remaining non-crime functions.
One could also assert that police should focus only on separating viola-
tors from non-violators and help to bring the violators before the courts.
From this perspective, decreasing crime is not a police function at all, but
the responsibility of others. If police actions prevent crime this is only an
unintended side benefit.
These normative positions stand independent of the fact that they may
be difficult to achieve. Empirical evidence only helps show what will need
to change in order to achieve these goals.
Similarly, taking a problem-oriented approach is an assertion that the
police should focus on diminishing the harm from the diverse array of
 Science, values, problem-oriented policing 123

issues brought to police attention. People demand a great deal of the

police because other institutions fail. The police must be the agency that
takes the lead in preventing future occurrences because they are the sole
24-hour general-purpose public trouble-shooting agency and have the
greatest capacity to spearhead problem-solving-based prevention.
From a normative perspective problem-oriented policing is untestable.
It is what the police should be doing because it is a natural outgrowth of
the definition of policing. There will always be a need for an organization
that deals with things that go wrong. Fixing these failures often raises the
likelihood that force may be required. So a general-purpose agency with
the ability to use the coercive powers of the state will always be required.
This is policing (Bittner 1980). And if we desire policing to be effective
and responsive to the public, then scientific examination of the problems
is necessary.

Problem-oriented policing as science

Police have used science for decades, most obviously in forensics. That
does not make policing a science, but makes science a tool of policing.
Problem-oriented policing goes further. A problem-oriented approach
attempts to make policing an applied social science. On what grounds
does this statement rest? There are two. First, Goldstein (1979) laid out
four principles: (1) define problems with specificity; (2) study problems
in depth; (3) conduct a broad search for solutions; and (4) focus on
outcomes. These are basic empirical principles of any scientific inquiry.
That problem-oriented policing is an applied science can be seen when
it is put into practice (beyond a cosmetic application). Problem-solving
projects look like applied science. Good problem-solving efforts involve
theory, systematic measurement, comparison, and analysis (Clarke and
Eck 2003). Inquiries into problems were originally envisioned to be joint
police–social scientists endeavors (Goldstein 1990). There are a num-
ber of well-documented case studies illustrating such collaborations (for
example, Goldstein and Susmilch 1982a; Goldstein and Susmilch 1982c;
Eck and Spelman 1987; Forrester, Chatterton, and Pease 1988; Kennedy
et al. 2001; Clarke and Goldstein 2002; Clarke and Goldstein 2003).
Much problem solving may be deficient, but so is much applied science.
This puts problem-oriented policing, as a whole, outside of the domain
of scientific evaluations. Evaluations cannot determine if problem-
oriented policing is effective because to attempt to scientifically eval-
uate problem-oriented policing as a single entity invokes Epimenides’
paradox.2 If problem-oriented policing is scientifically testable, one must
be able to envision a test that could show that it fails to deliver what it
124 John E. Eck

claims. If such a test were conducted, and the core principles of problem-
oriented policing were refuted, then the test also would refute the scien-
tific procedures used to conduct the test. This calls into question the test’s
conclusions. So we have a circular argument.
Scientific tests can show that theories, methods, and interventions
used within a problem-oriented approach do not deliver what they are
supposed to deliver, and that other theories, methods, and interven-
tions hold greater promise. Studies could show that a particular theory
of problems is not as helpful as another theory (social disorganization
theories, for example, are probably of limited utility, but environmental
criminological theories show great promise). They could show that par-
ticular analytical techniques (e.g., mapping) produce more information
about some problems than other problems. Experiments, for example,
can demonstrate that crackdowns on drug hot spots have limited utility
(Sherman and Rogan 1995a; Weisburd and Green 1995; Cohen, Gorr,
and Singh 2003), and interventions with landlords at such places have
greater effectiveness (Weisburd and Green 1995; Eck and Wartell 1998;
Mazerolle, Roehl, and Kadleck 1998). Scientific studies could show that
locating problem analysis in one part of the police organization is dys-
functional, while locating it in other parts of the organization is highly
productive.

A problem-oriented approach as effective policing

Evidence can address important questions about problem-oriented polic-
ing. Two questions in particular are critical. How much analysis is needed
for problem-oriented policing to be more effective than standard policing?
And does expanding police actions beyond law enforcement improve its
effectiveness? If limited analysis and diversity of actions improve police
effectiveness, then incremental adoption of a problem-oriented approach
is feasible. If analysis and diversity of action must increase dramatically
to get a useful payoff, then implementing problem-oriented policing will
be very difficult.

How much analysis is required? Findings from numerous studies

show that even a little analysis makes detectable improvements in police
effectiveness. Let’s examine this question by looking at three unidimen-
sional analyses of problems: repeat places, repeat offenders, and repeat
victims (Eck 2001).
Detecting crime hot spots of is one of the most basic forms of analysis
and evidence shows that acting on such limited analysis helps reduce
crime – compared to acting without such analysis (Braga 2001; Sherman
 Science, values, problem-oriented policing 125

and Rogan 1995b; Sherman and Weisburd 1995). Is there additional

utility for analysis after detecting hot spots? When hot spots are randomly
assigned to problem-solving (treatment) and standard policing, the hot
spots receiving the problem solving fared better on average than those
receiving standard policing (Weisburd and Green 1995; Braga, Weisburd,
Waring et al. 1999). Based on the published descriptions, the analysis
undertaken in the treatment hot spots was relatively limited. So, even
small amounts of analysis improve police effectiveness.
Deterrence works, if it can be applied to specific people. This is a basic
lesson of one of the most ambitious case studies in problem-oriented
policing, the Boston Ceasefire project (Braga et al. 2001; Kennedy et al.
1997). Similar results were found in the Indianapolis Gun Patrol Experi-
ment (McGarrell, Chermak, Weiss, and Wilson 2001). Though the anal-
ysis applied is below the threshold for a problem-oriented approach, it
shows that even a bit of analysis is better than none.
Focusing on repeat victims also reduces crime. Several quasi-
experiments point to the effectiveness of focusing on people who have
previous victimization experience (Farrell 1995). Repeat victimization
programs typically use non-traditional interventions and often apply law
enforcement as a last resort (Anderson, Chenery, and Pease 1995), so
it is difficult to disentangle the effects of analysis from the effects of the
use of alternatives to enforcement. Nevertheless, the conclusions from
the repeat victimization experiences mirror those from repeat place and
repeat offending – even a small amount of analysis is helpful.

How productive is diversifying responses? Repeat victimization pro-

grams show that the use of alternatives to enforcement improves effec-
tiveness. There is evidence from other programs that diversifying police
action beyond enforcement has payoffs. Crackdowns on drug hot spots
produce small to modest crime or disorder reduction (Sherman and
Rogan 1995a), but when coupled with landlord interventions, the impact
is larger (Hope 1994; Green 1995; Eck and Wartell 1998; Mazerolle,
Roehl, and Kadleck 1998). Matthews (1997) noticed a similar phe-
nomenon with prostitution. Early enforcement resulted in short-term
improvements but not sustained change. When enforcement was cou-
pled with changes in street layout he noted longer-lasting and deeper
impacts (Matthews 1993). Recently, Cohen and colleagues (2003) found
that police crackdowns on illegal drug dealing at bars had noticeable but
temporary effects. What would have occurred if the interventions had
targeted the bar owners using a wider variety of interventions? There is a
small but useful set of evaluations suggesting that the impact would have
been greater (Homel, Hauritz, Wortley et al. 1997; Eck 2002b).
126 John E. Eck

In conclusion, even rudimentary application of a problem-oriented

approach produces meaningful positive results, relative to alternatives
(Weisburd and Eck 2004). Though I have separated the effects of analysis
from the application of alternative interventions, alternative interventions
may not be much more effective than enforcement unless coupled with
analysis. Without analysis it is difficult to determine which alternatives
should be applied. So to be really effective both analysis and diversity of
intervention need to be applied.
These conclusions are conservative. The payoff to careful analysis and
diversifying responses may be greater. This is because policing is a human
activity. In every human activity fine distinctions improve effectiveness
because fine distinctions allow carefully tailored actions. And analysis
allows the development of complex sets of actions. Some claim that the
progression from simple to complex is a universal part of all life (Solè and
Goodwin 2000). It would be odd if paying attention to detail and acting
on knowledge did not improve police effectiveness when it improves all
other endeavors.
Additionally, I have assumed that there is a stable body of knowledge.
As noted earlier, the volume and specificity of problem analysis and inter-
vention guidance has increased over the last decade. Problem analysis will
become more penetrating as the knowledge grows and guidance becomes
more specific.
Problem-oriented policing itself is a source of innovation. When police
make systematic inquiries into a problem we learn things. Often what
we learn are small details. On occasion we learn general principles. Two
examples of problem-oriented policing creating new principles are exam-
ined in this volume: focused deterrence (Kennedy in this volume) and
third-party policing (Mazerolle in this volume). Thus, the more problem-
oriented the police become, the more they will learn about problems
and appropriate responses. In short, the practice of problem-oriented
policing produces a positive feedback that creates an increase in police
effectiveness.
All of this suggests that problem-oriented policing can be implemented
incrementally, in small steps.

Is there an alternative to problem-oriented policing?

Problem-oriented policing is not an historical necessity. There is no
logical reason that others could not apply problem solving to pre-
vent crime. High-school students have successfully undertaken problem-
solving activities (Kenney and Watson 1996). Some have argued that
the security industry is far better poised to analyze and address crime
 Science, values, problem-oriented policing 127

problems (Professional Security 2004). Policing could move away from

a problem-oriented approach. But if it does, it will become increasingly
marginalized (Eck 1993).
We have seen this occur with other emergency services. Fire and emer-
gency medical services mitigate the harmful effects of incidents rather
than prevent the incidents. Modern fire services focus on fire suppression
with most prevention left to other private and public agencies. Emergency
medical services respond to vehicle accidents, shooting injuries, and heart
attacks, but do little to prevent these events. Policing could go this route
as well.
But if policing is to have a prevention role, then there is no alterna-
tive to a problem-oriented approach. The standard model of policing
currently applied has limited effect on preventing crime and disorder
(Committee to Review Research 2004; Weisburd and Eck 2004). There
is only limited evidence for the prevention effectiveness of community
policing when a problem-oriented approach is not also used. Community
policing’s value may be in fostering better police–community relations,
improving the legitimacy of police, and promoting democratic principles.
It is a complementary innovation to a problem-oriented approach instead
of a competitor to it. Other innovations, such as hotspots policing, third-
party policing, focused deterrence, and Compstat highlight aspects of a
problem-oriented approach. These are not alternatives, but elaborations
on problem identification, interventions, and management systems that
have been or could be adapted to a problem-oriented approach.
Broken windows theory also may be complementary to a problem-
oriented approach (Kelling and Sousa 2001). If broken windows the-
ory can incorporate analysis and innovation, then it too can be applied
within a problem-oriented approach. This theory is only antithetical to
a problem-oriented approach if it fosters a single response to all police
problems. Because broken windows theory makes no suggestions regard-
ing a variety of police problems such as civil disturbances and traffic
accidents it is more limited than a problem-oriented approach. Another
limitation is that broken windows theory is at best only weakly sup-
ported by evidence (Harcourt 2001; Committee to Review Research
2004).
In the end, there is only one criticism of problem-oriented policing: a
problem-oriented approach is too difficult to implement. If this were valid
then an alternative to a problem-oriented approach would be needed. Any
alternative would have to be simpler than a problem-oriented approach.
But then it will be too simple to be effective for the demands on policing.
And that will bring us back to where we were when Goldstein proposed
a problem-oriented approach.
128 John E. Eck

Fortunately, this criticism is invalid. As we have seen, problem-solving

“light” is better than no problem-solving at all. Progress in policing may
be one baby step after another. Even if change is incremental and slow
that does not make problem-oriented policing unrealistic. Enacting a
problem-oriented approach just requires diligence, hard work, and a great
deal of patience. But then, there is no real alternative.

   
This paper is the result of twenty years of hanging out with some very smart
people. Each deserves thanks. Unfortunately, space and memory limit those
I can mention. So in alphabetical order, “thank you” to Anthony Braga,
Ronald Clarke, Ron Glensor, Herman Goldstein, Johannes Knutsson, Gloria
Laycock, Lorraine Mazerolle, Nancy McPherson, Andy Mills, Graeme
Newman, Rana Sampson, Karin Schmerler, Michael Scott, Lawrence
Sherman, William Spelman, Darrel Stephens, Nick Tilley, Julie Wartell, David
Weisburd, and Deborah Weisel. Tamara Madensen deserves special credit for
helping prepare the manuscript for publication. I alone am responsible for any
errors, omissions, and illogical leaps.
1. The COPS office has funded a companion guide for US crime analysts.
2. Epimenides was a sixth-century BC scholar from Crete who said, “All Cretans
are liars.” This has been sharpened to the Liars paradox, “This statement is
false” (Hofstadter 1979).


Anderson, D., Chenery, S., and Pease, K. (1995). Preventing repeat victimization:
A report on progress in Huddersfield. London: Home Office, Police Research
Group.
Associated Press. (1998, May 19). As dying teen bleeds nearby, hospital staff
stays inside – incident prompts change in policy. Washington Post, p. A9.
Bittner, E. (1980). The functions of police in modern society. Rockville, MD: National
Institute of Mental Health.
Boba, R. (2003). Problem analysis in policing. Washington, DC: Police Foundation.
Braga, A. A. (2001). The effects of hot spots policing on crime. Annals of the
American Academy of Political and Social Science, 578, 104–125.
(2002). Problem-oriented policing and crime prevention. Monsey, NY: Criminal
Justice Press.
Braga, A. A., Kennedy, D. M., Waring, E. J., and Piehl, A. M. (2001). Problem-
oriented policing, deterrence, and youth violence: An evaluation of Boston’s
operation ceasefire. Journal of Research in Crime and Delinquency, 38 (3),
195–225.
Braga, A. A., Weisburd, D. L., Waring, E. J. Mazerolle, L. G., Spelman, W., and
Gajewski, F. (1999). Problem-oriented policing in violent crime places: A
randomized controlled experiment. Criminology, 37 (3), 541–580.
Bynum, T. (2001). Using analysis for problem-solving: A guidebook for law enforce-
ment. Washington, DC: Office of Community Oriented Policing Services.
 Science, values, problem-oriented policing 129

Center for Problem-Oriented Policing. (2004). InPART: Interactive problem anal-

ysis and response tool. Retrieved from www.popcenter.org.
Clarke, R. V. and Eck, J. E. (2003). Become a problem-solving crime analyst: In 55
small steps. London: Jill Dando Institute of Crime Science.
Clarke, R. V. and Goldstein, H. (2002). Reducing theft at construction sites:
lessons from a problem-oriented project. In N. Tilley (ed.), Analysis for crime
prevention, 13 (pp. 89–130). Monsey, NY: Criminal Justice Press.
(2003). Thefts from cars in center-city parking facilities: A case study in imple-
menting problem-oriented policing. In J. Knutsson (ed.), Problem-oriented
policing: From innovation to mainstream, 15 (pp. 257–298). Monsey, NY:
Criminal Justice Press.
Cohen, J., Gorr, W., and Singh, P. (2003). Estimating intervention effects in
varying risk settings: Do police raids reduce illegal drug dealing at nuisance
bars? Criminology, 41 (2), 257–292.
Committee to Review Research on Police Policy and Practices. (2004). Fair-
ness and effectiveness in policing: The evidence. Washington, DC: National
Academies Press.
Cordner, G. W. (1986). Fear of crime and the police: An evaluation of a fear-
reduction strategy. Journal of Police Science and Administration, 14 (2), 223–
233.
Eck, J. E. (1993). Alternative futures for policing. In D. Weisburd and C. Uchida
(eds.), Police innovation and control of the police (pp. 59–79). New York:
Springer Verlag.
(2001). Policing and crime event concentration. In R. Meier, L. Kennedy, and
V. Sacco (eds.), The process and structure of crime: criminal events and crime
analysis (pp. 249–276). New Brunswick, NJ: Transactions.
(2002a). Assessing responses to problems: An introductory guide for police problem-
solvers. Washington, DC: Office of Community Oriented Policing Services,
US Department of Justice.
(2002b). Preventing crime at places. In L. W. Sherman, D. Farrington, B.
Welsh, and D. L. MacKenzie (eds.), Evidence-based crime prevention (pp. 241–
294). New York: Routledge.
(2003). Police problems: The complexity of problem theory, research and
evaluation. In J. Knutsson (ed.), Problem-oriented policing: From inno-
vation to mainstream, 15 (pp. 67–102). Monsey, NY: Criminal Justice
Press.
Eck, J. E. and Clarke, R. V. (2003). Classifying common police problems: A
routine activity approach. In M. J. Smith and D. B. Cornish (eds.), Theory for
practice in situational crime prevention, 16 (pp. 7–40). Monsey, NY: Criminal
Justice Press.
Eck, J. E. and LaVigne, N. (1993). A police guide to surveying citizens and their
environment. Washington: DC: Bureau of Justice Assistance.
Eck, J. E. and Rosenbaum, D. (1994). The new police order: Effectiveness, equity
and efficiency in community policing. In D. Rosenbaum (ed.), The challenges
of community policing: Testing the promises (pp. 3–26). Thousand Oaks, CA:
Sage Publications.
Eck, J. E. and Spelman, W. (1987). Problem-solving: Problem-oriented policing in
Newport News. Washington, DC: Police Executive Research Forum.
130 John E. Eck

Eck, J. E. and Wartell, J. (1998). Improving the management of rental properties

with drug problems: A randomized experiment. In L. G. Mazerolle and J.
Roehl (eds.), Civil Remedies and Crime Prevention, 9 (pp. 161–83). Monsey,
NY: Criminal Justice Press.
Eck, J. E., Chainey, S., and Cameron, J. (2005). Mapping crime: Understanding
hot spots. Washington, DC: US Department of Justice, National Institute of
Justice.
Farrell, G. (1995). Preventing repeat victimization. In M. Tonry and D. P.
Farrington (eds.), Building a safer society: Strategic approaches to crime pre-
vention, 19 (pp. 469–534). Chicago, IL: University of Chicago Press.
Forrester, D., Chatterton, M., and Pease, K. (1988). The Kirkhold burglary pre-
vention project, Rochdale. London: Home Office.
Goldstein, H. (1979). Improving policing: A problem-oriented approach. Crime
and Delinquency, 25 (2), 236–258.
(1990). Problem-oriented policing. New York: McGraw-Hill.
Goldstein, H. and Susmilch, C. E. (1982a). The drinking-driver in Madison:
A study of the problem and the community’s response. Madison, WI: Law
School, University of Wisconsin.
(1982b). Experimenting with the problem- oriented approach to improve
police service: A report and some reflections on two case studies. Madison,
WI: Law School, University of Wisconsin.
(1982c). The repeat sexual offender in Madison: A memorandum on the prob-
lem and the community’s response. Madison, WI: Law School, University
of Wisconsin.
Green, L. (1995). Policing places with drug problems: The multi-agency response
team approach. In J. E. Eck and D. Weisburd (eds.), Crime and place, 4
(pp. 199–216). Monsey, NY: Criminal Justice Press.
Harcourt, B. E. (2001). Illusion of order: The false promise of broken windows policing.
Cambridge, MA: Harvard University Press.
Harries, K. (1999). Mapping crime: Principle and practice. Washington, DC:
National Institute of Justice, Crime Mapping Research Center.
Helmick, D. O. (2002). Safety and farm labor vehicle education program. Sacra-
mento, CA: California Highway Patrol.
Hester, J. and Rice, K. (2001). Operation crystal clear: In an effort to reduce
glass related street violence, the message remains ‘crystal clear’. Liverpool,
UK: Merseyside Police.
Hoare, M. A., Stewart, G., and Purcell, C. M. (1984). The problem oriented
approach: Four pilot studies. London: Metropolitan Police, Management
Services Department.
Hofstadter, D. R. (1979). Gödel, Escher, Bach: An eternal golden braid. New York:
Vintage.
Homel, R., Hauritz, M., Wortley, R., McIlwain, G., and Carvolth, R. (1997). Pre-
venting alcohol-related crime through community action: The surfers par-
adise safety action project. In R. Homel (ed.), Policing for prevention: Reducing
crime, public intoxication and injury, 7 (pp. 35–90). Monsey, NY: Criminal
Justice Press.
 Science, values, problem-oriented policing 131

Hope, T. (1994). Problem-oriented policing and drug market locations: Three

case studies. In R. V. Clarke (ed.), Crime prevention studies, 2 (pp. 5–31).
Monsey, NY: Criminal Justice Press.
Kelling, G. L. and Sousa, W. H., Jr. (2001). Do police matter? An analysis of the
impact of New York City’s police reforms. Civic Report. New York: Manhat-
tan Institute for Policy Research.
Kennedy, D. M., Braga, A. A., and Piehl, A. M. (1997). The (un)known universe:
Mapping gangs and gang violence in Boston. In D. Weisburd and T. McEwen
(eds.), Crime mapping, 8 (pp. 219–262). Monsey NY: Criminal Justice
Press.
(2001). Developing and implementing operation ceasefire. In National Insti-
tute of Justice (ed.), Reducing gun violence: The Boston gun project’s operation
ceasefire. Washington, DC: National Institute of Justice.
Kenney, D. and Watson, T. S. (1996). Reducing fear in the schools: Managing
conflict through student problem solving. Education and Urban Society, 28
(4), 436–455.
Matthews, R. (1993). Kerb-crawling, prostitution and multi-agency policing.
London: Home Office.
(1997). Developing more effective strategies for curbing prostitution. In R.
V. Clarke (ed.), Situational crime prevention: Successful case studies, 2nd ed.
(pp. 74–82). Guilderland, NY: Harrow and Heston.
Mazerolle, L.G., Roehl, J., and Kadleck, C. (1998). Controlling social disor-
der using civil remedies: Results from a randomized field experiment in
Oakland, California. In L. Mazerolle and J. Roehl (eds.), Civil remedies and
crime prevention, 9 (pp. 141–160). Monsey, NY: Criminal Justice Press.
McGarrell, E. F., Chermak, S., Weiss, A., and Wilson, J. (2001). Reducing
firearms violence through directed police patrol. Criminology and Public
Policy, 1 (1), 119–148.
Office of Community Oriented Policing Services. (1998). Problem-solving tips:
A guide to reducing crime and disorder through problem-solving partnerships.
Washington, DC: US Department of Justice, Office of Community Oriented
Policing Services.
Professional Security. (2004, January). Manual of crime science: New breed aims
to outwit the criminals. Professional Security, 64.
Read, T. and Tilley, N. (2000). Not rocket science? Problem-solving and crime reduc-
tion. London: Home Office.
Scott, M. S. (2000). Problem-oriented policing: Reflections on the first 20 years.
Washington, DC: US Department of Justice, Office of Community Oriented
Policing Services.
(2002). Rave parties. Washington, DC: Office of Community Oriented Policing
Services.
Sherman, L. W. and Rogan, D. P. (1995a). Deterrent effects of police raids on
crack houses: A randomized, controlled, experiment. Justice Quarterly, 12
(4), 755–781.
(1995b). Effects of gun seizures on gun violence: ‘Hot spots’ patrol in Kansas
City. Justice Quarterly, 12 (4), 673–693.
132 John E. Eck

Sherman, L. W. and Weisburd, D. (1995). General deterrent effects of police

patrol in crime “hot spots”: A randomized, controlled trial. Justice Quarterly,
12 (4), 625–648.
Solè, R. and Goodwin, B. (2000). Signs of life: How complexity pervades biology.
New York: Basic Books.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder
and fear? Annals of the American Academy of Political and Social Science, 593,
42–65.
Weisburd, D. and Green, L. (1995). Policing drug hot spots: The Jersey City
drug market analysis experiment. Justice Quarterly, 12 (4), 711–735.
Weisburd, D., Mastrofski, S., McNally, A. M., Greenspan, R., and Willis, J. J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2 (3), 421–456.
Weisel, D. (1999). Conducting community surveys: A practical guide for law enforce-
ment agencies. Washington, DC: Bureau of Justice Statistics and Office of
Community Oriented Policing.
(2002). Burglary of single-family homes. Washington, DC: Office of Community
Oriented Policing Services.
7 Critic
Problem-oriented policing: the disconnect
between principles and practice

Anthony A. Braga and David Weisburd

Problem-oriented policing works to identify why things are going wrong

and to frame responses using a wide variety of innovative approaches
(Goldstein 1979). Using a basic iterative approach of problem identifi-
cation, analysis, response, assessment, and adjustment of the response,
this adaptable and dynamic analytic approach provides an appropriate
framework to uncover the complex mechanisms at play in crime prob-
lems and to develop tailor-made interventions to address the under-
lying conditions that cause crime problems (Eck and Spelman 1987;
Goldstein 1990). Researchers have found problem-oriented policing to
be effective in controlling a wide range of specific crime and disorder
problems, such as convenience store robberies (Hunter and Jeffrey 1992),
prostitution (Matthews 1990), and alcohol-related violence in pubs and
clubs (Homel, Hauritz, Wortley et al. 1997). Indeed, there is very promis-
ing evidence of the effectiveness of the approach (Braga 2002; Committee
to Review Research 2004; Weisburd and Eck 2004).
But is the “problem-oriented policing” that researchers have evaluated
similar to the model of problem-oriented policing that its originators
proposed? There is substantial evidence that, too often, the principles
envisioned by Herman Goldstein are not being practiced in the field.
Deficiencies in current problem-oriented policing practices exist in all
phases of the process. A number of scholars have identified challenging
issues in the substance and implementation of many problem-oriented
policing projects, including: the tendency for officers to conduct only a
superficial analysis of problems and then rush to implement a response,
the tendency for officers to rely on traditional or faddish responses rather
than conducting a wider search for creative responses, and the tendency
to completely ignore the assessment of the effectiveness of implemented
responses (Cordner 1998; but also see Clarke 1998; Read and Tilley;
Scott and Clarke 2000). Indeed, the research literature is filled with a
long history of cases where problem-oriented policing programs tend
to lean toward traditional methods and where the problem-solving pro-
cess is weak (Goldstein and Susmilch 1982; Eck and Spelman 1987;

133
134 Anthony A. Braga and David Weisburd

Buerger 1994; Capowich, Roehl, and Andrews 1995; Read and Tilley
2000). In his recent review of several hundred submissions for the Police
Executive Research Forum’s Herman Goldstein Award for Excellence
in Problem-Oriented Policing, Clarke (1998) laments that many recent
examples of problem-oriented policing projects bear little resemblance to
Goldstein’s original definition. Eck (2000) comments that the problem-
oriented policing that is practiced today is but a shadow of the original
concept.
In this chapter we examine the nature of problem solving and problem-
oriented policing as it is practiced in the field. Our main conclusion is that
there is a disconnect between the rhetoric and reality of problem-oriented
policing, and that this is not likely to change irrespective of the efforts of
scholars and policymakers. Indeed, we take a very different approach to
this problem than others who have examined the deficiencies of problem-
oriented approaches. We argue that there is much evidence that what
might be called “shallow” problem-solving responses can be effective in
combating crime problems. This being the case, we question whether the
pursuit of problem-oriented policing, as it has been modeled by Goldstein
and others, should be abandoned in favor of the achievement of a more
realistic type of problem solving. While less satisfying for scholars, it is
what the police have tended to do, and it has been found to lead to real
crime prevention benefits.

Defining the problem

To understand our argument it is useful to review the problems that
police encounter at each stage of the well-known SARA model (Scanning,
Analysis, Response, Assessment) that was developed in Newport News,
Virginia, to crystallize the problem-oriented process (Eck and Spelman
1987). In practice, many problem-oriented policing projects do not follow
the linear and separate steps of the SARA model. The steps tend to occur
simultaneously or in a fashion in which officers move back and forth
between scanning, analysis, and responses (Capowich and Roehl 1994;
Braga 1997). Nevertheless, the model provides a useful way to structure
our discussion of the disconnect between principles and practice in the
problem-oriented policing process.

Scanning
Scanning involves the identification of problems that are worth look-
ing at because they are important and amenable to solution. Herman
Goldstein (1990) suggests that the definition of problems be at the
 Problem-oriented policing 135

street level of analysis and not be restricted by preconceived typologies.

Goldstein further clarifies what is meant by a problem by specifying the
term as: “a cluster of similar, related, or recurring incidents rather than
a single incident; a substantive community concern; or a unit of police
business” (1990: 66). To ensure that problems are specified correctly, it
is important to break down larger categories of crime into more specific
kinds of offenses (Clarke 1997). For example, auto theft should be fur-
ther specified into particular types of auto theft such as “joy riding” or
“stripping for auto parts.” Close analysis of these specific problems may
lead to very different types of intervention that might be more appropriate
for each component of an auto theft problem when compared to a more
general approach.
There are many ways a problem might be nominated for police atten-
tion. A police officer may rely upon his or her informal knowledge of a
community to identify a problem that he or she thinks is important to
the well-being of the community. Another approach to identifying prob-
lems is through consultation with community groups of different kinds,
including other government agencies. Another possibility is to identify
problems from the examination of citizen calls for service coming into
a police department. This approach is implicitly recommended by those
who advocate “repeat call analysis” or the identification of “hot spots”
(Sherman, Gartin, and Buerger 1989; Weisburd and Mazerolle 2000).
The notion is that citizens will let the police know what problems concern
them by making calls as individuals. Crime mapping has become a very
popular way for police departments to identify existing crime problems
(Weisburd and McEwen 1997). A recent Police Foundation report found
that seven in ten departments with more than 100 sworn officers reported
using crime mapping to identify crime hot spots (Weisburd, Mastrofski,
and Greenspan 2001).
Relative to their performance in other phases of the process, police
officers are generally good at identifying problems (Bynum 2001). How-
ever, Clarke (1998) suggests that problem-oriented police officers often
fail to make appropriate specifications of the problems they are address-
ing in one of two ways: they undertake a project that is either too small
or too big. Small, beat level projects, such as dealing with a confused
lonely old man who is seeking some daily companionship by repeatedly
calling the police for a variety of concerns, may be better handled as a
citywide project addressing older citizens who live alone and who gener-
ate a large number of calls for trivial matters (Clarke 1998). Some overly
ambitious initiatives, such as dealing with “gang delinquency” or focus-
ing on a “problem neighborhood,” do not represent a single problem, but
a collection of problems. For example, in the “problem neighborhood,”
136 Anthony A. Braga and David Weisburd

there could be a diverse set of problems such as drug markets, auto theft,
and domestic violence. These are separate problems that require indi-
vidual attention. Problems in problem-oriented policing programs are
sometimes broadly identified; this destroys the discrete problem focus of
the project and leads to a lack of direction at the beginning of analysis
(Clarke 1998).

Analysis
The analysis phase challenges police officers to analyze the causes of
problems behind a string of crime incidents or substantive community
concern. Once the underlying conditions that give rise to crime problems
are known, police officers develop and implement appropriate responses.
The challenge to police officers is to go beyond the analysis that natu-
rally occurs to them; namely, to find the places and times where par-
ticular offenses are likely to occur, and to identify the offenders who
are likely to be responsible for the crimes. Although these approaches
have had some operational success, this type of analysis usually produces
directed patrol operations or a focus on repeat offenders. The idea of
analysis was intended to go beyond this. Unfortunately, as Boba (2003)
observes, while problem-oriented policing has blossomed in both con-
cept and practice, problem analysis has been the slowest part of the
process to develop. In his twenty-year review of problem-oriented polic-
ing, Michael Scott (2000) concludes that problem analysis remains the
aspect of problem-oriented policing that is most in need of improve-
ment. The Police Executive Research Forum’s national assessment of the
US Community Oriented Policing Services (COPS)-sponsored Problem
Solving Partnerships program also found that problem analysis was the
weakest phase of the problem-oriented policing process (Police Executive
Research Forum 2000).
Bynum (2001) suggests that police have difficulty clearly defining prob-
lems, properly using data sources, conducting comprehensive analyses,
and implementing analysis-driven responses. Some officers skip the anal-
ysis phase or conduct an overly simple analysis that does not adequately
dissect the problem or does not use relevant information from other agen-
cies (such as hospitals, schools, and private businesses) (Clarke 1998).
Based on his extensive experience with police departments implement-
ing problem-oriented policing, Eck (2000) suggests that much problem
analysis consists of a simple examination of police data coupled with
the officer’s working experience with the problem. In their analysis of
problem-oriented initiatives in forty-three police departments in England
and Wales, Read, and Tilley (2000) found that problem analysis was
 Problem-oriented policing 137

generally weak with many initiatives accepting the definition of a problem

at face value, using only short-term data to unravel the nature of the prob-
lem, and failing to adequately examine the genesis of the crime problems.
Analyzing problems at crime hot spots may present a particularly vexing
challenge to problem-oriented police officers. Given the contemporary
popularity of crime mapping in the identification of problems worthy of
police attention, we feel that it is important to highlight the difficulties
in analyzing problems at hot spots. High-activity crime places tend to
have multiple problems and the problems at crime places can be quite
complex and involved. In their close analysis of hot spots in Minneapolis,
Weisburd and his colleagues (1992) suggest that a heterogeneous mix of
crime types occur at high-activity crime places rather than a concentration
of one type of crime occurring at a place. In their examination of problem-
oriented policing in San Diego, Capowich and Roehl (1994) observed
that multiple problems tend to coincide at places and report, “at the beat
level, there are no pure cases in which the problem can be captured under
a single classification. The range of problems is wide, with each presenting
unique circumstances” (144). In Oakland’s Beat Health program to deal
with drug nuisance locations, officers encountered difficulties unraveling
what was happening at a place and deciding how it should be addressed
(Green 1996).
In the Jersey City (NJ) problem-oriented policing in violent crime hot
spots experiment, the number of identified problems per place ranged
from three to seven, with a mean of 4.7 problems per place (Braga 1997;
Braga, Weisburd, Waring et al. 1999). Table 7.1 presents the problems
identified by officers for further analysis and the key characteristics of vio-
lent crime hot spots in the Jersey City problem-oriented policing exper-
iment. From their training and the reading materials made available to
them, the problem-oriented officers expected that they would be prevent-
ing violence at each place by focusing on very specific underlying charac-
teristics or situational factors. After examining places closely, the officers
observed that they would be controlling a multitude of crime problems.
All twelve places were perceived to suffer from social disorder problems
such as loitering, public drinking, and panhandling; eleven places suf-
fered from physical disorder such as trash-filled streets, vacant lots, and
abandoned buildings. Seven places had problems with illicit drug selling
and three places had problems with property crimes.
Across and within places in the Jersey City problem-oriented policing
experiment, few problems were analyzed thoroughly (Braga 1997). Eck
and Spelman (1987) suggest two classifications for the depth of problem
analysis: limited analysis and extended analysis. Eck and Spelman (1987)
grouped problem-solving efforts by the Newport News (VA) Police
138 Anthony A. Braga and David Weisburd

Table 7.1 Problems and characteristics of 12 violent crime places

Place Problems Relevant Characteristics

1 Assault and robbery of commuters Train and bus terminal

Shoplifting Restaurants and retail stores
Pick-pocketing Abandoned buildings
Drug selling Abandoned automobiles
Homeless loitering and panhandling Holes in fences around terminal
Disorderly groups of youths Poor lighting
Physical disorder Piles of lumber used by loiterers
Piles of trash deposited by dumpers
2 Street fights Major thoroughfare
Drug market Bodega
Public drinking and drug use Tavern
Disorderly groups of youths Abandoned buildings
Physical disorder Vacant lot
3 Assault and robbery of college students College campus
Burglary and larceny on campus Multiple bus stops
Loitering Trash-strewn vacant lot w/tall weeds
Minor drug selling Low-income apartment building
Physical disorder Trash-strewn alley
Poor lighting
4 Assault and robbery of students Grammar/middle school
Car-jacking Community college
Homeless loiterers Retail stores
Disorderly groups of youths Major intersection
5 Robbery of retail stores Retail stores
Street fights Restaurants
Disorderly groups of youths Major thoroughfare
Public drinking Trash-strewn streets
Physical disorder
6 Drug market Liquor store
Loitering Major thoroughfare
Public drinking Poor lighting
Physical disorder
7 Robbery of elderly Senior citizen housing complex
Disorderly groups of youths Low-income housing project
Physical disorder Interstitial area between middle-class
and lower-class neighborhoods
Poor lighting
Bus stop
8 Drug market Large low-income apartment building
Street fights Tavern
Bar fights Retail stores
Disorderly groups loitering
Public drinking
Physical disorder
 Problem-oriented policing 139

Table 7.1 (cont.)

Place Problems Relevant Characteristics

9 Active indoor and outdoor drug market Abandoned buildings

Street fights Low-income apartment building
Loitering Major thoroughfare
Public drinking Vacant lots
Physical disorder
10 Robbery of convenience stores Retail stores
Disorderly groups of youths Major thoroughfare
Loitering Trash-strewn streets
Burglaries
Physical disorder
11 Street fights Abandoned buildings
Drug market Residential neighborhood
Public drinking Poor lighting
Loitering Vacant lots
Physical disorder Graffiti
Trash-strewn street
12 Drug market Park
Bar fights Taverns
Street fights Poor lighting
Public drinking Retail stores and restaurants
Loitering Trash-strewn streets
Disorderly groups of youths
Physical disorder

Department by determining whether there were obvious information

sources that were not used, given the nature of the problem; if there were
not any obvious unused sources, the effort was classified as extended.
Using these definitions, slightly less than one third (31.1 percent;
19 of 61) of the identified problems in the Jersey City study received
what could be described as an extended analysis. It must be noted that,
for certain problems, a superficial analysis was all that was necessary (e.g.,
alleviating a trash problem by recognizing that there were no trash recep-
tacles at the place). Weak problem analyses occurred for two reasons.
First, at ten of the twelve places, the officers believed that they “knew
what was going on” based on their working knowledge and “on the spot”
appraisals of problems. Second, the officers believed that most of the
street crime at the place could be linked directly or indirectly to the phys-
ical and social disorder of a place. From their problem-oriented policing
training, the officers were familiar with the broken windows thesis and
much preferred the simplicity of a general plan to restore order and reduce
140 Anthony A. Braga and David Weisburd

crime at places over the specifics of the SARA model to control the mul-
tiple and interrelated problems of a place. Rather than conducting rigor-
ous analysis of crime problems and developing tailor-made solutions, the
officers generally attempted to control their places via aggressive order
maintenance and making physical improvements such as securing vacant
lots or removing trash from the street.

Responses
After a problem has been clearly defined and analyzed, police officers
confront the challenge of developing a plausibly effective response. The
development of appropriate responses is closely linked with the analysis
that is performed. The analysis reveals the potential targets for an inter-
vention, and it is at least partly the idea about what form the intervention
might take that suggests important lines of analysis. As such, the reason
police often look at places and times where crimes are committed is that
they are already imagining that an effective way to prevent the crimes
would be to get officers on the scene through directed patrols. The rea-
son they often look for the likely offender is that they think that the most
effective and just response to a crime problem would be to arrest and
incapacitate the offender. However, the concept of “problem-oriented
policing” as envisioned by Herman Goldstein (1990), calls on the police
to make a much more “uninhibited” search for possible responses and
not to limit themselves to getting officers in the right places at the right
times, or identifying and arresting the offender (although both may be
valuable responses). Effective responses often depend on getting other
people to take actions that reduce the opportunities for criminal offend-
ing, or to mobilize informal social control to drive offenders away from
certain locations.
The available research evidence suggests that the responses of many
problem-oriented policing projects rely heavily upon traditional police
tactics (such as arrests, surveillance, and crackdowns) and neglect the
wider range of available alternative responses (Clarke 1998). Many eval-
uations of problem-oriented policing efforts have documented a prepon-
derance of traditional policing tactics (see e.g., Capowich and Roehl
1994; Cordner 1994). Read and Tilley (2000) found that officers selected
certain responses prior to, or in spite of, analysis; failed to think through
the need for sustained crime reduction; failed to think through the mech-
anisms by which the response could have a measurable impact; failed to
fully involve partners; narrowly focused responses, usually on offenders;
as well as a number of other weaknesses in the response development
process. Cordner (1998) observed that some problem-oriented policing
 Problem-oriented policing 141

projects gravitate toward faddish responses rather than implementing a

response that truly fits the nature of the problem.
As suggested above, the complexity of problems at crime hot spots
may discourage police officers from developing innovative responses. For
example, frustrated narcotics officers may choose to chase drug dealers at
a place rather than implementing a plan to change the multiple underly-
ing characteristics of a place that make it an attractive spot for illicit drug
sales (Green 1996; Eck and Wartell 1998; Taylor 1998). In the Jersey City
problem-oriented policing experiment, the numerous responses imple-
mented at the violent crime hot spots mirrored the complexity of the
problems they were intended to control (Braga 1997). Twenty-eight types
of responses were implemented across the places (Table 7.2). The num-
ber of responses per place ranged from one to twelve, with a mean of 6.7
responses per place. Traditional policing tactics were used at all places
and comprised between 14 percent (Place 7) and 100 percent (Place 4
and Place 10) of the responses per place. On average, almost a third
(31.1 percent) of the responses implemented per place involved tra-
ditional enforcement. Situational interventions intended to modify the
characteristics of a place were implemented at ten of twelve treatment
places. The situational strategies varied according to the nuances of the
problems at places (e.g., razing an abandoned building or the code inspec-
tion of a tavern). However, at all locations, a package of aggressive order
maintenance interventions was used to control the social disorder of the
places. These tactics included repeat foot and radio car patrols, dispers-
ing groups of loiterers, issuing summons for public drinking, “stop and
frisks” of suspicious persons, and so on. At nine places, drug enforce-
ment was used to disrupt drug sales at the place. The officers believed
that an increased presence and the harassment of illicit users of the
place would quell egregious social disorder, at least temporarily, until
a better plan could be developed and implemented. According to the
officers, aggressive order maintenance was a treatment that could affect
all illicit activity no matter what the variation: drug selling, loitering by
disorderly youth, homeless panhandlers, predatory robbers – all could
be affected by an increased presence in a bounded geographical area.
The Jersey City officers much preferred this strategy to the additional
work necessary to implement alternative situational responses (Braga
1997).

Assessment
The crucial last step in the practice of problem-oriented policing is
to assess the impact the intervention has had on the problem it was
142 Anthony A. Braga and David Weisburd

Table 7.2 Responses to problems at violent crime places

Place Responses

1 Aggressive order maintenance to disperse loitering homeless and disorderly youths

Drug enforcement to disrupt street-level drug sales around train and bus terminals
Robbery investigation to apprehend robbery crew
Public works removed trash at place to discourage illegal dumping
Hung “No dumping – Police take notice” sign to discourage illegal dumping
Fixed holes in fence to secure access to the bus and train terminal
Boarded and fenced abandoned buildings to prevent drug activity and squatting
Helped homeless find shelter and substance abuse treatment
Dispensed crime prevention literature to commuters
Removed piles of lumber to discourage loitering
2 Aggressive order maintenance to disperse groups of disorderly youths
Drug enforcement to disrupt street-level drug sales on street corner and from bodega
Required store owners to clean trash from store fronts
Erected fences to secure vacant lot from illicit drug activity
Cleaned vacant lot
Boarded and fenced abandoned buildings to prevent drug activity
Removed drug selling crew’s stashed guns
Housing code enforcement to repair dilapidated building
Eviction of drug seller from low-income apartment at place
Surveillance of drug offenders using videotapes
3 Aggressive order maintenance to disperse loiterers from vacant lot and apartment
building area
Drug enforcement to disrupt street-level drug sales from vacant lot, alley, and
apartment building
Robbery investigation to apprehend robbery crew
Cut weeds and removed trash from vacant lot
Erected fence around vacant lot to prevent drug sales and robberies
Public works removed trash and drug paraphernalia from alley
Public works trimmed trees to increase lighting around vacant lot, alley, and
apartment building
Dispensed crime prevention literature to students and campus security
Improved apartment building security by adding new door locks; prevented access to
drug dealers
4 Aggressive order maintenance to disperse vagrants and groups of disorderly youths
Directed patrol before and after school hours
Robbery investigation to apprehend robbery crew
5 Aggressive order maintenance to disperse groups of disorderly youths
Required store owners to clean store fronts
Changed style of trash cans to prevent loitering
6 Aggressive order maintenance to disperse loiterers
Drug enforcement to disrupt street-level drug sales in front of liquor store
Required store owners to clean store fronts
Opened and cleaned vacant lot to allow neighborhood kids to play basketball
Repaired street lights to increase lighting in area
Housing code enforcement to clean apartment building used as stash house
 Problem-oriented policing 143

Table 7.2 (cont.)

Place Responses

7 Aggressive order maintenance to disperse loitering groups of disorderly youths

Housing code enforcement to remove trash from within housing project
Evicted squatters and drug sellers from unoccupied apartments in low-income
housing project
Public works removed trash from around low-income housing project
Improved housing project security by adding new locks to doors; prevented access to
drug dealers
Improved lighting around housing project by installing floodlights and replacing
streetlights
Enforced parking ordinances around housing project to impound cars of drug sellers
and buyers
8 Aggressive order maintenance to disperse loiterers
Intensive drug enforcement to disrupt street-level sales and dismantle major drug
organization
Required store owners to clean store fronts
Code investigation of taverns
9 Aggressive order maintenance to disperse loiterers
Drug enforcement to disrupt street-level drug sales
Housing code enforcement to clean up low-income property
Evicted drug sellers from low-income apartment building
Boarded and fenced abandoned buildings to prevent drug activity
Razed abandoned buildings
Cleaned vacant lots
Public works removed trash from street
Erected fences around vacant lots to secure from drug activity
Surveillance of offenders at place using video tapes
Hung sign “No drug selling – Area under video surveillance by police”
10 Aggressive order maintenance to disperse groups of disorderly youths
11 Aggressive order maintenance to disperse loiterers
Drug enforcement to disrupt street-level drug sales
Surveillance of offenders at place using video tapes
Hung sign “No public drinking – Area under video surveillance by police”
Public works trimmed trees to increase lighting in area
Boarded and fenced abandoned buildings to prevent drug activity
Cleaned vacant lot
Erected fence around vacant lot to secure from drug activity
Removed graffiti from building
Mobilized residents to keep streets clean in front of their homes
Public works removed trash from street
Added trash receptacles to prevent litter
12 Aggressive order maintenance to disperse loiterers and groups of disorderly youth
Drug enforcement to disrupt drug sales at park and in and around bars
Code investigation of taverns; one bar shut down and other bar owners discontinue
sales of “take-out” alcohol
Required storeowners to clean storefronts
144 Anthony A. Braga and David Weisburd

supposed to solve. Assessment is important for at least two different

reasons. The first is to ensure that police remain accountable for their
performance and for their use of resources. Citizens and their represen-
tatives want to know how the money and freedom they surrendered to
the police are being used, and whether important results in the form of
less crime, enhanced security, or increased citizen satisfaction with the
police have been achieved. A second reason assessment is important is to
allow the police to learn about what methods are effective in dealing with
particular problems. Unless the police check to see whether their efforts
produced a result, it will be hard for them to improve their practices.
The degree of rigor applied to the assessment of problem-oriented ini-
tiatives will necessarily vary across the size and overall importance of the
problems addressed (Clarke 1998). Serious, large and recurrent prob-
lems such as controlling gang violence or handling domestic disputes
deserve highly rigorous examinations. Other problems that are less seri-
ous, or common, such as a lonely elderly person making repeat calls to the
police for companionship, are obviously not worth such close examina-
tions. Unfortunately, Scott and Clarke (2000) observe that assessment of
responses is rare and, when undertaken, it is usually cursory and limited
to anecdotal or impressionistic data.
Clarke’s (1998) review of submissions for the Herman Goldstein award
provides a good summary of the common limitations of assessments
performed by police departments engaged in problem-oriented policing
projects (319–320):
r Police report reductions in calls for service or arrests without relating the
results to specific actions taken. In other words, the police frequently
fail to examine whether variations in “dosage levels” of the response
correlate with variations in relevant crime statistics.
r Police consider assessment only as an afterthought, rather than building
it into the original outline of the project.
r Police fail to present any control data. For example, if some action has
been taken in one location that appears to have produced a decrease in
the number of calls for service, the police do not generally document
what, if anything, happened in a similar nearby location where no action
was taken.
r On the rare occasion when control data are presented, police fail to
ensure that the control is adequate. For example, the control location
may be so different from the original site that it is impossible to conclude
anything from the comparison.
r Police fail to study displacement. Many agencies do not realize that
their response may simply have pushed the problem elsewhere.
 Problem-oriented policing 145

Is “shallow” problem solving enough?

Despite the gap between the desired application of the approach and
its actual implementation, there is considerable evidence that problem-
oriented policing is effective in preventing crime (see e.g., Braga 2002;
Committee to Review Research 2004; Weisburd and Eck 2004). This
suggests that problem-oriented policing interventions may not need to
be implemented in the ways envisioned by Herman Goldstein in order
to produce a crime prevention effect. Perhaps, simply focusing police
resources on identifiable risks that come to the attention of problem-
oriented policing projects, such as high-activity offenders, repeat victims,
and crime hot spots, may be enough to produce crime control gains.
This is a striking result considering the large body of research that shows
the ineffectiveness of many police crime prevention efforts (see Visher
and Weisburd 1998 for a review). Therefore, we think that it is criti-
cally important to refine and extend current definitions of police crime
prevention efforts.
Given the long police tradition of responding to high-crime locations
and the current popularity of hot spots policing, our concluding discus-
sion focuses on police crime efforts at high-crime places. These con-
cepts could also be applied to offender- and victim-focused responses.
Figure 7.1 presents a continuum of strategies, ranging from traditional
to innovative, that police can use to control crime at high-crime places. At
one extreme, police departments use traditional, incident-driven strate-
gies to control crime in the community. Although these activities coinci-
dentally cluster in space and time, these ad hoc enforcement strategies
are not specifically targeted at problem places and the limitations of
this approach are, by now, well known. Based on Eck’s (1993) examina-
tion of alternative futures for problem-oriented policing, problem-solving
efforts can be divided into “enforcement” and “situational” problem-
oriented policing programs. “Enforcement” problem-oriented policing
interventions have moved the police response forward by focusing mostly
traditional tactics at high-risk times and locations. Although these pro-
grams are “problem oriented” in a global way, their tactics do not employ
the individualized treatments of crime problems advocated by Herman
Goldstein (1990). These interventions have included optimizing patrol
time in hot spots (Koper 1995; Sherman and Weisburd 1995), conducting
aggressive Terry 1 searches for guns in areas that are known for high levels
of firearms violence (Sherman, Shaw, and Rogan 1995), and implement-
ing systematic crackdowns in street-level drug markets (Weisburd and
Green 1995).
146 Anthony A. Braga and David Weisburd

Traditional Innovative

Incident-Driven Enforcement POP Situational POP

• Random Patrol • Directed Patrol • Understanding causal
mechanisms
• Investigation and Arrest • Crackdowns • Implementing alternative
responses
• Rapid Response • Terry Searches • Collaborating with others

Figure 7.1 Continuum of police strategies to control high-crime places

These enforcement problem-oriented policing interventions focus

mainly on the time and location of crime events, rather than focusing
on the characteristics and dynamics of a place that make it a hot spot
for criminal activity. Although these interventions have produced crime
control gains and have added to law enforcement’s array of crime pre-
vention tools, it is commonly assumed that police could be more effective
if they focused their efforts on the criminogenic attributes that cause a
place to be “hot.” In other words, adding an increased level of guardian-
ship at a place by optimizing patrols is a step in controlling crime, but
reducing criminal opportunities by changing site features, facilities, and
the management at a place (e.g., adding streetlights, razing abandoned
buildings, and mobilizing residents) may have a more profound effect on
crime. At the innovative end of the continuum is Goldstein’s (1990) vision
of “situational” problem-oriented policing: members of police agencies
undertake a thorough analysis of crime problems at places, collaborate
with community members and other city agencies, and conduct a broad
search for situational responses to problems.
Similar to other studies, translating problem-oriented policing theory
into practice was difficult for Jersey City problem-oriented police offi-
cers. On the continuum of police strategies to control high-activity crime
places, the program as implemented would fit between enforcement
and situational problem-oriented interventions (see Figure 7.1). Despite
falling short of Goldstein’s vision, the Jersey City problem-oriented polic-
ing program was still found to be effective in preventing crime (Braga et al.
1999). Some scholars would describe the efforts of the Jersey City
police officers as “problem solving” rather than problem-oriented polic-
ing (Cordner 1998). Problem solving is a “shallow” version of problem-
oriented policing, as these efforts are smaller in scope, involve rudimen-
tary analysis, and lack formal assessments. According to Cordner (1998),
 Problem-oriented policing 147

problem solving better describes what an officer does to handle a par-

ticular dispute or drug house rather than initiatives to deal with prob-
lems of substantial magnitude like prostitution in a downtown area and
thefts from autos in parking lots. Enforcement problem-oriented initia-
tives would also fit what we would consider shallow problem solving.
When the complexities of crime problems at places and the difficulties
police officers experience in unraveling these problems and implementing
individualized responses are considered, shallow problem solving can be
a very practical way for police departments to prevent crimes at certain
places.

Should we expect police agencies to develop more

in-depth problem solving?
Problem-oriented policing advocates lament the current state of the art
in problem solving and offer a laundry list of suggestions to improve the
capacity of police agencies to engage in strategic problem solving (see
e.g., Scott 2000; Braga 2002; Boba 2003). Is it realistic to expect police
agencies to develop their capacity to conduct more in-depth problem-
oriented policing so their officers can go beyond the shallow problem
solving that is currently practiced in the field? We believe that it is unlikely
that police agencies will enhance their problem-solving efforts in the
foreseeable future. There seem to be, at least, two obstacles. First, the
hierarchical organizational structure of policing in itself tends to inhibit
innovation and creativity. In their close examination of Compstat, Willis,
Mastrofski, and Weisburd (2004) found that, while the program holds out
the promise of allowing police agencies to adopt innovative technologies
and problem-solving techniques, it actually hindered innovative problem
solving while strengthening the existing hierarchy through the added pres-
sure of increased internal accountability. Officers were reluctant to brain-
storm problem-solving approaches during Compstat meetings for fear of
undermining authority or the credibility of their colleagues. Moreover,
the danger of “looking bad” in front of superior officers discouraged mid-
dle managers from pursuing more creative crime strategies with a higher
risk of failure.
Second, the organizational culture of policing is resistant to this in-
depth approach to problem solving. The community policing move-
ment that emerged in the 1980s stressed greater police recognition
of the role of the community and emphasized “decentralization” and
“debureaucratization” to empower rank-and-file officers to make deci-
sions about how better to serve the neighborhoods to which they were
assigned (Skolnick and Bayley 1986; Mastrofski 1998). Weisburd and
148 Anthony A. Braga and David Weisburd

his colleagues (2003) suggest that the popularity and rapid diffusion of
Compstat programs across larger police departments in the United States
during the 1990s could be interpreted as an effort to maintain and rein-
force traditional police structures rather than an attempt to truly reform
American policing. The rapid rise of Compstat within police agencies did
not enhance their strategic problem-solving capacity at the beat level as
Compstat departments were found to be reluctant to relinquish power
that would decentralize some key elements of decisionmaking such as
allowing middle managers to determine beat boundaries and staffing lev-
els, enhancing operational flexibility, and risking going beyond the stan-
dard tool kit of police tactics and strategies (Weisburd et al. 2003). The
overall effect of the spread of Compstat, whether intended or not, was to
reinforce the traditional bureaucratic model of command and control.
More generally, the type of in-depth problem solving that Goldstein
and other problem-oriented policing advocates have proposed seems
unrealistic in the real world of policing, especially at the street level where
it is often envisioned. In the real world of police organizations, commu-
nity and problem-oriented police officers rarely have the latitude neces-
sary to assess and respond to problems creatively. As William Bratton
(1998: 199) observed in his assessment of community policing in New
York City at the beginning of his tenure, “street-level police officers were
never going to be empowered to follow through.” Compstat was offered
as a solution to this problem, but the ability of the rank-and-file officer to
make decisions changed little as middle managers were held accountable
for achieving organizational goals of successful crime control (Willis et al.
2004). Of course, decentralizing decisionmaking power to line-level offi-
cers does not guarantee effective problem solving. In their assessment of
problem solving in Chicago, a police department well known for its efforts
to decentralize its organization to facilitate community policing, Skogan
and his colleagues (1999: 231) gave a failing grade ranging from “strug-
gling” to “woeful” for problem-solving efforts in 40 percent of the beats
they studied. Inadequate management, poor leadership and vision, lack
of training, and weak performance measures were among the operational
problems identified as affecting problem solving in the poorly performing
beats (Skogan et al. 1999).

Conclusion
Problem-oriented policing represents an important innovation in
American policing. Unfortunately, the practice of problem-oriented
policing too often falls short of the principles suggested by Herman
Goldstein (1990). Given the extensive literature on the gap between
 Problem-oriented policing 149

principles and practice, it seems unrealistic to expect line-level officers to

conduct in-depth problem-oriented policing as their routine way of deal-
ing with crime problems. Successful applications of problem-oriented
policing usually involve larger-scale problems, the involvement of aca-
demic researchers and crime analysis units, and the solid support of the
police command staff to implement alternative responses. The Boston
Police Department’s Operation Ceasefire intervention to prevent gang
violence (Braga et al. 2001), the Newport News Police Department’s
efforts to reduce thefts from cars parked in shipyard parking lots (Eck
and Spelman 1987), and the Charlotte Mecklenburg Police Department’s
program to reduce theft from construction sites (Clarke and Goldstein
2002) involved resources that were far beyond the reach of the rank-and-
file police officer.
Research evidence suggests that shallow problem-solving efforts, rang-
ing from focused enforcement efforts in high-risk places and at high-risk
times to problem solving with weak analyses, mostly traditional responses,
and limited assessments, are enough to prevent crime. Perhaps it is time
to stop trying to achieve the ideal of strategic problem-oriented policing at
the street level and embrace the reality of what has been shown to lead to
real crime prevention value – ad hoc shallow problem-solving efforts that
focus police on high-risk places, situations, and individuals. We suggest
that line-level officers should be encouraged to problem solve at the beat
level and more sophisticated problem-oriented policing projects should
be engaged higher up in the organization with the support of a crime
analysis unit and in collaboration with other criminal justice agencies,
academic researchers, and community-based groups. However, it is time
for police practitioners and policymakers to set aside the fantasy of street-
level problem-oriented policing and embrace the reality of what they can
expect from the beat officer in the development of crime prevention plans
at the street level.

  
In Terry v. Ohio (1968), the Supreme Court upheld police officers’ right to
conduct brief threshold inquiries of suspicious persons when they have reason
to believe that such persons may be armed and dangerous to the police and
others. In practice, this threshold inquiry typically involves a safety frisk of the
suspicious person.


Boba, R. (2003). Problem analysis in policing. Washington, DC: Police Foundation.
Braga, A. A. (1997). Solving violent crime problems: An evaluation of the Jersey
City Police Department’s pilot program to control violent places. Doctoral
150 Anthony A. Braga and David Weisburd

Dissertation, Rutgers University. Ann Arbor, MI: University Microfilm

International.
(2002). Problem-oriented policing and crime prevention. Monsey, NY: Criminal
Justice Press.
Braga, A. A., Kennedy, D., Waring, E., and Piehl, A. (2001). Problem-oriented
policing, deterrence, and youth violence: An evaluation of Boston’s Opera-
tion Ceasefire. Journal of Research in Crime and Delinquency, 38, 195–225.
Braga, A. A., Weisburd, D., Waring, E., Mazerolle, L. G., Spelman, W., and
Gajewski, F. (1999). Problem-oriented policing in violent crime places: A
randomized controlled experiment. Criminology, 37, 541–580.
Bratton, W. (1998). Turnaround: How America’s top cop reversed the crime epidemic.
New York: Random House.
Buerger, M. (1994). The problems of problem-solving: Resistance, interdepen-
dencies, and conflicting interests. American Journal of Police, 13, 1–36.
Bynum, T. (2001). Using analysis for problem-solving: A guidebook for law enforce-
ment. Washington, DC: Office of Community Oriented Policing Services,
US Department of Justice.
Capowich, G. and Roehl, J. (1994). Problem-oriented policing: Actions and effec-
tiveness in San Diego. In D. Rosenbaum (ed.), The challenge of community
policing: Testing the promises. Thousand Oaks, CA: Sage Publications.
Capowich, G., Roehl, J., and Andrews, C. (1995). Evaluating problem-oriented
policing outcomes in Tulsa and San Diego. Final Report to the National Institute
of Justice. Alexandria, VA: Institute for Social Analysis.
Clarke, R. (ed.) (1997). Situational crime prevention: Successful case studies. 2nd
ed. Albany, NY: Harrow and Heston.
(1998). Defining police strategies: Problem solving, problem-oriented polic-
ing and community-oriented policing. In T. O’Connor Shelley and A. C.
Grant (eds.), Problem-oriented policing: Crime-specific problems, critical issues,
and making POP work. Washington, DC: Police Executive Research Forum.
Clarke, R. and Goldstein, H. (2002). Reducing theft at construction sites: Lessons
from a problem-oriented project. In N. Tilley (ed.), Analysis for crime preven-
tion. New York: Criminal Justice Press.
Committee to Review Research on Police Policy and Practices. (2004). Fair-
ness and effectiveness in policing: The evidence. Washington, DC: National
Academies Press.
Cordner, G. (1994). Foot patrol without community policing: Law and order in
public housing. In D. Rosenbaum (ed.), The challenge of community policing:
Testing the promises. Thousand Oaks, CA: Sage Publications.
(1998). Problem-oriented policing vs. zero tolerance. In T. O’Connor Shel-
ley and A. C. Grant (eds.), Problem-oriented policing: Crime-specific problems,
critical issues, and making POP work. Washington, DC: Police Executive
Research Forum.
Eck, J. E. (1993). Alternative futures for policing. In D. Weisburd and C. Uchida
(eds.), Police innovation and control of the police: Problems of law, order, and
community. New York: Springer-Verlag.
(2000). Problem-oriented policing and its problems: The means over ends
syndrome strikes back and the return of the problem-solver. Unpublished
manuscript. Cincinnati, OH: University of Cincinnati.
 Problem-oriented policing 151

Eck, J. E. and Spelman, W. (1987). Problem-solving: Problem-oriented policing in

Newport News. Washington, DC: Police Executive Research Forum.
Eck, J. E. and Wartell, J. (1998). Improving the management of rental properties
with drug problems: A randomized experiment. In L. G. Mazerolle and J.
Roehl (eds.), Civil Remedies and Crime Prevention, 9 (pp. 161–183). Monsey,
NY: Criminal Justice Press.
Goldstein, H. (1979). Improving policing: A problem-oriented approach. Crime
and Delinquency, 25, 236–258.
(1990). Problem-oriented policing. Philadelphia, PA: Temple University Press.
Goldstein, H. and Susmilch, C. (1982). The drinking-driver in Madison: A study
of the problem and the community’s response. Madison, WI: University of
Wisconsin Law School.
Green, L. (1996). Policing places with drug problems. Thousand Oaks, CA: Sage
Publications.
Homel, R., Hauritz, M., Wortley, R., McIlwain, G., and Carvolth, R. (1997). Pre-
venting alcohol-related crime through community action: The surfers par-
adise safety action project. In R. Homel (ed.), Policing for prevention: Reducing
crime, public intoxication, and injury. New York: Criminal Justice Press.
Hunter, R. and Jeffrey, C. R. (1992). Preventing convenience store rob-
bery through environmental design. In R. Clarke (ed.), Situational crime
prevention: Successful case studies. Albany, NY: Harrow and Heston.
Koper, C. (1995). Just enough police presence: Reducing crime and disorderly
behavior by optimizing patrol time in crime hot spots. Justice Quarterly, 12,
649–672.
Mastrofski, S. (1998). Community policing and police organization structure. In
J. Brodeur (ed.), How to recognize good policing: Problems and issues. Thousand
Oaks, CA: Sage Publications.
Matthews, R. (1990). Developing more effective strategies for curbing prostitu-
tion. Security Journal, 1, 182–187.
Police Executive Research Forum. (2000). PSP National Evaluation Final Report.
Washington, DC: Police Executive Research Forum.
Read, T. and Tilley, N. (2000). Not rocket science? Problem-solving and crime reduc-
tion. Crime Reduction Series Paper 6. London: Policing and Crime Reduc-
tion Unit, Home Office.
Scott, M. (2000). Problem-oriented policing: Reflections on the first 20 years.
Washington, DC: US Department of Justice, Office of Community Oriented
Policing Services.
Scott, M. and Clarke, R.V. (2000). A review of submission for the Herman
Goldstein Excellence in Problem-Oriented Policing. In C. Sole Brito and
E. Gratto (eds.), Problem-oriented policing: Crime-specific problems, critical
issues, and making POP work, 3. Washington, DC: Police Executive Research
Forum.
Sherman, L. W., Gartin, P., and Buerger, M. (1989). Hot spots of predatory
crime: Routine activities and the criminology of place. Criminology, 27, 27–
56.
Sherman, L. W. and Weisburd, D. (1995). General deterrent effects of police
patrol in crime hot spots: A randomized controlled trial. Justice Quarterly,
12, 625–648.
152 Anthony A. Braga and David Weisburd

Sherman, L. W., Shaw, J., and Rogan, D. (1995). The Kansas City gun experiment.
Washington, DC: National Institute of Justice, US Department of Justice.
Skogan, W., Hartnett, S., DuBois, J., Comey, J., Kaiser, M., and Lovig, J. (1999).
On the beat: Police and community problem solving. Boulder, CO: Westview
Press.
Skolnick, J. and Bayley, D. (1986). The new blue line: Police innovations in six
American cities. New York: Free Press.
Taylor, R. (1998). Crime and small-scale places: What we know, what we can prevent,
and what else we need to know. In Crime and Place: Plenary Papers of the 1997
Conference on Criminal Justice Research and Evaluation. Washington, DC:
National Institute of Justice, US Department of Justice.
Visher, C. and Weisburd, D. (1998). Identifying what works: Recent trends in
crime prevention strategies. Crime, Law and Social Change, 28, 223–242.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder,
and fear? Annals of the American Academy of Political and Social Science, 593,
42–65.
Weisburd, D. and Green, L. (1995). Policing drug hot spots: The Jersey City
DMA Experiment. Justice Quarterly, 12, 711–736.
Weisburd, D. and Mazerolle, L. (2000). Crime and disorder in drug hot spots:
Implications for Theory and Practice in Policing. Police Quarterly, 3 (3),
331–349.
Weisburd, D. and McEwen, J. T. (eds.). (1997). Crime mapping and crime preven-
tion. Monsey, NY: Criminal Justice Press.
Weisburd, D., Maher, L., and Sherman, L. W. (1992). Contrasting crime gen-
eral and crime specific theory: The case of hot spots of crime. Advances in
Criminological Theory, 4, 45–69.
Weisburd, D., Mastrofski, S. D., and Greenspan, R. (2001). Compstat and Orga-
nizational Change. Washington, DC: Police Foundation.
Weisburd, D., Mastrofski, S. D., McNally, A. M., Greenspan, R., and Willis, J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2 (3), 421–456.
Willis, J., Mastrofski, S. D., and Weisburd, D. (2004). Compstat and bureaucracy:
A case study of challenges and opportunities for change. Justice Quarterly,
21 (3), 463–496.
Part IV

Pulling levers policing

8 Advocate
Old wine in new bottles: policing and the
lessons of pulling levers

David M. Kennedy

Introduction
Pity the poor soul who, ten years ago, would have predicted that hard-core
street cops would be making a practice of sitting down with serious violent
offenders and telling them politely to cease and desist. Yet, increasingly,
that is what cops are doing: and, more to the point, doing successfully.
The “pulling levers” strategies piloted in Boston in the mid-1990s and
implemented since then in a range of other jurisdictions are racking up
impressive results in preventing violent crimes. From one vantage, these
strategies are innovative, sometimes bordering on the bizarre, as in the
face-to-face meetings between authorities and offenders that many rely
upon. From another, they are really very traditional, relying on old and
simple ideas about offenders and their motivations, and about authori-
ties and their powers, and how the latter can influence the former. Old
and new meet in the design of the strategies, where ivory-tower research
methods rely on the street savvy of front-line cops; in images of offend-
ers, where very traditional and concrete ideas about serious crime meet
new and abstract notions; in the partnerships that carry out the work,
where traditional actors do some very traditional things in significantly
new ways; and in the underlying logic of the strategies, which seeks to do
something entirely old-fashioned – deter crime – in a strikingly original
way.
Pulling levers strategies are one fruit of the problem-oriented polic-
ing movement. The approach first emerged as part of the Boston Gun
Project, a problem-oriented policing project aimed at “gang” violence in
Boston, and has since spread to other jurisdictions and settings (Kennedy
1997; 1998; Kennedy and Braga 1998; Braga, Kennedy, Waring, and
Piehl 2001; Braga, Kennedy, and Tita 2002; Kennedy 2002a; Dalton
2003; McGarrell and Chermak 2003; Tita, Riley, Ridgeway et al. 2003;
Wakeling 2003). It was, in that sense, straight from the problem-oriented
copybook: it focused on a relatively narrow problem, utilized a wide vari-
ety of information sources and analytic techniques, relied heavily on the

155
156 David M. Kennedy

experience and insight of front-line law enforcement and community

actors, and sought a strategic intervention that would have a substan-
tial impact. It is also one of a growing number of policing interventions
that combines a set of relatively traditional tactics with a new strategic
focus, and adds to that mix new operational tools, resulting in a strategy
that looks both very traditional and quite new and different (Weisburd
and Eck 2004).
The best news about these strategies is that they seem to have worked
in a number of jurisdictions against serious crime problems that have tra-
ditionally been regarded as intractable, and even in principle impossible
for criminal justice actors to address. But beyond that, they suggest a
bridge between traditional and powerful – but often, in practice, unhap-
pily limited – ideas and actions, and variations on those ideas and actions
that can lead to potent, practical applications. This chapter looks at the
experience to date with pulling levers strategies. It is not a review of the
operational idea, its intellectual roots, particular interventions, or the for-
mal evaluation of impacts; those things have been done elsewhere. Rather,
it treats another set of concerns: what these strategies seem to have meant
to police departments and other criminal justice agencies, and why, and
what that experience suggests about next steps.

What it is
Pulling levers or focused deterrence strategies deploy enforcement, ser-
vices, the moral voice of communities, and deliberate communication in
order to create a powerful deterrent to particular behavior by particular
offenders. First deployed in Boston in the mid-1990s to prevent youth
violence, pulling levers operations have tended to follow a basic frame-
work that includes:
r Selection of a particular crime problem, such as youth homicide or
street drug dealing.
r Pulling together an interagency enforcement group, typically including
police, probation, parole, state and federal prosecutors, and sometimes
federal enforcement agencies.
r Conducting research, usually relying heavily on the field experience
of front-line police officers, to identify key offenders – and frequently
groups of offenders, such as street gangs, drug crews, and the like – and
the context of their behavior.
r Framing a special enforcement operation directed at those offenders
and groups of offenders, and designed to substantially influence that
context, for example by using any and all legal tools (or levers) to
 Old wine in new bottles 157

sanction groups such as crack crews whose members commit serious

violence.
r Matching those enforcement operations with parallel efforts to direct
services and the moral voices of affected communities to those same
offenders and groups.
r Communicating directly and repeatedly with offenders and groups to
let them know that they are under particular scrutiny, what acts (such as
shootings) will get special attention, when that has in fact happened to
particular offenders and groups, and what they can do to avoid enforce-
ment action. One form of this communication is the “forum,” “notifi-
cation,” or “call-in,” in which offenders are invited or directed (usually
because they are on probation or parole) to attend face-to-face meet-
ings with law enforcement officials, service providers, and community
figures.
In Boston, for example, probation officers pulled members of street
drug groups into meetings with authorities, service providers, and com-
munity figures where they were told that violence had led to comprehen-
sive enforcement actions – such as federal drug investigations – against
several violent groups, that violence by their groups would provoke the
same enforcement actions, that services were available to those who
wished them, that the affected communities desperately wanted the vio-
lence to stop, and that groups that did not act violently would not get such
unusual, high-level enforcement attention. Similar efforts in other juris-
dictions followed. Pulling levers has been a central theme in the Justice
Department’s Strategic Approaches to Community Safety Initiative and
Project Safe Neighborhoods initiative, and the basic ideas are increas-
ingly showing up in local operations (Dalton 2003). In general, “pulling
levers” operations seem to be well received by the affected communities,
apparently due at least in part to the usually wide array of community
actors involved and to the clear notice and options provided to offenders
(Winship and Berrien 1999).

Evidence of impact
There is good reason to believe that such efforts can work. These inter-
ventions do not lend themselves to the kind of high-level random assign-
ment experimental designs that would give the strongest evaluations of
their impact. Short of that, the available evaluations and a growing body
of site experience suggest strong effects. The best statistical evaluations
come from Boston and Indianapolis and show strikingly parallel results:
citywide reductions in homicide of around 50 percent, with larger effects
158 David M. Kennedy

on gun homicides of younger minority males at which both interventions

were primarily directed (Braga et al. 2001; McGarrell and Chermak
2003). In each case, the reductions were very sudden and sharp: what
Professor Edmund McGarrell, now of Michigan State University, who
helped organize the Indianapolis project, calls the “light switch” effect.
The same pattern has been seen in Minneapolis (Kennedy and Braga
1998); Stockton, California (Wakeling 2003); High Point and Winston-
Salem, North Carolina (Dalton 2003); Portland, Oregon (Dalton 2003);
and Rochester, New York.1 A weak, because only partially implemented,
version even appears to have been somewhat effective in East Los Angeles
(Tita et al. 2003). The range of cities now suggests that the basic model
is effective against violence by more informal, neighborhood-based drug
groups, as in Boston; by more structured, Chicago-style “gangs,” as in
Minneapolis and Indianapolis; and by West coast gangs, as in Stockton
and Los Angeles. It has been effective in jurisdictions in which violence
was more or less stable, as in Boston, and in which it was escalating, as
in Minneapolis and Indianapolis. Not all of the news is good. Several
jurisdictions have failed to sustain what were apparently successful inter-
ventions. Serious efforts in Baltimore and San Francisco did not reach
the point of full implementation (about which more later). There does
seem good reason to believe, however, that these strategies hold promise
of real effectiveness. Taken separately, as is often – perhaps always – the
case, there are plausible doubts about what might have happened in each
site. Taken together, the experience to date suggests the strategy has both
power and legs.
There is less experience with pulling levers interventions aimed at other
crime problems. The author is involved in pulling levers drug market
interventions in High Point and in Rochester, New York, both aimed at
eliminating particularly troublesome public forms of drug dealing such
as street markets and crack houses by warning dealers, buyers, and their
families that focused enforcement is imminent (Kennedy 1998). There
is as yet no experience with such interventions, but it is possible to revisit
provocative and successful earlier drug market operations and view them
in that light. The Houston Police Department’s Link Valley operation,
for example, used the deliberate announcement of a massive interagency
sweep to dry up a large open-air powder cocaine market without mak-
ing a single arrest (Kennedy 1990). The San Diego Police Department
eliminated a substantial crack market by reaching out to both dealers and
buyers and informing them of forthcoming special enforcement efforts
(San Diego Police Department 1998). The author has framed pulling
levers approaches to domestic violence (Kennedy 2002b). While there
is no evidence that these strategies would be successful on a large scale,
 Old wine in new bottles 159

they do pass the law enforcement “laugh test”: police, prosecutors, and
other enforcement personnel tend to think that they’re worth exploring.
Pulling levers operations, then, are slowly being accepted in enforcement
circles as useful additions both to the traditional operational repertoire
and to newer frameworks such as “broken windows” policing.

What cops know

As innovative, odd, and apparently self-defeating as these strategies may
be – just try telling a grizzled street cop that the way to stop violent crime
is to tell hard-core offenders exactly what your enforcement plans are –
their conditional appeal to practitioners makes a certain amount of sense.
These are interventions that are based on what cops (and many others in
law enforcement) know to be true. The whole of the strategy is certainly
not familiar, and frequently meets with deep skepticism: the idea that
offenders will pay any attention to a “stop it” message, for example – no
matter how direct, credible, and backed up by action – often provokes
stark disbelief. At the same time, though, much of what goes into these
operations, and what sets them apart from other operations, is knowledge
and outlook deeply ingrained in enforcement circles. Those sensibilities
both shape the strategies and shape the law enforcement response to
adopting them.
Police officers (and prosecutors, probation officers, and the like) know,
for example, that for many crime problems, repeat offenders and repeat
victims are at the center of the action. They know that much of that
action is in particular neighborhoods, and particular places within those
neighborhoods. They know that different kinds of groups, and various
kinds of problems within and between those groups, drive a lot of what
happens. They know that these groups and their members commit vast
numbers of crimes, probation and parole violations, and other offenses,
from failing to show up in court to not paying fines and child support
to drinking and drugging to driving illegally. They know this not from
reading criminology and social science – though all these threads are to
be found there (Kennedy 2003) – but from their own experience. They
know that if those groups and individual offenders were controlled, it
would make a big difference. Experienced practitioners know that our
usual attempts to shape offenders’ behavior, whether through traditional
enforcement or through services, counseling, social work, or such means,
are hugely ineffective: most experienced law enforcement personnel hold
out little or no hope that serious offenders will be influenced by any-
thing short of a cop standing at arm’s length. That frequently, perhaps
usually, makes them despair of doing anything effective. But they want
160 David M. Kennedy

to, and their inclination is to do so through enforcement: police officers

and their kin in other agencies believe in good guys and bad guys, and
they believe in the exercise of authority, and they want to draw lines in the
sand and enforce them. “Pulling levers” interventions speak to this orien-
tation. It suits cops to say, Stop it! They need to be persuaded that it can
work, but this mindset at least opens the door. It sets these interventions,
and other largely authority-based interventions such as broken windows
strategies, apart from other innovative problem-solving frameworks such
as, for example, crime prevention through environmental design.
Operationally, the wealth of very particular street knowledge at the
front lines of these agencies is an invaluable resource. Police officers, in
particular – but also frequently probation and parole officers, line pros-
ecutors, the sheriff’s officers who manage jail populations, and others –
frequently have extraordinary insight into who is doing what, and why,
and very often what will happen next. What from even a small distance
looks chaotic, senseless, and unpredictable is often known to these prac-
titioners as comprehensible and in its own way coherent. They know
who is in what street groups, and who is fighting with whom; what last
year’s antecedent to yesterday’s shooting was; who is committing the drug
robberies that are not even being reported to the police; who is selling
drugs on the corners; what mid-level dealer is running the crack houses
operated only by juveniles; what turf is claimed by which groups, and
who is allowed there and who is not; which domestic violence offenders
are currently most dangerous to what women. Police departments and
other enforcement agencies generally make little or no use of this front-
line knowledge, partly because it is often of no use in making cases – an
unreported drug robbery, to take a particularly clear example, cannot be
prosecuted – and partly because of the top-down management typical
of police agencies. As a result, what is common knowledge at the street
level can be utterly obscure just a few steps away. In Washington, DC,
for example, even some career homicide prosecutors argued vehemently
that that city’s extraordinary homicide rate was the product of solo gun-
slingers bumping into each other on the streets; there was, they believed,
no more sense to it than that, and no logic or context beyond chaotic
social breakdown. That was what they saw in the formal fact patterns they
reviewed and presented in prosecutions. Police officers were incredulous,
and for good reason; a review of homicide incidents based on front-line
knowledge showed clearly that homicide was rooted in very patterned
behavior by a fairly small number of very active and well-known, mostly
neighborhood-based drug crews (Travis, Kennedy, Roman et al. 2003).
But until what they knew was deliberately gathered and recorded, it had
no formal existence and could not shape policy or operations.
 Old wine in new bottles 161

Qualitative research can draw this information out. These methods –

incident reviews, gang and group mapping exercises, systems to gather
front-line knowledge about “hidden” crimes like drug robberies and
domestic violence – have become central both to unpacking particu-
lar crime problems and to implementing pulling levers interventions
(Kennedy, Braga, and Piehl 1997; Kennedy and Braga 1998; Dalton
2003; McGarrell and Chermak 2003; Wakeling 2003). The information
they produce is vital, and frequently astonishes both agency executives
and outsiders. In San Francisco, for example, a surge in homicide that
shocked the city and led to, among other things, a high-profile public
forum by the mayor, was the product, by front-line consensus, of perhaps
a hundred or fewer particularly dangerous, influential – and quite well-
known – offenders. The process of gathering and utilizing such knowledge
honors the front lines and greatly enhances the salience of subsequent
operations.

Reclaiming deterrence: futility and displacement

Pulling levers strategies are deterrence strategies. Most criminal justice
practitioners, most of the time, do not believe in deterrence. That the
strategies work, and beyond that work with offenders and crime problems
that have demonstrably been particularly resistant to ordinary enforce-
ment, is helping reclaim the idea and the practice of deterrence.
The failure of deterrence is everywhere in law enforcement. Practi-
tioners certainly believe that it is operating at some level – if police shut
down operations, crime would certainly go up – but they do not much
believe that it is operating at the margin. They do not believe that offend-
ers, particularly serious offenders, much care about official consequences
for their crimes or that those consequences can be increased such that
offenders will care. The evidence for this is manifest. Many offenders
are arrested and sanctioned repeatedly, and continue to offend. Many
commit crimes such as drug dealing in full public view. Many routinely
violate the terms and conditions of probation and parole. And many, by
their actions and words, show their apparent disregard for extreme and
predictable consequences of their behavior, such as being hurt or killed
on the street. When offenders don’t mind dying, authorities frequently
say, what can we possibly do that they will care about?
This does not undercut the faith police and other authorities have in
enforcement. But it severely limits the ways in which they think enforce-
ment can be effective. It may work to swamp an area with officers; as long
as that persists, offenders will keep their heads down. It will certainly work
to take an offender off the streets; while he’s inside, he won’t commit any
162 David M. Kennedy

crimes outside. But authorities tend to have little or no faith in their

ability to persuade offenders to change their behavior, or in fact to do
anything at all; one of the predictable hurdles in implementing “pulling
levers” operations is police officers’ conviction that probationers will not
even bother to show up when directed to appear at call-ins. They violate
their probation all the time, and nothing ever happens, officers say; why
should this be any different? (In fact, compliance at call-ins is routinely
very high, and in practice seeing probationers appear often opens the first
chink in line officers’ skepticism about the larger strategy.)
Closely linked is an absolute, nearly theological belief in displacement –
that anything short of incarceration will simply move offenders around.
The research on this point is by now pretty persuasive; displacement is far
less universal than has long been thought, is virtually never complete, and
in fact enforcement efforts frequently produce a “diffusion of benefits,”
the opposite of displacement (Clarke and Weisburd 1994; Braga 2002).
Police and other enforcement practitioners remain unmoved. This is not
what they see, and they do not believe it. One of the most important
results of the manifest impact of pulling levers strategies, then, is that
the basic idea of deterrence – a big and important idea – is to some
extent being rehabilitated. The facts are quite clear, in jurisdiction after
jurisdiction, the worst of offenders – gang members, violent offenders,
shooters – are substantially changing their behavior without being taken
off the street. It is but a short step to the provocative thought that it is
not that offenders cannot be deterred, but that what we normally do is
simply not deterring them, and that quite possibly we can do better.

The salience of groups and networks

Most pulling levers interventions have focused on various kinds of groups
and networks: gangs, drug crews, and the like. This is not necessarily
an essential part of the framework. Individual domestic violence offend-
ers, for example, could be identified as being particularly dangerous;
sanctioned using any available legal means (domestic violence charges,
charges on other sorts of legal exposure, special probation and parole
regimes); and that special focus marketed to other domestic violence
offenders in order to shape their behavior (Kennedy 2002b). But in
practice, pulling levers strategies have so far focused primarily on homi-
cide and serious violence in urban settings, and in practice those prob-
lems are overwhelmingly concentrated in groups and networks of chronic
offenders (Kennedy 1997; 1998; Kennedy and Braga 1998; Braga et al.
2001; Braga et al. 2002; Kennedy 2002a; Dalton 2003; McGarrell and
Chermak 2003; Tita et al. 2003; Wakeling 2003). Various sorts of groups
 Old wine in new bottles 163

of offenders representing a tiny fraction of the population have repeat-

edly been shown to be responsible for most homicides and large shares of
other violent crime (Kennedy 2003). The operational strategies have sim-
ply followed that presentation of the problem. Once again, the salience
of groups and networks is a commonplace in the literature, and also a
commonplace on the front lines. It has not, however, been a common-
place in official accounts of these problems or, frequently, in operational
responses to them. Elevating their recognition and the attention paid to
them is an important development.
There are a number of reasons groups and networks have gotten far
less attention than they should have in law enforcement practice. Police
and other authorities recognize the existence of gangs, but outside of a
few jurisdictions most of these groups and networks are not what are
generally thought of as gangs (Kennedy 2002a). Legal avenues rarely
allow a group or network to be held legally responsible for a criminal
act; a loose-knit drug group cannot in practice be charged collectively
for the shooting committed by one of its members. Homicide investi-
gators and prosecutors responsible for uncovering and presenting the
fact pattern will themselves frequently not know about the rest of the
group or the larger context of the shooting. Those who know the most
about the group and its behavior, such as conflicts with other groups,
will frequently be narcotics and precinct officers and have no role in the
investigation, or in the gathering and interpreting of departmental intel-
ligence. What results is a kind of open secret in which the existence of
offender groups and networks, and the significance of their behavior, is
both very well known to some practitioners and neither recognized nor
acted upon elsewhere. In New Haven, for example, authorities concerned
about gun crime insisted that gangs were not an issue; the Latin Kings,
who had previously been a presence, had been eliminated through fed-
eral prosecutions. Work with front-line police officers, however, brought
to the surface what they already knew – that loose neighborhood-
based groups of chronic offenders were responsible for most of
the violence (Dalton 2003). Similar findings are invariable in other
jurisdictions.
Some of the implications are immediate and practical. These groups
can in fact be attended to as groups, even when something like a con-
spiracy case or a RICO prosecution cannot be brought; a hand-to-hand
drug buy from each member can put the whole group out of action. The
prospect of using the many such legal avenues is in fact the origin of the
pulling levers label. Once recognized, these groups can also be commu-
nicated with: they can be told, for example, that a shooting by one of
their number will result in action against all. Their relations with other
164 David M. Kennedy

groups can be identified and monitored, so that allies and enemies in a

simmering group-on-group vendetta get special attention.
A particularly interesting possibility – less immediate, but perhaps
no less practical – is recognizing that such groups and networks of
groups produce and enforce their own codes, rules, and expectations.
This, again, is classic sociology, criminology, and psychology, which have
always placed great emphasis on norms (Kennedy 2003). Police and other
authorities have rarely matched this recognition. A drive-by shooting will
typically be viewed as an expression of the shooter’s character and choice,
or, more remotely, as the product of more global forces such as neighbor-
hood, class, or racial and ethnic factors. Authorities rarely consider that
something intermediate might be operating, such as the imposition by the
shooter’s crew of norms that demand violence in certain circumstances,
and the real benefits and risks that go along with complying or failing to
comply. Nor do they consider the possibility of deliberately countering or
even changing those norms, such as by letting all group members know
that violence by one will impose costs on all, or by mobilizing respected
community voices to challenge them. In Lowell, Massachusetts, for exam-
ple, police responded to shootings by young Asian gang members, against
whom they had very little legal leverage, by cracking down on the gam-
bling parlors operated by older gang members and telling them explicitly
that the shooting had triggered the crackdowns. The police could not con-
trol the shooters, but the older members could and did (Braga, McDevitt,
and Pierce 2005). Recognizing groups and group processes opens up an
enormous range of analysis and action that operates in between, and may
be more productive than, the usual very particular attention to individuals
and very broad attention to social factors.

The strategic use of traditional enforcement

Law enforcement likes enforcing the law. For better or for worse, the
appeal of the traditional operational toolkit – patrol, rapid response, inves-
tigation, arrest, prosecution, jail, prison, and the like – remains strong. A
large body of academic and, beyond that, practical experience suggesting
that these tools are often very weak indeed has done little to change their
primacy (Braga 2002; Weisburd and Eck 2004). Their failure in prac-
tice is much more likely to be ascribed to system or social failure – we’re
making the arrests, the prosecutors are letting us down; what can we do
when nobody in the neighborhood is finishing school or working? – than
it is to lead to fundamental reconsideration of how business is or ought
to be done. Police, prosecutors, and others believe, even in the midst of
great frustration, in what they do.
 Old wine in new bottles 165

Pulling levers strategies appeal to practitioners partly because they

make a virtue of those beliefs and of existing agency capacities. These
tools, as deployed, are usually unsatisfactory. Deploying them in new ways
may be very different. Law enforcement personnel, and many others,
think that the logic and the actions of enforcement should matter. People
should seek to avoid being arrested, should avoid going to prison, should
not want probation officers in their bedrooms. The manifest failure of
existing tools to win compliance from offenders fuels the common con-
clusion that they are, on that evidence, irrational. But what if here, too,
there is a middle ground? It is true that chronic offenders are frequently
arrested and sanctioned. But it is also true – and, again, well understood
both from research and from practical experience – that most crimes
are not reported, most that are reported are not cleared, most that are
cleared are not met with stiff sanctions, sanctions such as probation are
generally not rigorously enforced, and so on (Kennedy 2003). It is thus
possible for there to be a great deal of enforcement without that enforce-
ment being consistent, predictable, or very meaningful to offenders. A
street drug dealer may have been arrested a number of times. But each
day, when he considers whether to go work his corner, the chance that
he will be arrested today will be both small and essentially random. This
can be true even for very serious crimes. In one of the most violent areas
of Washington, DC, nearly three-quarters of homicides go unsolved.2
Police and others very often know what happened, but they cannot put
cases together, and so there are no official consequences (Braga, Piehl,
and Kennedy 1999).
Law enforcement personnel understand very well – better than most
others, short of offenders – that this is in fact true. Their usual responses,
which have to do with resources, dramatically altering the behavior of
various criminal justice agencies and judges, enhancing the willingness
of the public to come forward, and the like, are hopeless, and they know
that as well. Pulling levers strategies offer a way out. It is in fact possible
to warn each of sixty street gangs in a jurisdiction that the first to commit
a homicide will get overwhelming attention, and to back that promise
up. Experience shows that when offenders believe that promise, their
behavior changes. It may be possible to make sure that the worst 100
domestic abusers in a jurisdiction get very special attention, and then let
the next 1000 know that behavior on their part will win them a place in
the top group. It may be possible to warn a dozen street drug markets that
there will be a crackdown in a week, and that they can protect themselves
by shutting down before it comes.
This is very much old wine in new bottles, traditional building blocks
assembled in commonsense but original ways. It bows in both directions:
166 David M. Kennedy

to the conviction on the part of police and other authorities that enforce-
ment is (or should be) powerful, justified, and important, and to the
painfully clear fact that enforcement as usual too often simply does not
work. Many in criminal justice have long been aware of the latter without
being willing or able to give up the former. These strategies, at least to
some extent, offer a way forward.

Communication, relationship, and common ground

Communication with offenders is a central element in pulling levers
strategies. Communication and persuasion have long been recognized as
essential elements in the deterrence process; sanctions offenders do not
know about cannot deter them, and sanctions offenders do not believe in
will not deter them (Zimring and Hawkins 1973; Kennedy 2003). Com-
munication has in practice gotten very little attention, however, either
theoretically or practically. We tend to assume what we know in fact to be
false: that offenders understand the complicated and often rapidly shift-
ing enforcement environment they face. One central lesson of pulling
levers operations is the clear power of direct communication with offend-
ers. A city’s-worth of drug crews may utterly ignore the fact – may not
collectively even be aware of the fact – that one crew has just been taken
off the street. But tell them all that this has happened, and beyond that,
that attention to groups is now standard operating procedure, and beyond
that, that the next group that commits violence will be treated likewise,
and beyond that, that enforcement attention will be heightened if groups
are active in street drug sales – and matters may change. The evidence is
that matters do change.
As experience with these strategies has accumulated, two related
themes have emerged. One is that relationships between authorities and
offenders matter a great deal. There is no acceptable place for this notion
in traditional enforcement thinking. The operational and legal conven-
tions of traditional law enforcement and case processing – call by call,
incident by incident, case by case – run counter. The atmosphere and
resonances of policing and much of the rest of law enforcement, with
its division of the world into white hats and black hats, also very much
run counter. Where the idea of relationship does have a place, as is some-
times true in probation and parole, there is a tendency to partition out the
exercise of authority – leading, for example, to the framing of supervisees
as “clients,” which leads to much eye-rolling by the more enforcement
minded.
Pulling levers strategies structure a very different kind of relation-
ship. Authorities meet with offenders, as in call-ins, in a setting that is
 Old wine in new bottles 167

respectful, civil, but also unabashedly about authority and consequences.

Such contacts are often repeated. They are often supplemented as needed
in other settings, such as pointed warnings on the street when front-line
intelligence suggests that trouble is brewing between particular groups.
They frequently include other elements, such as the offering of services
and impassioned statements by community figures. One of the most strik-
ing features of call-ins for police officers and other enforcement figures is
that offenders tend to be well behaved, attentive, and often visibly moved
by what ministers, mothers who have lost their children to the street,
and neighborhood elders have to say to them. It is not uncommon for
offenders to applaud authorities who speak to them uncompromisingly
but also respectfully. Once the call-in process becomes known on the
streets, offenders sometimes show up uninvited and ask to be allowed to
sit in. The contrast with the conventional wisdom – that offenders are
wild, self-destructive, unreachable, and the like – is graphic.
It is but a short step to the idea that there is in fact common ground
between authorities and offenders. There is also no place for this in tradi-
tional enforcement thinking, which assumes a constant, zero-sum strug-
gle between the forces of good and bad. In fact, there is a great deal of
common ground. Offenders do not wish to die; for the most part they do
not wish to be arrested and go to prison; they do not want their mothers
to grieve; they do not want their younger siblings following in their foot-
steps or walking deadly streets; they would, all else being equal, prefer to
do their business quietly and safely. These are, at least sometimes, goals
authorities and offenders can seek together. Each gang may put down its
guns if it understands that all are putting them down. Drug dealers may
shift the ways in which they do business if they understand that police will
be paying particular attention to the street markets and crack houses that
most damage and endanger communities. Such common ground is cer-
tainly not absolute, or always to be found, but it is also not unknown. The
clear fact that offenders respond to, and sometimes quite evidently wel-
come, the new rules established in pulling levers strategies is beginning
to suggest to authorities that offenders might want things to be different,
or at least that they may be far less resistant than is invariably assumed.
It is, like the rest of the pulling levers framework, a notion that is both
commonsensical and radical.

The rest of the story

All this is promising and provocative. Not all the news, of course, is
good. These strategies are demonstrably potent in addressing serious
violence. Whether they can be extended to other substantive problems
168 David M. Kennedy

remains to be seen. Where they have been effective, they have had a large
impact on very particular problems; they have not had broad effects on
a whole range of crime issues, as has happened, for example, in New
York City. While some police departments and other criminal justice
agencies are intrigued by these ideas, resistance and skepticism remain
deep. Their staying power within implementing agencies remains very
much in question. Boston, where the strategy originated, let Operation
Ceasefire wither, as did – after an even more striking apparent initial
success – Minneapolis (Kennedy and Braga 1998). Others have done
better; the initiative in Indianapolis, for example, remains healthy and
continues to evolve. But the same elements that make these operations
promising – the innovative thinking, the focused operations, the intera-
gency partnerships, the structuring of new relationships with offenders –
also present challenges to getting them off the ground and keeping them
in place. Local political strife, for example, scuttled the enterprise in
Baltimore just at the point that it was demonstrating initial success in the
field (Braga, Kennedy, and Tita 2002). In particular, the lion’s share of
the work inevitably falls on police departments, who have most of the
necessary street knowledge and enforcement capacity. Trying to mobi-
lize and sustain police action frequently shows, in the rudest imaginable
ways, the shockingly poor management and accountability mechanisms
within police departments, and departments’ consequent abilities sim-
ply to refuse – obviously and successfully – what their executive levels
demand of them.
At this point, it seems fair to say that the “pulling levers” framework is a
bright spot in the larger community and problem-solving policing agenda.
That agenda is not being realized as one might wish: nearly all police
agencies still do nearly all their work the same way they did twenty years
ago at the beginning of the “new policing” movement (Goldstein 2002).
Occasional standout initiatives do not change that clear truth. The spread
of pulling levers strategies stands in some small way in opposition to
that stasis. Like the other main exception, the spread of broken windows
policing, the strategies have what appeal they have to law enforcement
practitioners primarily because they allow mostly traditional tactics to be
deployed in new ways with the promise of considerably greater results.
We shall see whether even that movement is sustained, and beyond that
whether the possibly quite considerable potential for even more innovative
thought and action can be realized.

   
1. These data are available upon request from author.
2. Matt Miranda, Office of the United States Attorney, personal communication.
 Old wine in new bottles 169


Braga, A. (2002). Problem-oriented policing and crime prevention. Monsey, New
York: Criminal Justice Press.
Braga, A., Kennedy, D., and Tita, G. (2002). New approaches to the strategic
prevention of gang and group-involved violence. In C. R. Huff (ed.), Gangs
in America, 3rd ed. Thousand Oaks, CA: Sage Publications.
Braga, A., McDevitt, J., and Pierce, G. (2005). Understanding and preventing
gang violence: Problem analysis and response development in Lowell, Mas-
sachusetts. Police Quarterly, 8.
Braga, A., Piehl, A., and Kennedy, D. (1999). Youth homicide in Boston: An
assessment of supplementary homicide report data. Homicide Studies, 3,
277–299.
Braga, A., Kennedy, D., Waring, E., and Piehl, A. (2001). Problem-oriented
policing, deterrence, and youth violence: An evaluation of Boston’s operation
ceasefire. Journal of Research in Crime and Delinquency, 38, 195–225.
Clarke, R. and Weisburd, D. (1994). Diffusion of crime control benefits: Obser-
vations on the reverse of displacement. Crime prevention studies, 2, 165–
184.
Dalton, E. (2003). Lessons in preventing homicide. Project Safe Neighborhoods
Report. Lansing, MI: School of Criminal Justice, Michigan State University.
Goldstein, H. (2002). On further developing problem-oriented policing: The
most critical need, the major impediments, and a proposal. In J. Knutsson
(ed.), Mainstreaming problem-oriented policing. Monsey, NY: Criminal Justice
Press.
Kennedy, D. (1990). Fighting the drug trade in Link Valley. Case C16-90-
935.0. Cambridge, MA: John F. Kennedy School of Government, Harvard
University.
(1997). Pulling levers: Chronic offenders, high-crime settings, and a theory of
prevention. Valparaiso University Law Review, 31, 449–484.
(1998). Pulling levers: Getting deterrence right. National Institute of Justice
Journal, July 2–8.
(2002a). A tale of one city: Reflections on the Boston Gun Project. In
G. Katzmann (ed.), Securing our children’s future: New approaches to juvenile
justice and youth violence. Washington, DC: Brookings Institution Press.
(2002b). Controlling domestic violence offenders. Report submitted to the Hewlett-
Family Violence Prevention Fund. Cambridge, MA: John F. Kennedy School
of Government, Harvard University.
(2003). Reconsidering deterrence. Final report submitted to the US National
Institute of Justice. Cambridge, MA: John F. Kennedy School of Govern-
ment, Harvard University.
Kennedy, D. and Braga, A. (1998). Homicide in Minneapolis: Research for prob-
lem solving. Homicide Studies, 2, 263–290.
Kennedy, D., Braga, A., and Piehl, A. (1997). The (un)known universe: Mapping
gangs and gang violence in Boston. In D. Weisburd and J. T. McEwen (eds.),
Crime Mapping and Crime Prevention. Monsey, NY: Criminal Justice Press.
McGarrell, E. and Chermak, S. (2003). Strategic approaches to reducing firearms
violence: Final report on the Indianapolis violence reduction partnership. Final
170 David M. Kennedy

report submitted to the US National Institute of Justice. East Lansing, MI:

School of Criminal Justice, Michigan State University.
San Diego Police Department. 1998. Operation “hot pipe/smoky haze/rehab.”
Paper published in the proceedings of the 9th Annual International Problem-
Oriented Policing Conference. Washington, DC: Police Executive Research
Forum.
Tita, G., Riley, K. J., Ridgeway, G., Grammich, C., Abrahamse, A., and Green-
wood, P. (2003). Reducing gun violence: Results from an intervention in East
Los Angeles. Santa Monica, CA: Rand Corporation.
Travis, J., Kennedy, D., Roman, C., Beckman, K., Solomon, A., and Turner, E.
(2003). An analysis of homicide incident reviews in the District of Columbia.
Unpublished report. Washington, DC: Justice Policy Center, The Urban
Institute.
Wakeling, S. (2003). Ending gang homicide: Deterrence can work. Perspectives on
violence prevention, 1. Sacramento, CA: California Attorney General’s Office
and the California Health and Human Services Agency.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder,
and fear? Annals of the American Academy of Political and Social Science, 593,
42–65.
Winship, C. and Berrien, J. (1999). Boston cops and black churches. The Public
Interest, 136, 52–68.
Zimring, F. and Hawkins, G. (1973). Deterrence: The legal threat in crime control.
Chicago: University of Chicago Press.
9 Critic
Partnership, accountability, and innovation:
clarifying Boston’s experience with pulling
levers

Anthony A. Braga and Christopher Winship

The pulling levers focused deterrence strategy has been embraced by

the US Department of Justice as an effective approach to crime pre-
vention. In his address to the American Society of Criminology, former
National Institute of Justice Director Jeremy Travis (1998) announced
“[the] pulling levers hypothesis has made enormous theoretical and prac-
tical contributions to our thinking about deterrence and the role of the
criminal justice system in producing safety.” Pioneered in Boston to halt
youth violence, the pulling levers framework has been applied in many
American cities through federally sponsored violence prevention pro-
grams such as the Strategic Alternatives to Community Safety Initiative
and Project Safe Neighborhoods (Dalton 2002). In its simplest form, the
approach consists of selecting a particular crime problem, such as youth
homicide; convening an interagency working group of law enforcement
practitioners; conducting research to identify key offenders, groups, and
behavior patterns; framing a response to offenders and groups of offend-
ers that uses a varied menu of sanctions (“pulling levers”) to stop them
from continuing their violent behavior; focusing social services and com-
munity resources on targeted offenders and groups to match law enforce-
ment prevention efforts; and directly and repeatedly communicating with
offenders to make them understand why they are receiving this special
attention (Kennedy 1997; Kennedy in this volume).
Despite the enthusiasm for the approach, there is relatively little
rigorous scientific evidence that pulling levers deterrence strategies
have been useful in preventing violence beyond the Boston experience
(Wellford, Pepper, and Petrie 2005). Even in Boston, the exact contribu-
tion of pulling levers to the reduction of youth violence remains unclear
(Wellford et al. 2005). Moreover, high-profile replications of the Boston
approach have been difficult to implement and sustain. In Baltimore,
local political problems undermined the implementation process (Braga,
Kennedy, and Tita 2002). In Minneapolis, the strategy was abandoned as

171
172 Anthony A. Braga and Christopher Winship

the participating agencies returned to their traditional methods of dealing

with violence (Kennedy and Braga 1998).
We believe that the difficulties experienced by other jurisdictions stem
from a limited understanding of the larger Boston story. Boston’s suc-
cess in reducing youth violence has been attributed to a wide variety of
programs and strategies: public health interventions, police–probation
partnerships, enhanced federal prosecutions, police–black minister part-
nerships, and the pulling levers focused deterrence strategy known as
Operation Ceasefire. Many observers have suggested that these are iso-
lated and competing explanations. For example, in his discussion of Oper-
ation Ceasefire and the Boston Police Department’s collaboration with
activist black ministers, Fagan (2002) describes these as “two distinct
and contrasting narratives [that] comprise the Boston story” (136).
In reality, the Boston story consists of multiple interconnected lay-
ers. As we discuss below, the implementation of Ceasefire was possible
because of newly formed relationships among the police and other law
enforcement and social service agencies and between the police and the
community, with the latter creating important mechanisms for police
accountability. Thus, although available quantitative evidence suggests
that Operation Ceasefire was the key initiative associated with a signifi-
cant reduction in youth violence, a fuller and more nuanced description
of the Boston experience is needed. A narrow and inappropriate inter-
pretation of Boston’s success as simply being due to Operation Ceasefire
creates the danger of unrealistic expectations of success, serious imple-
mentation problems in replicating the Ceasefire program, and an inability
to sustain implemented violence prevention programs.
In order to understand the innovations that took place in Boston’s
policing strategies during the 1990s it is necessary to examine the impor-
tance of two key elements that created the foundation that made change
possible. First, in order for the Boston Police to develop an innovative
program involving a variety of partners, it was essential to have estab-
lished a “network of capacity” consisting of dense and productive rela-
tionships from which partners could be drawn. Second, because of the
long history of perceived racism by the Boston Police, a new mechanism
of police accountability was necessary in order to create trust that new
programs would be beneficial to the community. This trust was essential
for establishing needed community and political support for innovative
efforts by the Boston Police. Operation Ceasefire simply could not have
been launched without either a network of partners who were a central
component of its design or the trust that derived from accountability.
This chapter begins by briefly describing the key elements of Cease-
fire. It then examines the available evidence on Ceasefire’s effect on
 Partnership, accountability, and innovation 173

serious violence in Boston and elsewhere. It subsequently discusses

the implementation of Ceasefire within the changing political con-
text of police–community relationships and evolving police partnerships
with other agencies. Finally, it analyses the implications of the fuller
Boston story for replicating and sustaining the Boston approach in other
jurisdictions.

The Boston Gun Project and Operation Ceasefire

Like many American cities during the late 1980s and early 1990s, Boston
suffered an epidemic of youth violence that had its roots in the rapid
spread of street-level crack-cocaine markets (Kennedy, Piehl, and Braga
1996). The Boston Gun Project was a problem-oriented policing project
aimed at preventing and controlling serious youth violence. The problem
analysis phase of the Project began in early 1995 and the Operation Cease-
fire strategy was implemented in mid 1996. The trajectory of the Project
and of Ceasefire has been extensively documented (see e.g., Kennedy
et al. 1996; Kennedy, Braga, and Piehl 2001; Kennedy in this volume).
Briefly, a problem-solving working group of law enforcement personnel,
youth workers, and researchers diagnosed the youth violence problem in
Boston as one of patterned, largely vendetta-like hostility amongst a small
population of highly active criminal offenders, and particularly amongst
those involved in some sixty loose, informal, mostly neighborhood-based
gangs. Based on the problem-analysis findings, the Boston Gun Project
working group crafted the Operation Ceasefire initiative that was tightly
focused on disrupting ongoing conflicts among youth gangs.
The Boston Police Departments’s Youth Violence Strike Force
(YVSF), an elite unit of some forty officers and detectives, coordinated
the actions of Operation Ceasefire. An interagency working group, com-
posed of law enforcement personnel, youth workers, and members of
Boston’s Ten Point Coalition of activist black clergy, was convened on
a biweekly basis to address outbreaks of serious gang violence. Opera-
tion Ceasefire’s pulling levers strategy was designed to deter gang vio-
lence by reaching out directly to gangs, saying explicitly that violence
would no longer be tolerated, and backing up that message by “pulling
every lever” legally available when violence occurred (Kennedy 1997).
These law enforcement levers included disrupting street-level drug mar-
kets, serving warrants, mounting federal prosecutions, and changing the
conditions of community supervision for probationers and parolees in
the targeted group. Simultaneously, youth workers, probation and parole
officers, and clergy offered gang members services and other kinds of
help. If gang members wanted to step away from a violent lifestyle, the
174 Anthony A. Braga and Christopher Winship

Ceasefire working group focused on providing them with the services and
opportunities necessary to make the transition.
The Ceasefire Working Group delivered their anti-violence message
in formal meetings with gang members; through individual police and
probation contacts with gang members; through meetings with inmates
of secure juvenile facilities in the city; and through gang outreach workers
(Kennedy et al. 2001). The deterrence message was not a deal with gang
members to stop violence. Rather, it was a promise to gang members
that violent behavior would evoke an immediate and intense response. If
gangs committed other crimes but refrained from violence, the normal
workings of police, prosecutors, and the rest of the criminal justice system
dealt with these matters. But if gang members hurt people, the Working
Group focused its enforcement actions on them.
A large reduction in the yearly number of Boston youth homicides
followed immediately after Operation Ceasefire was implemented in mid-
1996. This reduction was sustained for the next five years (see Figure 9.1).
The Ceasefire program, as designed, was in place until 2000. During
the early years of the new millennium, the Boston Police experimented
with a broader approach to violence prevention by expanding certain
Ceasefire tactics to a broader range of problems such as serious repeat
violent gun offenders, the re-entry of incarcerated violent offenders back
into high-risk Boston neighborhoods, and criminogenic families in hot
spot areas. These new approaches, known broadly as Boston Strategy II,
seemed to diffuse the ability of Boston to respond to ongoing conflicts
among gangs. Homicide, most of which is gang related, has returned
as a serious problem for the City of Boston. Homicide has been rising
since 2001 with the sharpest increase among victims aged 25 and older
(Figure 9.1). In Fall 2004, the Boston Police implemented a new violence
prevention campaign, which borrows heavily from Ceasefire’s tight focus
on disrupting cycles of violent gang retribution (Winship forthcoming).

Evidence on the impact of Ceasefire on serious violence

A US Department of Justice (DOJ)-sponsored evaluation of Operation
Ceasefire used a non-randomized control group design to analyze trends
in serious violence between 1991 and 1998. The evaluation reported that
the intervention was associated with a 63 percent decrease in the monthly
number of Boston youth homicides, a 32 percent decrease in the monthly
number of shots-fired calls, a 25 percent decrease in the monthly number
of gun assaults, and, in one high-risk police district given special attention
in the evaluation, a 44 percent decrease in the monthly number of youth
gun assault incidents (Braga, Kennedy, Waring, and Piehl 2001). The
 CEASEFIRE
INTERVENTION
May 15, 1996
80

70

60

50

40

30

Number of Victims
20

10

76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 993 994 995 996 997 998 999 000 001 002 003
19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 1 1 1 1 1 1 1 2 2 2 2
Year

Figure 9.1 Youth homicide in Boston, 1976–2003, victims ages 24 and under
176 Anthony A. Braga and Christopher Winship

evaluation also suggested that Boston’s significant youth homicide reduc-

tion associated with Operation Ceasefire was distinct when compared to
youth homicide trends in most major US and New England cities.
Other researchers, however, have observed that some of the decrease in
homicide may have occurred without the Ceasefire intervention in place
as violence was decreasing in most major US cities. Fagan’s (2002) cur-
sory review of gun homicide in Boston and in other Massachusetts cities
suggests a general downward trend in gun violence that existed before
Operation Ceasefire was implemented. Levitt (2004) analyzed homicide
trends over the course of the 1990s and concluded that the impact of
innovative policing strategies, such as Operation Ceasefire in Boston and
broken windows policing and Compstat in New York, on homicide was
limited. Other factors, such as increases in the number of police, the
rising prison population, the waning crack-cocaine epidemic, and the
legalization of abortion, can account for nearly all of the national decline
in homicide, violent crime, and property crime in the 1990s.
The National Academies’ Panel on Improving Information and Data
on Firearms (Wellford et al. 2005) concluded that the Ceasefire evalua-
tion was compelling in associating the intervention with the subsequent
decline in youth homicide. However, the Panel also suggested that many
complex factors affect youth homicide trends and it was difficult to specify
the exact relationship between the Ceasefire intervention and subsequent
changes in youth offending behaviors. While the DOJ-sponsored evalua-
tion controlled for existing violence trends and certain rival causal factors
such as changes in the youth population, drug markets, and employ-
ment in Boston, there could be complex interaction effects among these
factors not measured by the evaluation that could account for some mean-
ingful portion of the decrease. The evaluation was not a randomized,
controlled experiment. Therefore, the non-randomized control group
research design cannot rule out these internal threats to the conclusion
that Ceasefire was the key factor in the youth homicide decline.
Like the Panel, we believe that Ceasefire was responsible for a mean-
ingful proportion of the youth homicide decline. However, it is difficult
to determine the exact contribution of Ceasefire to the decline. Clearly,
other factors were responsible for some of the decline. As Figure 9.2
reveals, there was a parallel decrease of approximately equal magnitude
in adult homicide that can only be partly explained by Ceasefire’s poten-
tial effect on the behavior of adults participating in violent gang dynamics.
Figure 9.1 also presents a long-term picture of youth homicide that sug-
gests some of the decline may result from a “regression to the mean”
phenomenon. Youth homicides dramatically increase in 1989 and remain
historically high through 1995. While post-Ceasefire youth homicide
 CEASEFIRE
INTERVENTION
May 15, 1996
90

80

70

60

50

40

Number of Victims
30

20

10

76 77 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 000 001 002 003
19 19 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2
Year

Figure 9.2 Adult homicide in Boston, 1976–2003, victims ages 25 and older
178 Anthony A. Braga and Christopher Winship

counts are lower than counts during the 1976–88 period, it seems plau-
sible that some portion of the decline was part of a natural return to an
average count of youth homicide. This certainly raises questions about the
effectiveness of Ceasefire. Experimental research is necessary to uncover
the true crime prevention benefits of engaging a pulling levers strategy.
The National Academies’ Panel also found that the evidence on the
effectiveness of the pulling levers focused deterrence strategy is quite
limited (Wellford et al. 2005). The available evidence on the effects
of pulling levers programs in other jurisdictions is scientifically weak.
Assessments of these programs in other jurisdictions did not use control
groups and usually consisted of simple pre-post measurements of trends
in violence (see, e.g., Braga et al. 2002; McGarrell and Chermak 2003).
In Baltimore and Minneapolis, two well-known replications of the Boston
experience, violence prevention initiatives rapidly unraveled and were
abandoned (Kennedy and Braga 1998; Kennedy in this volume). As dis-
cussed further below, we believe that the difficulty other jurisdictions
have had in replicating and sustaining a pulling levers focused deterrence
strategy may, in part, stem from a weak understanding of the context in
which the Boston intervention was implemented.

The larger Boston story I: The development of a

“network of capacity”
Missing from the account of Operation Ceasefire reported in most law
enforcement circles is the larger story of an evolving collaboration that
spanned the boundaries that divide criminal justice agencies from one
another, criminal justice agencies from human service agencies, and crim-
inal justice agencies from the community. Such collaborations are nec-
essary to legitimize, fund, equip, and operate complex strategies that are
most likely to succeed in both controlling and preventing youth violence
(Moore 2002). The solid working relationships that were at the heart of
the interagency working group process were developed long before the
Boston Gun Project commenced in 1995. In essence, Boston created a
very powerful “network of capacity” to prevent youth violence (Moore
2002). This network was well positioned to launch an effective response
to youth violence because criminal justice agencies, community groups,
and social service agencies coordinated and combined their efforts in
ways that could magnify their separate effects. Ceasefire capitalized on
these existing relationships by focusing the network on the problem of
serious gang violence.
Criminal justice agencies work largely independent of each other,
often at cross-purposes, often without coordination, and often in an
 Partnership, accountability, and innovation 179

atmosphere of distrust and dislike (Kennedy 2002). Until the height of the
youth violence epidemic, this observation was certainly true in Boston.
It was painfully apparent that no one agency could mount a meaningful
response to the gang violence that was spiraling out of control in the city.
The crisis forced Boston criminal justice agencies to work together and
develop new approaches to deal with the violence problem. YVSF officers
and detectives and line-level workers from other criminal justice agencies
collaborated on a variety of innovative programs, including: Operation
Night Flight – a police–probation partnership to ensure at-risk youth were
abiding by the conditions of their release into the community (Corbett,
Fitzgerald, and Jordan 1998); Safe Neighborhoods Initiatives – a com-
munity prosecution program that was rooted in a partnership between the
Suffolk County District Attorney’s Office, the Boston Police, and com-
munity members in hot spot neighborhoods (Coles and Kelling 1999);
and a partnership between the Boston Police, the Bureau of Alcohol,
Tobacco, and Firearms (ATF), and the US Attorney’s Office to identify
and apprehend illegal gun traffickers who were providing guns to violent
gangs (Kennedy et al. 1996).
The YVSF also formed working relationships with social service and
opportunity provision agencies. For certain prevention initiatives, the
YVSF was the lead agency involved in the program such as the Summer
of Opportunity program that provides at-risk youth with job training and
leadership skills that could be transferred to workplace, school, or home
settings. More often, however, the police supported the activities of youth
social service providers from community-based organizations such as the
Boston Community Centers’ streetworker program and the Dorchester
Youth Collaborative. YVSF officers and detectives would encourage at-
risk youth to take advantage of these resources and also consider the input
of youth workers in determining whether certain gang-involved youth
would be better served by prevention and intervention actions rather
than enforcement actions.
When the Boston Gun Project was initiated, the YVSF had already
developed a network of working relationships that could be powerfully
channeled by a more focused initiative like Operation Ceasefire. Crimi-
nal justice agency partnerships provided a varied menu of enforcement
options that could be tailored to particular gangs. Without these partner-
ships, the available “levers” that could be pulled by the working group
would have been limited. Social service and opportunity provision agen-
cies were also integrated into Ceasefire interventions to provide a much-
needed “carrot” to balance the law enforcement “stick.” The inclusion of
prevention and intervention programs in the Ceasefire intervention was
vitally important in securing community support and involvement in the
180 Anthony A. Braga and Christopher Winship

program. We believe that the legitimacy conferred upon the Ceasefire ini-
tiative by key community members was an equally important condition
that facilitated the successful implementation of this innovative program.

The larger Boston story II: Accountability and

police–community relations
There was a radical change in the relationship between the Boston Police
and Boston’s minority communities that pre-dated Ceasefire and had a
profound influence on the trajectory of the Ceasefire intervention. This
collaborative relationship, led by Ten Point Coalition activist black minis-
ters, developed in the context of a high level of community dissatisfaction
with policing strategies and tactics engaged by the Boston Police (Winship
and Berrien 1999). When the violence epidemic started in the late 1980s,
the Boston Police were ill equipped to deal with the sudden increase in
serious youth violence. The Boston Police relied upon highly aggressive
and reportedly indiscriminate policing tactics to deal with street gang
violence (Winship and Berrien 1999; Berrien and Winship 2002; 2003).
A series of well-publicized scandals emanating from an indiscriminate
policy of stopping and frisking all black males in high-crime areas out-
raged Boston’s black community. Perhaps the most important was the
1989 murder of Carol Stuart, a pregnant white woman on her way home
from Boston City Hospital. Initially, Charles Stuart, the victim’s husband
who was the actual murderer, led Boston Police investigators to believe
that the murderer was a black male. The police responded by blanketing
the Mission Hill housing projects for a suspect. Abusive police conduct
was reported to be widespread as coerced statements led to the wrongful
arrest of a black male. The black community and the local media were
outraged and condemned the discriminatory actions of the investigating
officers. The Carol Stuart case and other scandals led to the establish-
ment of the St. Clair Commission, an independent committee appointed
to investigate the policies and practices of the Boston Police. In 1992
it released its report, which cited extensive corruption and incompetent
management, and called for extensive reform including the replacement
of top personnel.
In response, the Boston Police overhauled its organization, mission,
and tactics during the early 1990s. The existing command staff, includ-
ing the commissioner, were replaced with new officers who were known
to be innovative and hardworking; investments were made to improve
the department’s technology to understand crime problems; a neighbor-
hood policing plan was implemented; and beat-level officers were trained
in the methods of community and problem-oriented policing. In 1991,
 Partnership, accountability, and innovation 181

the Anti-Gang Violence Unit (AGVU) was created and charged with
disrupting ongoing gang conflicts rather than mounting an aggressive
campaign to arrest as many offenders as possible. By 1994, the AGVU
evolved into the YVSF and its mandate was broadened beyond control-
ling outbreaks of gang violence to more general youth violence preven-
tion. While these changes were important in creating an environment
where the police could collaborate with the community, residents of
Boston’s poor minority neighborhoods remained wary of and dissatisfied
with a police department that had a long history of abusive and unfair
treatment.
In 1992, a loosely allied group of activist black clergy formed the Ten
Point Coalition after a gang invasion of the Morningstar Baptist Church.
During a memorial for a slain rival gang member, mourners were attacked
with knives and guns (Winship and Berrien 1999; Kennedy et al. 2001;
Berrien and Winship 2002; 2003). In the wake of that outrage, the Ten
Point Coalition ministers decided they should attempt to prevent the
youth in their community from joining gangs, and also that they needed
to send an anti-violence message to all youth, whether gang-involved or
not.
Initially, the ministers assumed an adversarial role to the Boston Police
and were highly critical in the public media of police efforts to prevent
youth violence. However, as the ministers worked the streets, they started
to form effective relationships with particular YVSF officers and devel-
oped a shared understanding of the nature of youth violence in Boston:
only a small number of youths in the neighborhoods were involved in
violence, many of these gang-involved youth were better served by inter-
vention and prevention strategies, and only a small number of these gang-
involved youths needed to be removed from the streets through arrest and
prosecution strategies.
The Ten Point ministers also sheltered the police from broad pub-
lic criticism while the police were engaged in activities the ministers
deemed to be of interest to the community and its youth. In 1995, Paul
McLaughlin, a local gang prosecutor who was white, was murdered on
his way home from work. The initial description of the assailant (“young
black male wearing a hooded sweatshirt and baggy pants”) was vague
enough to cause concern to many in the black community that an “open
season on young black males” similar to that during the Carol Stuart
investigation would occur (Grunwald and Anand 1995). Fortunately,
these initial fears were unfounded as the black ministers and the Boston
Police supported each other in the handling of the media and the ensuing
investigation. The black ministers publicly praised the police for showing
restraint in their conduct and the police praised the ministers for their
182 Anthony A. Braga and Christopher Winship

willingness to provide help and keep the community calm (Berrien and
Winship 2002; 2003).
Prior to Ceasefire, the Ten Point ministers also helped the Boston Police
manage negative publicity by the local media after several potentially
explosive events ranging from the beating of a black undercover offi-
cer by uniformed police officers (Chacon 1995) to the accidental death
of 75-year-old retired minister who suffered a fatal heart attack after a
botched drug raid (Mallia and Mulvihill 1994). In these cases, the minis-
ters took two positions. First, they demanded that the police department
take responsibility for its actions – investigate incidents thoroughly and
hold those involved accountable. Second, after it was clear that the Boston
Police was accepting responsibility, the ministers communicated to the
community that the police were in fact reacting appropriately. This, in
turn, prevented these situations from becoming racially explosive and
provided the police with the continued political support they needed in
order to undertake policy innovations, such as Ceasefire. In more recent
years, the ministers have continued to play this dual role with regards to
fatal police shootings, eight of which occurred over a 22-month period
between 2000 and 2002 (Tench 2002; Winship forthcoming).
While the Ten Point ministers were not involved in the design of the
Ceasefire intervention, they were influential as an informal “litmus test”
of the types of enforcement actions that would and would not be toler-
ated by the community. The youth workers participating in the design
of Ceasefire would voice their concerns about community reaction to
any proposed enforcement tactics that could be viewed as overly aggres-
sive. However, what usually ended discussions was the recognition of
the political vulnerability of the Boston Police to the consequences of
the Ten Point ministers potentially reporting any questionable practices
to local media and, more importantly, exerting pressure on the Mayor’s
Office to deal with perceived inappropriate actions by the Department.
For example, while discussing plausible interventions, the working group
considered the notable gun violence reduction results of the Kansas City
Gun Experiment, which involved intensive enforcement of laws against
illegally carrying concealed firearms via safety frisks during traffic stops,
plain view, and searches incident to arrest in gun violence hot spot areas
(Sherman and Rogan 1995). After some discussion, the working group
rejected the idea of engaging a hot spots policing strategy as the Boston
Police did not want to adopt an enforcement program that could be
viewed by the Ten Point ministers as a return to the indiscriminate “stop
and frisk” policies of the past.
When Ceasefire was ready to be implemented, the commander of
the YVSF presented the program to key black ministers to obtain their
 Partnership, accountability, and innovation 183

approval of and involvement in the initiative. The Boston Police knew

that they would need the political support of the Ten Point Coalition to
pursue aggressive enforcement actions against hardcore gang members
who were central to violent conflicts. While the Ceasefire initiative was
a violence prevention campaign, given the Carol Stuart case and other
incidents, the community and local media could have easily misunder-
stood the enforcement tactics as simply another law enforcement initiative
designed to arrest large numbers of young black men. The ministers rec-
ognized the value of the Ceasefire approach to violence prevention as it
was carefully focused only on violent gang-involved youth and provided
social services and opportunities to gang youth who desired them. After
Ceasefire was implemented, Ten Point Coalition ministers became regu-
lar members of the working group. Ministers played key roles in working
with the police to identify dangerous gang-involved youth, communicat-
ing the anti-violence deterrence message to all youth and, with the help
of social service providers, offering assistance to gang youth who wanted
to step away from their violent lifestyles.
By including the ministers in the Ceasefire working group, the Boston
Police developed a mechanism for transparency and accountability that
was very desirable to Boston’s minority community. Through their
involvement in Ceasefire, the ministers became part of the process of
determining which gang interventions would be done and when. In addi-
tion, they, along with others, gave gang members the message that they
had a choice: stop the violence and they would be helped – with school,
a job, family; continue and the full weight of the law (and the commu-
nity) would come down on them, with every possible lever being used to
see that they were incarcerated. At a more general level, a shared under-
standing of the reality of youth violence and the actions that were neces-
sary to prevent and control that violence emerged (Berrien and Winship
2002; 2003). The transparency and involvement in the enforcement pro-
cess built trust and further solidified a functional working relationship
between the community and the Boston Police. In turn, by engaging
in a process through which they were meaningfully and appropriately
accountable to the community, the Boston Police created the political
support, or “umbrella of legitimacy,” that they needed to pursue more
focused and perhaps more aggressive intervention than would have been
possible otherwise (Berrien and Winship 2002).

Implications of the larger story for other jurisdictions

Operation Ceasefire became a nationally recognized model for youth
violence reduction programs and many jurisdictions quickly started to
184 Anthony A. Braga and Christopher Winship

experiment with the approach (Kennedy in this volume). Unfortunately,

despite some initially promising results, many of these replications were
never fully implemented or were eventually abandoned. Braga has been
involved in replication efforts in a number of cities, including Baltimore,
Minneapolis, and San Francisco, and these jurisdictions simply did not
have an adequate network of capacity in place before adopting a Ceasefire-
like approach to youth violence. Operation Ceasefire was a “relationship
intensive” intervention based on trust and the ability of a diverse set of
individuals to work together toward a common goal. The narrow descrip-
tion of Ceasefire that currently circulates in criminal justice circles is, in
many ways, a recipe for frustration and eventual failure as it simplifies the
trajectory of the Boston experience.
Effective collaborations and the trust and accountability that they entail
are essential in launching a meaningful response to complex youth vio-
lence problems. However, the fact that such collaborations are needed
does not guarantee that they inevitably a rise or, once developed, that
they are sustained. There are many significant obstacles to their develop-
ment and maintenance such as giving up control over scarce resources
that could compromise agencies’ traditional missions, aligning agencies’
individual work efforts into a functional enterprise, and developing a col-
lective leadership among a group of individuals aligned with the needs of
their individual organizations (Bardach 1998).
A central problem in creating and managing effective capacity-building
collaborations is overcoming the problem of distrust (Bardach 1998).
Distrust corrodes the creative process that criminal justice agencies and
community-based organizations are necessarily engaged in. Like most
cities, distrust characterized the relationship among criminal justice agen-
cies and between criminal justice agencies and the inner city community
in Boston. Practitioners and community members in Boston were able to
overcome their historical distrust and form productive working relation-
ships. These relationships existed before Ceasefire and were the founda-
tion upon which it was built. Of course, working groups can be forced
together and, sometimes, can implement short-term programs that have
promising initial results. However, if the initiative is not based on a shared
understanding of the problem and cemented through functional partner-
ships, the initiative will fall apart. These are key issues for other jurisdic-
tions to consider in replicating Operation Ceasefire and in sustaining the
collaborative effort once it has been launched.
In many community and problem-oriented policing projects, commu-
nity members serve as informants who report to the police on unac-
ceptable community conditions and the particulars of crime problems
(Skogan and Hartnett 1997; Braga 2002). They are rarely engaged as
 Partnership, accountability, and innovation 185

“partners” or “co-producers” of public safety. Police officers remain the

“experts” on crime who are primarily responsible for developing and
managing interventions to address crime problems. Through their collab-
oration with Ten Point ministers, the Boston Police discovered a system
whereby they were accountable to the community. This accountability
to the community became a great asset to the police. By engaging the
ministers in their violence prevention efforts and creating a sense of joint
ownership of the youth violence problem, the Boston Police created the
political support necessary for both innovation and more focused and
aggressive intervention. With the Ten Point’s approval of and involve-
ment in Operation Ceasefire, the community supported the approach as
a legitimate violence prevention campaign. Police strategies can acquire
true legitimacy within the inner city only if the community partner sup-
ports police tactics when they are appropriate as well as publicly criti-
cizes activities that are not (Berrien and Winship 2002; 2003; Winship
forthcoming). Given the potentially harsh law enforcement levers that
can be pulled as part of a Ceasefire-like program, we feel that com-
munity involvement is critical in replicating and sustaining such inten-
sive violence prevention initiatives. Without the political support of the
community, the police cannot pursue an innovative enforcement strategy
that targets truly dangerous youth at the heart of urban youth violence
problems.



Bardach, E. (1998). Getting agencies to work together. Washington, DC: Brookings

Institution Press.
Berrien, J. and Winship, C. (2002). An umbrella of legitimacy: Boston’s Police
Department – ten point coalition collaboration. In G. Katzmann (ed.), Secur-
ing our children’s future: New approaches to juvenile justice and youth violence.
Washington, DC: Brookings Institution Press.
(2003). Should we have faith in the churches? The ten-point coalition’s effect
on Boston’s youth violence. In B. Harcourt (ed.), Guns, crime, and punishment
in America. New York: New York University Press.
Braga, A. (2002). Problem-oriented policing and crime prevention. Monsey, NY:
Criminal Justice Press.
Braga, A., Kennedy, D., and Tita, G. (2002). New approaches to the strategic
prevention of gang and group-involved violence. In C. R. Huff (ed.), Gangs
in America (3rd ed.). Thousand Oaks, CA: Sage Publications.
Braga, A., Kennedy, D., Waring, E., and Piehl, A. (2001). Problem-oriented
policing, deterrence, and youth violence: An evaluation of Boston’s operation
ceasefire. Journal of Research in Crime and Delinquency, 38, 195–225.
Chacon, R. (1995). Boston police investigators seek cause of undercover officer’s
injuries. The Boston Globe, February 4, 22.
186 Anthony A. Braga and Christopher Winship

Coles, C. and Kelling, G. (1999). Prevention through community prosecution.

The Public Interest, 136, 69–84.
Corbett, R., Fitzgerald, B., and Jordan, J. (1998). Boston’s operation night light:
An emerging model for police-probation partnerships. In J. Petersilia (ed.),
Community corrections: Probation, parole, and intermediate sanctions. New York:
Oxford University Press.
Dalton, E. (2002). Targeted crime reduction efforts in ten communities: Lessons
for the Project Safe Neighborhoods initiative. US Attorney’s Bulletin, 50, 16–
25.
Fagan, J. (2002). Policing guns and youth violence. The Future of Children, 12,
133–151.
Grunwald, M. and Anand, G. (1995). Authorities praised; Some blacks wary.
The Boston Globe, September 30, 80.
Kennedy, D. (1997). Pulling levers: Chronic offenders, high-crime settings,
and a theory of prevention. Valparaiso University Law Review, 31, 449–
484.
(2002). A tale of one city: Reflections on the Boston gun project. In G. Katz-
mann (ed.), Securing our children’s future: New approaches to juvenile justice and
youth violence. Washington, DC: Brookings Institution Press.
Kennedy, D. and Braga, A. (1998). Homicide in Minneapolis: Research for prob-
lem solving. Homicide Studies, 2, 263–290.
Kennedy, D., Braga, A., and Piehl, A. (2001). Developing and implementing
operation ceasefire. In Reducing gun violence: The Boston gun project’s operation
ceasefire. Washington, DC: National Institute of Justice, US Department of
Justice.
Kennedy, D., Piehl, A., and Braga, A. (1996). Youth violence in Boston: Gun
markets, serious offenders, and a use-reduction strategy. Law and Contem-
porary Problems, 59, 147–196.
Levitt, S. (2004). Understanding why crime fell in the 1990s: Four factors that
explain the decline and six that do not. Journal of Economic Perspectives, 18,
163–190.
Mallia, J. and Mulvihill, M. (1994). Minister dies as cops raid wrong apartment.
The Boston Herald, March 26, 1.
McGarrell, E. and Chermak, S. (2003). Strategic approaches to reducing firearms
violence: Final report on the Indianapolis violence reduction partnership. Final
report submitted to the US National Institute of Justice. East Lansing, MI:
School of Criminal Justice, Michigan State University.
Moore, M. (2002). Creating networks of capacity: The challenge of managing
society’s response to youth violence. In G. Katzmann (ed.), Securing our chil-
dren’s future: New approaches to juvenile justice and youth violence. Washington,
DC: Brookings Institution Press.
Sherman, L. and Rogan, D. (1995). Effects of gun seizures on gun violence: “Hot
spots” patrol in Kansas City. Justice Quarterly, 12, 673–694.
Skogan, W. and Hartnett, S. (1997). Community policing, Chicago style. New York:
Oxford University Press.
Tench, M. (2002). Group offers support for Evans, points to progress in curbing
violence. The Boston Globe, September 21, 34.
 Partnership, accountability, and innovation 187

Travis, J. (1998). Crime, justice, and public policy. Plenary presentation to

the American Society of Criminology. Washington, DC: November 12.
(http://www.ojp.usdoj.gov/nij/speeches/asc.htm).
Wellford, C., Pepper, J., and Petrie, C. (eds.). (2005). Firearms and violence: A
critical review. Committee to Improve Research Information and Data on
Firearms. Committee on Law and Justice, Division of Behavioral and Social
Sciences and Education. Washington, DC: The National Academies Press.
Winship, C. (Forthcoming). The end of a miracle? Crime, faith, and partnership
in Boston in the 1990s. In R. Drew Smith (ed.), The public influence of black
churches, II. Durham, NC: Duke University Press.
Winship, C. and Berrien, J. (1999). Boston cops and black churches. The Public
Interest, 136, 52–68.
Part V

Third-party policing
10 Advocate
The case for third-party policing

Lorraine Mazerolle and Janet Ransley

Third-party policing is defined as police efforts to persuade or coerce

organizations or non-offending persons, such as public housing agen-
cies, property owners, parents, health and building inspectors, and
business owners to take some responsibility for preventing crime or reduc-
ing crime problems (Buerger and Mazerolle 1998: 301). In third-party
policing, the police create or enhance crime control nodes in locations
or situations where crime control guardianship was previously absent or
non-effective. Sometimes the police use cooperative consultation with
community members, parents, inspectors, and regulators to encourage
and convince third parties to take on more crime control or prevention
responsibility. Central, however, to third-party policing is the police use
of a range of civil, criminal, and regulatory rules and laws, to engage,
coerce (or force) third parties into taking some crime control responsi-
bility. It is the regulatory and legal provisions that dictate the process for
third-party intervention.
In third-party policing, laws and legal mechanisms are directed at will-
ing or unwilling non-offending third parties, with the object of facilitating
or coercing them into helping to control the behavior of offending ulti-
mate targets. This type of policing, however, is not new. Some might argue
that this is just good, proactive policing. Indeed, for many years the police
have sought alternative solutions to regulate activities and solve crime
problems (Eck and Spelman 1987; Goldstein 1990; Goldstein 2003).
The police often create “place managers” to guard and protect problem
places (Eck 1994; Felson 1994) and they have used situational responses
for many years to deal with on-going crime problems (Clarke 1992). The
police have been carrying out this mode of policing for years, whether
as part of their routine patrol activities, their problem-oriented policing
program, or as a crime prevention initiative.
A systematic review of third-party policing evaluations reveals that
about less than a third of third-party policing initiatives occur within
the context of problem-oriented policing (Mazerolle and Ransley 2005).
We suggest, therefore, that third-party policing (e.g., legal levers, civil

191
192 Lorraine Mazerolle and Janet Ransley

remedies, coercion of third parties) is more than a subset of problem-

solving responses (see Goldstein 1990; Scott 2000; Goldstein 2003).
We argue that third-party policing is distinguished from other models
of policing through its intrinsic links with societal trends in regulation
more generally and the legal/regulatory provisions that dictate the inter-
vention processes. Indeed, we argue that the proliferation of third-party
policing has not occurred in a vacuum as an idea born at the grassroots
of policing or because the police have just gotten better at good, proac-
tive policing. Rather, we argue that the pace, context, and prominence of
third-party policing initiatives has escalated in recent years as one of many
consequences in the move from centralized state control to a system of
decentralized networks of governance and crime control agents. The pro-
liferation and development of third-party policing is, we suggest, reflective
of external pressure and the general transformation of government and
governance taking place in contemporary society.
In this chapter we provide an analysis of third-party policing in a societal
context. We begin with an overview of the trends in regulation more
generally that have set the stage for third-party policing to emerge as
a popular crime control tactic. We then describe the key dimensions of
third-party policing and summarize the evaluation evidence. We conclude
our chapter with a short examination of the side-effects, fairness, and
equity concerns that are raised through third-party policing and propose
ways in which third-party policing could be appropriately implemented
into mainstream policing activities.

Governance, regulation, and third-party policing

The proliferation of third-party policing is part of a pattern of major
change, indeed a transformation, of government and governance taking
place in contemporary society. This political, legal, economic, and social
transformation has affected the institutions of government and civil soci-
ety and is also altering how we think about crime, its prevention, and its
control (Braithwaite 2000). Big, organizing themes like governance, risk,
and plurality are now intrinsically affecting crime control and policing
and the rise of third-party policing has emerged in the context of these
broader trends in governance and crime control.
Recent transformations of governance in Western democracies have
involved a movement away from state sovereignty and control to net-
works of power (see Braithwaite 2000). These changes have influenced
contemporary police practice and driven the societal push to third-party
policing. In this contemporary model of governance the state provides
only one of many nodes in a network of regulation.
 The case for third-party policing 193

The main mechanism for this transformation has been the develop-
ment of the notion of risk – economic, social, and political activities are
less subject to central control, and more likely to be monitored for risks
that need to be managed (Ericson and Haggerty 1997; O’Malley 2000).
Garland (1996; 1997) argues that risk and insurance promote a form of
“responsibilized” autonomy (Garland 1996: 452–455). That is, the bur-
den of managing risk has been shifted from governments to individuals,
who must become responsible for the outcomes of their own decisions
on risk. Prudent individuals will obtain insurance, identify and minimize
risk, and manage their own security.
Feeley and Simon (1994) apply the concepts of risk to criminal justice,
to develop the notion of actuarial justice. They describe an old and a new
penology – the old marked by “concern for individuals, and preoccu-
pied with such concepts as guilt, responsibility and obligation, as well as
diagnosis, intervention and treatment of an individual offender” (Feeley
and Simon 1994: 173). The new penology, however, is “concerned with
techniques for identifying, classifying and managing groups assorted by
levels of dangerousness. It takes crime for granted. It accepts deviance as
normal” (Feeley and Simon 1994: 173). Interventions are less directed at
determining responsibility or ensuring accountability, and more at man-
aging and regulating risky groups of people or places.
The impact of these trends on crime control and policing has been to
change the focus from state responsibility for preventing and correcting
criminal behavior to a system where crime control and prevention net-
works are responsible for identifying and managing risks. Public police
form one node of these networks, with private police, insurance com-
panies, regulatory agencies, communities, schools, and parents as other
nodes. These networks may exist within legislated frameworks, but are
often episodic and ad hoc. The prime concern of police and the criminal
justice system is less with the detection and rehabilitation of individual
offenders, and more with identifying and corralling risky groups – repeat
offenders, sex offenders, drug users, homeless or mentally ill people.
The new technologies involve systematic identification of target groups
or places (Eck and Weisburd 1995), and then new forms of surveil-
lance, preventive detention, and incapacitation via longer or mandatory
sentences.
It is now well documented that the rhetoric of market solutions and
privatization espoused by Thatcherist and Reaganist proponents of the
1980s–90s was not deregulation, but rather a new form of regulation
(Braithwaite 1999; 2000). For every privatized industry there was created
a new regulatory agency to supervise, monitor, and investigate the new
private players. But the new regulators differ from the old, state-centered
models. They rely on regulatory techniques ranging from voluntary and
194 Lorraine Mazerolle and Janet Ransley

enforced self-regulation through to selective use of the old command-

and-control techniques and sanctions. They work with industry bodies,
and increasingly as part of globalized regulatory networks (Braithwaite
and Drahos 2000). The practical effect has been a shift in regulatory
model, from state control to market models, from hierarchical systems of
command and control to responsive regulation. The new forms of gover-
nance actually require state control of the direction of regulation and risk
management, with many of the operational regulatory and compliance
functions then shifted out to the market, community, and other social
institutions.
The scope of regulation has been expanding significantly since the
1960s. In recent years traditional areas of regulation (such as health,
education, and taxation) have been joined by new forms of social regu-
lation (such as occupational health and safety, building codes, consumer
protection, and environmental regulation). In many of these areas, for-
mal regulation, or law, has been only one form of social control, with
self, peer, and professional regulation also being significant. Recogni-
tion of this plurality of regulatory methods, coupled with ideologically
inspired shrinkage of state activities, has led to a departure from reliance
on command and control as the only way of securing compliance with
regulation. Nowadays, the regulatory pyramid (Ayres and Braithwaite
1992) sets out a tiered ranking of techniques, beginning with persuasion
and self-regulation, progressing through professional discipline, adverse
publicity and fines, through to prosecution and the withdrawal of occupa-
tional licenses. Responsive regulation requires a logical working through
of these techniques, resulting in the use of the most coercive only in a
small number of cases of intransigence. Hence, regulation becomes a
layered web, with strands contributed by public regulatory agencies, pro-
fessional and community organizations and individuals, and international
organizations.
The impact of these methods on regulatory agencies is to trans-
form them from reactive, hierarchical command structures to problem-
oriented, team-based units focused on risk management (Sparrow 1994;
2000). The emphasis moves from after the event use of formal legal sanc-
tions, to cooperation, persuasion, and the creation of incentives for com-
pliance. The attraction for the regulated is the comfort that the “big stick”
of coercive sanctions will only be used as a last resort, and also that those
who are regulated will have some input into the rule-making and com-
pliance processes. The attraction for governments is also twofold – first,
persuasion and the other techniques are cheaper and give quicker results
than formal legal process (see Cheh 1998), but more importantly, they
help build an image of government as supportive of business, rather than
focused on bureaucracy and red tape.
 The case for third-party policing 195

These changes in societal context have far-reaching implications for

police and the organization of police work. One oft-cited effect of the
shift in regulation has been the movement away from state dominated
policing to the situation where most developed economies have more
private than state police (Shearing and Stenning 1987). As private secu-
rity guards replace police in public and private buildings, community
centers, even public spaces, and as private prison administration pro-
liferates, the role of the state increasingly becomes one of regulating
standards rather than actually controlling policing and criminal justice
functions. The end result is a “reconstitution of policing as a mechanism
of governance oriented to the management of conduct across civil soci-
ety, and the advent of a loosely coupled network of policing agencies”
(Loader 2000: 333–334, emphasis added) and a partial shift in the con-
trol of policing away from the state towards political subcenters (Shearing
1996).
Perhaps the most pervasive impact of the shift in governance is the
assumption now that the public police no longer (if they ever did) have a
monopoly over responding to and preventing crime, but are expected to
work in partnership with a range of other institutions, agencies, and indi-
viduals. There is no clear framework for these types of partnerships, but
rather a set of expectations that police will work cooperatively with their
partners (what we define as “third parties”) in identifying and respond-
ing to crime, in ways that are likely to vary from community to commu-
nity and problem to problem. We note that the notion of police part-
nerships is now entrenched in legislation in the United Kingdom with
the requirement for police and local authorities to cooperatively develop
crime reduction policies.
Police have become, as Ericson and Heggarty (1997) describe, the
brokers for these partnerships – coordinating, providing information and
resources, and responding to the risks. Risks are identified through the
analysis of statistical data and technologies such as computerized crime
mapping systems and utilized systematically by the police through man-
agement systems like Compstat (see Weisburd, Mastrofski, McNally et al.
2003). Interagency task forces and intervention teams with multiagency
membership all utilize these technologies for crime control purposes.
We argue that the proliferation of third-party policing is an intrin-
sic outcome of these transformations in governance and the rise of the
new regulatory state. Third-party policing is the result of global, regula-
tory processes and the accompanying pressures on the police to conform
to contemporary regulatory practice. It is no coincidence that cooper-
ation, risk management, problem identification and solving, and part-
nerships are the new primary foci for regulatory practice, as well as
policing.
196 Lorraine Mazerolle and Janet Ransley

Dimensions of third-party policing

Third-party policing is defined and distinguished from other models of
policing through the intrinsic links it has with transformations in gov-
ernance more generally and in regulatory trends more particularly. The
general form of third-party policing includes the police forming willing
(or unwilling) partnerships and using legal levers to control and prevent
crime. We describe the key dimensions of third-party policing below.

Purpose of action
We identify two primary purposes of third-party policing activities: crime
prevention or crime control. In crime prevention, the police seek to antic-
ipate crime problems and reduce the probability of an escalation of the
underlying conditions that may cause crime problems to develop. Third-
party policing that has crime prevention as its purpose of action operates
to control those underlying criminogenic influences that may (or may
not) lead to future crime problems. By contrast, third-party policing that
seeks to control existing crime problems explicitly aims to alter the rou-
tine behaviors of those parties that the police believe might have some
influence over the crime problem.

Initiators of third-party policing

In our analysis of third-party policing we focus on the public police as
the initiators of third-party policing. There are, however, a variety of col-
lectivities and individuals that have (or could) initiate similar activities.
We define these activities, however, as third-party crime control in order
to distinguish the activities of the public police as initiators from other
potential initiators such as prosecutors, individual citizens, private secu-
rity, community groups, and law enforcement agents in regulatory agen-
cies. Whilst we focus exclusively in this chapter on the public police as
the initiators of third-party policing, we recognize the plurality of poten-
tial initiators of similar activities and that different communities, states,
and countries will most likely emphasize different types of third-party
initiators.

Focal point
The focal point of third-party policing can be people, places, or situations
(see Mazerolle and Roehl 1998; Smith 1998). Sometimes third-party
policing efforts are directed specifically at categories of people such as
young people, gang members or drug dealers. To address some types
of crime problems, the focal point of third-party policing efforts might
 The case for third-party policing 197

be directed against specific places, more often than not places that have
been defined by the police as hot spots of crime. Drug dealing corners,
parks where young people hang out, and public malls are typically the
focal point of third-party policing activities that address specific places as
opposed to certain categories of people.
The third focal point of third-party policing activities includes situa-
tions that give rise to criminogenic activity. An example of a criminogenic
situation that gives rise to third-party policing responses include fights
occurring in entertainment precincts. In third-party policing licensees
are encouraged (or coerced) into partnerships that induce them to con-
trol the supply of alcohol and the off-site conduct of patron behavior
(see the Valley Alcohol Management Partnership, an interdepartmental
initiative in Brisbane, Queensland). In this third-party policing example,
the police draw on legal provisions to coerce third-party partners such as
government agencies (e.g., Department of Communities, Brisbane City
Council), statutory authorities (Liquor Licensing Authority), and busi-
nesses (local pubs, bars, and nightclubs) into signing onto an Alcohol
Management Accord. The accord governs the training of bartenders, the
ratio of trained versus untrained personnel working at any time, the supply
of alcohol to patrons, and the assistance provided by the local businesses
to better disperse patrons throughout the night.

Types of problems
Third-party policing can be directed against a broad range of crime and
quality of life problems (see Finn and Hylton 1994; National Crime
Prevention Council 1996). However, most examples and evaluations of
third-party policing comprise police efforts to control drug problems (see
Green 1996; Eck and Wartell 1998; Mazerolle, Kadleck, and Roehl 1998)
and disorderly behavior.
There is one main reason why third-party policing tends to proliferate
in efforts to control low-level, street types of crime activity. In the new
regulatory state, nodes for crime control are created not in a vacuum but
as the result of centralized state policymaking. A higher-order, complex
crime problem such as child sexual assault is the type of crime that grabs
newspaper headlines and challenges the state to create wide-ranging poli-
cies to guide intervention. The crime control partners (doctors, police,
educators, child safety officers) and legal levers (e.g., mandatory notifi-
cation) for these types of problems are usually initiated at a state level
and generally not by the police. As such, the police are just one partner
(albeit important) in the process, but not the initiator of the third-party
intervention.
198 Lorraine Mazerolle and Janet Ransley

Ultimate targets
The ultimate targets of third-party policing efforts are people involved in
deviant behavior. In theory, the ultimate targets of third-party policing
could include those persons engaged in any type of criminal behavior
including domestic violence, white-collar offending, street crime, and
drug dealing. In practice, however, the ultimate targets of third-party
policing are typically street level offenders. Young people (see White
1998), gang members, drug dealers (Green 1996), vandals, and petty
criminals typically feature as the ultimate targets of third-party policing.

Proximate targets, burden bearers and third parties

A key, defining feature of third-party policing is the presence of some type
of third person (or third collectivity or regulatory node) that is utilized
by the police in an effort to prevent or control crime. The list of poten-
tial third parties is extensive and can include property owners, parents,
bar owners, shop owners, local and state governments, insurance com-
panies, business owners, inspectors, and private security guards. Indeed,
any person or entity that is engaged by the police to take on some type
of role in controlling or preventing crime could potentially be identified
as a third-party or what Buerger and Mazerolle (1998) refer to as “prox-
imate targets” and what Mazerolle and Roehl (1998) have referred to as
“burden-bearers.” These are the people or entities that are coerced by the
police and who carry the burden for initiating some type of action that is
expected to alter the conditions that allow crime activity to grow or exist.
Proximate targets of third-party policing are often stakeholders or reg-
ulators that are identified by the police as being useful levers in con-
trolling a crime problem. Indeed, the roles in third-party policing can
change rapidly, they vary depending on the situation, sometimes recip-
rocal in nature and idiosyncratic to the problem at hand. Indeed, the
proximate targets of a third-party policing activity in one context may
become the ultimate targets of third-party policing in another context.
Moreover, cooperative police partners in one context might become hos-
tile “partners” in another context. We suggest that the dynamic nature
of third-party policing reflects the fluidity and chaotic nature of crime
prevention and crime control more generally.
There are many potential partners in third-party policing including
education authorities who prosecute the parents of truants, bar owners
who work with police to reduce street drunkenness, and property own-
ers who screen potential tenants and maintain the physical conditions of
their properties. However, a predominant group involved with police in
third-party policing networks comprises the regulatory authorities. These
 The case for third-party policing 199

are government agencies or officials with a function of regulating and

maintaining standards in some legal activity, such as housing, build-
ing, business, or industry. Typical regulatory officials who might become
involved in third-party policing include building, health and safety inspec-
tors, and environmental protection officers. These officials are attractive
to police because their functions are often accompanied by coercive pow-
ers to enter properties, inspect and search, issue closure orders, or take
other retaliatory action against people in breach of the regulatory scheme.
For police, partnering with such officials can act as a de facto extension
of their own powers, as well as increasing their potential weapons and
sanctions against people they suspect of involvement in criminal activity.
But it is important to remember that crime control and prevention are
not the primary aims of regulators, who instead have specific statutory
functions to fulfil.

Legal basis
In third-party policing, compliance is obtained through the threat, or
actual use of, some type of legal provision. As such, a key defining fea-
ture of third-party policing is that there must be some sort of legal basis
(statutes, delegated or subordinate legislation or regulations, contractual
relationships, torts laws) that shapes police coercive efforts to engage a
third-party to take on a crime prevention or crime control role. The most
common legal basis of third-party policing includes local, state, and fed-
eral statutes (including municipal ordinances and town by-laws), health
and safety codes, uniform building standards, and drug nuisance abate-
ment laws, and liquor licensing. We point out that the legal basis does
not necessarily need to be directly related to crime prevention or crime
control. Indeed, most third-party policing practices utilize laws and reg-
ulations that were not designed with crime control or crime prevention
in mind (e.g., Health and Safety codes, Uniform Building Standards).
For the vast majority of third-party policing activities, the legal basis
that provides the coercive power for police to gain the “cooperation” of
third parties derives from delegated legislation and obscure, non-criminal
sources.
The partnerships between police and third parties occur within a legal
framework that authorizes the conduct of the third-party – the building
inspector, local authority, licensing agency, or parent. This legal frame-
work establishes the source of authority of the third-party, the extent to
which they can partner with police, the contexts in which they can do that,
the types of action they can take against targets (criminogenic places and
individuals), and the limits of their legal ability to cooperate with, or be
coerced by, police.
200 Lorraine Mazerolle and Janet Ransley

In third-party policing, laws and legal mechanisms are directed at

willing or unwilling non-offending third parties, with the object of
facilitating or coercing them into helping to control the behavior of
offending ultimate targets. The types of law used can be criminal or civil,
and the distinction between these two categories is becoming increasingly
blurred (see also Cheh 1991). The laws or legal devices used may be
established specifically for the relevant crime control purpose, or may be
directed at some other issue but coopted, by police, community groups,
victims, or regulators, to achieve crime control or prevention goals.
Much of what we describe as third-party policing has arisen as an
unintended consequence of law, rather than as its specific object (notable
exceptions include Britain’s Crime and Disorder Act 1998, which specif-
ically mandates police networks). Third-party policing has arisen largely
in the space known as “the shadow of the law” – that is, it is influenced
by the law, uses the law where necessary, is limited in scope by the law,
but often occurs in practice as an extralegal activity.

Types of sanctions and penalties

Civil sanctions and remedies vary greatly including court-ordered repairs
of properties, fines, forfeiture of property or forced sales to meet fines
and penalties, eviction, padlocking or temporary closure (typically up to
a year) of a rented residential or commercial property, license restrictions
and/or suspensions, movement restrictions, lost income from restricted
hours and ultimately arrest and incarceration (see Mazerolle and Roehl
1998). Oftentimes, several civil remedies and sanctions may be initiated
simultaneously to solve one problem.

Tools and techniques

Dozens of examples can be provided to illustrate the processes by which
third parties are recruited and used by the police. Against the backdrop of
a legal foundation to force a third-party to cooperate, the police operate
on a continuum to engage third parties in their crime prevention or crime
control activities. At the more benign end of the spectrum, the police can
approach third parties and politely ask them to cooperate. The police
might consult with members of the community as well as local property
owners and ask them about ways that they see fit to control an existing
crime problem or help them to alter underlying conditions that the police
believe might lead to future crime problems. For these types of cases the
ultimate sanctions that might be unleashed on third parties most likely go
unnoticed. The police may themselves consciously utilize their persuasive
powers, yet be unconscious about the alternative methods of coercion
 The case for third-party policing 201

that they may resort to if the third-party target proves to be an unwilling

participant. At the more coercive end of the spectrum the police engage
third parties to participate in their crime control activities by threatening
or actually initiating actions that compel the third party to cooperate.
We point out that there are several stages in the forcible initiation of third
parties in taking a crime control role: the first stage may involve a building
services agency issuing citations to a property owner following building
inspections of their property (see Green 1996). The latter stages of this
most coercive practice involve the initiation of prosecutions against the
non-compliant landowner and ultimately court-forced compliance by the
third party.

Types of implementation
There are many different ways that the police implement third-party
policing practices including third-party policing within the context of
problem-oriented policing or situational crime prevention programs.
Problem-oriented policing provides the management infrastructure (see
Goldstein 1990) and step-wise approach to solving a crime problem (Eck
and Spelman 1987) and situational crime prevention offers the police
an appreciation for situational opportunities that might be exploited
using third-party policing tactics (Clarke 1992; 1995). A review of the
Goldstein awards from 1993 to 2003 reveals that about 50 percent of
problem-oriented policing submissions incorporated third-party policing
activities (see Mazerolle and Ransley 2005), yet just one third of third-
party policing occurs within a situational crime prevention or problem-
oriented policing framework.
The most common manifestation of third-party policing, however, is
the ad hoc utilization of third-party principles initiated by patrol officers.
These police are simply “flying by the seats of their pants,” there is no
script for them to follow, no police department policy that they are work-
ing within, and generally very little accountability for their actions.

Evaluation evidence
Measuring program performance and being able to say whether or not an
intervention works is the bread and butter of sound policy decisionmak-
ing. We use performance measurement for “evidenced-based policymak-
ing” or making policies based on the outcomes of well-designed and exe-
cuted evaluations. In the policing arena, Sherman and colleagues (1997)
draw analogies with “evidence-based medicine” and defines evidenced-
based policing as the use of the best available research on the outcomes
of police work to implement guidelines and evaluate agencies, units, and
Table 10.1 Summary of third-party policing evaluation evidence

Studies where
an effect size
All Studies was calculated
Most common
N percent N percent Effect size outcomes third-party from Most common type
77 100 12 100 (N = 23) all 77 studies of legal lever

1. Drugs 21 27.3 8 66.7 13/18 (72 percent) Residential and Municipal ordinances
outcomes were desirable commercial property (e.g., building & health
owners & safety codes, nuisance
abatement)
2. Violent crime 21 27.3 2 16.7 2/2 (100 percent) Licenced premises Code of practice with
outcomes were desirable owners restrictions on serving of
alcohol
3. Places 15 19.4 0 0 NA Local councils Municipal ordinances
4. Young people 11 14 1 8.3 1/1 (100 percent) Parents Curfews
outcome was desirable
5. Property crime 9 12 1 8.3 2/2 (100 percent) Business owners Theft prevention policies
outcomes were undesirable

Source: Mazerolle and Ransley (2005).

 The case for third-party policing 203

officers. In short, evidenced-based policing “uses the best evidence to

shape the best practice” (Sherman et al. 1997: 4).
Mazerolle and Ransley (2005) conducted a search of the interna-
tional literature to uncover evaluations of police tactics involving third-
party policing. We categorized the studies into five groups: the use of
third-party policing in controlling drugs, violent crimes, property crime,
youth problems and crimes at criminogenic places. Our search strate-
gies resulted in the identification of seventy-seven studies that included
an evaluation component and also involved a third-party policing
tactic. Of these, only twelve studies included sufficient data to calcu-
late effect sizes (see Table 10.1). These twelve studies generated twenty-
three effect sizes (odds ratios and standardized mean differences) across
a variety of outcome measures (calls for service, arrests, field con-
tacts, observations). The heterogeneous units of analysis and the variety
of ways researchers had operationalized their outcome measures pre-
cluded calculation of mean effect sizes. As such, the individual out-
comes were examined independently (see Mazerolle and Ransley 2005).
Table 10.1 summarizes the evaluation evidence.
As the table shows, it appears that third-party policing is an effective
mechanism to control drug problems. It is likely that it is an effective
strategy for controlling violent crime and for dealing with young peo-
ple. Whilst the number and quality of the reported outcomes limits our
assessment of the effectiveness of third-party policing, Table 10.1 also
indicates that third-party policing is somewhat ineffective at controlling
property crime problems. Our review suggests that the majority of evi-
dence collated about third-party policing involves property owners (com-
mercial property owners in particular) as third parties. We were also able
to uncover limited, yet important, evidence that identifies parents, local
councils, housing authorities, and victims as third parties.
Despite our systematic attempts to uncover the evidence surrounding
third-party policing tactics, limited evaluations of third-party policing
initiatives exist in the research literature and very little systematic effort
has been expended to document, collate, and consolidate the third-party
policing research evidence. We suggest that there is a great need to con-
tinue to evaluate, document, and assess the effectiveness of third-party
policing.

Conclusions
In this chapter we provided a synopsis of third-party policing and dis-
cussed the regulatory context of third-party policing. We described how
third-party policing is part of a general transformation in contemporary
society that has seen a movement from state sovereignty and control to
204 Lorraine Mazerolle and Janet Ransley

networks of power. We argued that third-party policing is a manifestation

of these broader societal trends. We have also discussed what we know
about the effectiveness of third-party policing in preventing or responding
to crime.
We argue that the future of the third-party policing approach is rein-
forced by a number of factors: first, third-party policing is not a bottom-
up, grassroots innovation in the way that problem-oriented policing has
been articulated and developed (see Goldstein 2003). Rather, third-party
policing is part of a broader transformation in regulation. As such, the
partnership approach is likely to be foisted upon the police in the future
not from within, but rather from external forces. An example of this
is Britain’s Crime and Disorder Act 1998, which specifically mandates
police networks. We are likely to see more of this type of top-down
approach in the future. Second, the future of third-party policing is rein-
forced by early evidence that suggests it is a successful way of handling at
least street level crime problems. Obviously, success breeds success and
the more a strategy is found to reduce a problem, then the more likely it
is to be used again.
But to examine the context, mechanics and effectiveness of third-party
policing is to consider only half the equation – we also need to ask about
the side effects of third-party policing, intentional and unintentional, pos-
itive and negative, on the partners who work with police, as well as on
other groups in the community, on the proximate and ultimate targets,
and even on the police organizations themselves (see Meares in this vol-
ume). Some of the questions raised in third-party policing include: How
does third-party policing affect the community in which it is practiced?
How equitably does it affect different communities, both internally and
in comparison to other communities? Just as importantly, who is held
accountable for the outcomes and impacts of third-party policing, and
how? Are traditional police accountability mechanisms adaptable to take
account of this new way of doing business, or is there a need for new
mechanisms to be developed? In short, is third-party policing an ethical
policing practice?

  
Many of the arguments presented in this chapter are more fully developed in
Mazerolle and Ransley (2006).



Ayres, I. and Braithwaite, J. (1992). Responsive regulation: Transcending the dereg-

ulation debate. New York: Oxford University Press.
 The case for third-party policing 205

Braithwaite, J. (1999). Accountability and governance under the new regulatory

state. Australian Journal of Public Administration, 58 (1), 90–97.
(2000). The new regulatory state and the transformation of criminology. British
Journal of Criminology, 40, 222–238.
Braithwaite, J. and Drahos, P. (2000). Global business regulation. Melbourne:
Cambridge University Press.
Buerger, M. E. and Mazerolle, L. G. (1998). Third-party policing: A theoretical
analysis of an emerging trend. Justice Quarterly, 15 (2), 301–328.
Cheh, M. (1991). Constitutional limits on using civil remedies to achieve crim-
inal law objectives: Understanding and transcending the criminal-civil law
distinction. Hastings Law Journal, 42, 1325–1413.
(1998). Civil remedies to control crime: Legal issues and constitutional chal-
lenges. In L. G. Mazerolle and J. Roehl (eds.), Civil remedies and crime
prevention: Crime prevention studies, 9 (pp. 45–66). New York: Criminal Jus-
tice Press.
Clarke, R. V. (1992). Situational crime prevention: Successful case studies. New York:
Harrow and Heston.
(1995). Situational crime prevention. In M. Tonry and D. P. Farrington (eds.),
Building a safer society: Strategic approaches to crime prevention. Crime and
justice, 19 (pp. 91–150). Chicago: University of Chicago Press.
Eck, J. E. (1994). Drug places: Drug dealer choice and the spatial structure of
illicit drug markets. Unpublished doctoral thesis, University of Maryland.
Eck, J. E. and Spelman, W. (1987). Problem solving: Problem-oriented policing
in Newport News. Washington, DC: Police Executive Research Forum and
National Institute of Justice.
Eck, J. E. and Wartell, J. (1998). Improving the management of rental properties
with drug problems. In L. Green Mazerolle and J. Roehl (eds.), Civil remedies
and crime prevention: Crime prevention studies, 9 (pp. 161–83). Monsey, NY:
Criminal Justice Press.
Eck, J. E. and Weisburd, D. (1995). Crime prevention studies: Crime and place, 4.
Monsey, NY: Criminal Justice Press.
Ericson, R. V. and Haggerty, K. D. (1997). Policing the risk society. Toronto:
Toronto University Press.
Feeley, M. and Simon, J. (1994). Actuarial justice: The emerging new criminal
law. In D. Nelken (ed.), The futures of criminology (pp. 173–201). London:
Sage Publications.
Felson, M. (1994). Crime and everyday life: Insights and implications for society.
Thousand Oaks, CA: Pine Forge Press.
Finn, P. and Hylton, M. O. (1994). Using civil remedies for criminal behavior: Ratio-
nal case studies and constitutional issues. Washington, DC: National Institute
of Justice, US Department of Justice.
Garland, D. (1996). The limits of the sovereign state. British Journal of Criminol-
ogy, 36, 445–471.
(1997). “Governmentality” and the problem of crime: Foucault, criminology,
sociology. Theoretical Criminology, 1 (2), 173–214.
Goldstein, H. (1990). Problem-oriented policing. New York: McGraw-Hill.
(2003). On further developing problem-oriented policing: The most critical
need, the major impediments and a proposal. In J. Knutsson (ed.), Crime
prevention studies, 15. Monsey, NY: Criminal Justice Press.
206 Lorraine Mazerolle and Janet Ransley

Green, L. (1996). Policing places with drug problems. Thousand Oaks, CA: Sage
Publications.
Loader, I. (2000). Plural policing and democratic governance. Social and legal
studies, 9 (3), 323–345.
Mazerolle, L. G. and Ransley, J. (2006). Third-party policing. Cambridge: Cam-
bridge University Press.
Mazerolle, L. G. and Roehl, J. (1998). Civil remedies and crime prevention: An
introduction. In L. Green Mazerolle and J. Roehl (eds.), Civil remedies and
crime prevention: Crime prevention studies, 9 (pp. 1–20). Monsey, NY: Criminal
Justice Press.
Mazerolle, L.G., Kadleck, C., and Roehl, J. (1998). Controlling drug and disor-
der problems: The role of place managers. Criminology, 36, 371–404.
National Crime Prevention Council. (1996). New ways of working with local laws
to prevent crime. Washington, DC: National Crime Prevention Council.
O’Malley, P. (1992). Risk, power and crime prevention. Economy and Society, 21
(3), 252–275.
(2000). Risk, crime and prudentialism revisited. In K. Stenson and R. Sullivan
(eds.), Risk, crime and justice: The politics of crime control in liberal democracies
(ch. 5). London: Willan.
Scott, M. (2000). Problem-oriented policing: Reflections on the first 20 years.
Washington, DC: Office of Community Oriented Policing Services, US
Department of Justice.
Shearing, C. (1996). Reinventing policing: Policing as governance. In O. Marenin
(ed.), Policing change, changing police: International perspectives. Current Issues
in Criminal Justice, 4 (pp. 285–308). New York: Garland.
Shearing, C. and Stenning, P. (1987). Private policing. Thousand Oaks, CA: Sage
Publications.
Sherman, L.W., Gottfredson, G., MacKenzie, D., Eck, J., Reuter, P., and
Bushway, S. (1997). Preventing crime: What works, what doesn’t, what’s
promising: A report to the United State Congress. College Park: Department
of Criminology and Criminal Justice, University of Maryland.
Smith, M. (1998). Regulating opportunities: Multiple roles for civil remedies in
situational crime prevention. In L. Green Mazerolle and J. Roehl (eds.), Civil
remedies and crime prevention: Crime prevention studies, 9 (pp. 67–88). Monsey,
NY: Criminal Justice Press.
Sparrow, M. (1994). Imposing duties: Government’s changing approach to compli-
ance. Westport: Praeger.
Sparrow, M. (2000). The regulatory craft: Controlling risks, solving problems, and
managing compliance. Washington, DC: Brookings Press.
Weisburd D., Mastrofski, S. D., McNally, A. M., Greenspan, R., and Willis, J. J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2 (3), 421–456.
White, R. (1998). Curtailing youth: A critique of coercive crime prevention. In
L. Green Mazerolle and J. Roehl (eds.), Civil remedies and crime prevention:
Crime prevention studies, 9 (pp. 117–140). Monsey, NY: Criminal Justice
Press.
11 Critic
Third-party policing: a critical view

Tracey L. Meares

According to Lorraine Mazerolle and Janet Ransley (2003), “third-party

policing” describes police efforts to persuade or to coerce third parties,
such as landlords, parents, local government regulators, and business
owners to take on some responsibility for preventing crime or reducing
crime problems. Obviously this definition seeks to distinguish policing
directed at those who are and who might be criminal offenders from
policing efforts directed at non-offending “others.” Thus, the definition
emphasizes the affinities that third-party policing has with other forms
of civil regulation. Examples of such regulation abound. In an effort to
ensure that corporations do not defraud stockholders, regulators place
constraints, both civil and criminal, on accountants and lawyers. In an
effort to make sure that employers do not violate civil rights laws, legisla-
tors have structured statutes so that violators pay plaintiffs’ attorneys fees
should the plaintiffs win. In this way, plaintiffs are persuaded to become
“private attorneys general” helping public officials to enforce the law. In
an effort to ensure that athletes do not take illegal drugs, the National
Football League requires teams to complete random urine tests of players.
The federal government is now pressuring the Major League Baseball to
do the same. In this way, the sports leagues become third-party enforcers
of laws prohibiting the use of certain drugs.
Given the pervasive forms of such regulation today, that such “third-
party” efforts are becoming common in the enterprise of street crime
control should hardly be surprising. In fact, we can redescribe even those
efforts typically conceptualized as directed primarily at offenders in terms
of third-party controls. One common consequence of criminal offending
is a sanction in the form of a prison sentence or fine. However, it is silly
to think that only the offender suffers the opprobrium that accompa-
nies such sanctions. Families of the offender, as well as friends, can also
suffer (e.g., Braman 2004). In a world in which the costs associated with
formal punishment extend beyond the actual offender to non-offenders,
we should expect such third parties to engage in efforts to persuade, or
even informally coerce, potential offenders to refrain from crime. And,

207
208 Tracey L. Meares

social theorists have long explained how informal norms combine with
formal sanctions to effect deterrence.1 When informal social controls are
supported by legal sanctions, social costs are made more salient so that
behavior is better regulated.
Importantly, informal controls are not relevant simply when they sup-
port formal legal sanctions. Informal controls also involve the normative
processes and ethics of social interaction that regulate everyday social
life, as well as the mobilization of community that occurs in response to
problem behaviors (Doyle and Luckenbill 1991). Thus, informal social
controls are effective in several ways: inhibition of problem behaviors,
facilitation of conformity, and restraint of social deviance once it appears.
The key is to see that that evaluation of the deterrent effect of a policy can-
not depend simply on the likely impact of formal punishment, but, rather,
must also include some kind of assessment of the reciprocity between legal
and social controls. It is not enough to ask whether a potential offender
will be persuaded not to offend by assessing how he will compare the costs
of potential punishment against the benefit of engaging in the crime. We
also need to know about the potential offender’s community context and
social role to be able to begin to make assessments regarding the potential
effectiveness of a planned formal legal sanction.
My task here is not to bring all policing efforts under the potentially
commodious umbrella of third-party policing, however. While it should
be clear that much of traditional criminal law enforcement and policing
can be described as “third-party policing,” in this volume the meaning
of the term is more limited. Contributors to this volume have pinpointed
third-party policing as those relatively recent efforts by many policing
agencies that recognize that much social control is exercised by institu-
tions other than the police and that crime can be managed through agen-
cies besides those concerned primarily with dispensing criminal justice,
such as the police. Specific examples include nuisance abatement, anti-
gang loitering laws, and public housing eviction policies among others
(Mazerolle and Ransley in this volume).
The primary purpose of this chapter is to review the arguments made
by the critics of third-party policing efforts. Although the concerns of
most of these critics are based in constitutional law, some critics make
a special effort to explicate the potentially destructive racial dynamics of
some third-party policing strategies. These arguments target the distri-
butional effects of exactly which groups bear the burden of third-party
law enforcement. After reviewing these arguments, this chapter will con-
clude by attempting to address the costs and benefits of third-party polic-
ing strategies with special emphasis on the race-based critiques of these
approaches.
 Third-party policing 209

Civil libertarian critique

One basic criticism of government targeting of third-party non-offenders
to persuade or informally coerce the so-called “real offenders” to refrain
from lawbreaking is that by targeting third parties the government uses
its power to restrict the liberty of those who “really aren’t doing anything
wrong.” The essence of civil libertarianism is that coercive government
power should be deployed only in those instances in which it is necessary
to protect individuals from the use of force or fraud by others. Under this
view, a threat to levy a civil fine against a landlord in order to persuade her
to better scrutinize and identify potential drug-dealing renters is patently
impermissible. According to libertarian beliefs, state power ought to be
trained upon the drug dealer, not the landlord, as the dealers are the true
wrongdoers.
Private groups, legislators, and other governmental officials have an
answer to this foundational criticism. Criminal offenders are sometimes
difficult to locate and bring to justice through the conventional criminal
justice apparatus. In fact, criminal liability itself has, over time, expanded
to address this problem. We punish offenders not only for the crime of
robbery, but also for the offense of attempted robbery. Further, prosecu-
tors often utilize offenses such as conspiracy and solicitation in order to
ferret out and punish crime before it takes place. Neither the doctrine
of conspiracy nor of solicitation actually requires that the prohibited act
occur. The doctrines require simply that a person agrees with another to
commit a criminal offense (conspiracy),2 or request that another engage
in conduct that constitutes a crime (solicitation).3 Additionally, we pun-
ish those who assist others in criminal offending under the doctrine of
accomplice liability even when the accomplice has not himself or her-
self committed any act that most would deem criminal. According to the
doctrine of accomplice liability, individuals can be punished when they
make a decision to further the criminal act of another and engage in an
act that furthers the principal’s criminal act. In addition to all of this, civil
remedies have long been used to hold offenders to account outside of the
traditional criminal justice context.
Prophylactic measures, combined with civil remedies, provide law
enforcers with a wider range of alternatives to crime control that are less
costly than traditional criminal justice approaches. Third-party polic-
ing, then, is simply a small extension of this trend. There is, however,
one basic difference between third-party policing efforts and prophylac-
tic criminal and civil remedies directed at wrongdoers: those co-opted
into enforcement efforts are not themselves wrongdoers in the usual
sense.
210 Tracey L. Meares

This basic difference is a key difference for critics of third-party policing

concerned with civil liberties. Such critics might point to Bennis v. Michi-
gan as a paradigmatic case of third-party policing gone awry. In Ben-
nis, a car was seized from Mr. Bennis after he had been convicted of
committing an indecent act with a prostitute while the two were in the
car. Mrs. Bennis claimed that her interest in the car could not be for-
feit to the State of Michigan because she was completely unaware of her
husband’s activities. The United States Supreme Court, in a plurality
opinion, nonetheless ruled against her. As a result of the Court’s deci-
sion, the State of Michigan was allowed to keep the Bennis family 1977
Pontiac, which was paid for primarily with money Mrs. Bennis earned
from babysitting.
The Bennis plurality stated that there was a long history of permitting
forfeiture against innocent owners in order to prevent further use of illicit
property. The reasoning is obvious: if an innocent owner knows that there
is a risk that her property might be forfeit if someone else uses the property
for illegal purposes, she likely will take greater care than she otherwise
would to ensure that her property is not so used. That is, she will become
a third-party police officer.4
The problem with this reasoning, according to civil libertarians, is not
whether the procedure works or not; rather, the problem is that the for-
feiture process is fundamentally unfair under basic principles of due pro-
cess of law. For one thing, the deterrence rationale that applies to Mrs.
Bennis potentially implies that Michigan can punish Mrs. Bennis for her
husband’s crime. Such punishment is clearly inconsistent with due pro-
cess. The Supreme Court has recognized the due process right of an
innocent person to be free of punishment.5 Mrs. Bennis was not tried
and convicted before her punishment – the forfeiture of her interest in
the family car – was meted out by the state. Indeed, she could not pos-
sibly have been convicted for the crime her husband committed. Mrs.
Bennis had no awareness of her husband’s activities, so she possessed
no criminal intent to engage in the offense or to further it in any way.
Moreover, Mrs. Bennis had engaged in no act that could be considered
criminal. While her husband was involved with a prostitute, Mrs. Bennis
was simply waiting for her husband to come home for dinner.
Even if one were convinced that Michigan’s actions in this case were not
punitive, but instead were merely regulatory,6 one might still conclude
that the state’s action offended due process. Deterring an activity such as
prostitution is, of course, a legitimate aim of a state’s police powers; how-
ever, due process principles require that government respect the rights
of property owners and afford them procedural protections under civil
law as well as criminal law. The Constitution’s text is clear: government
 Third-party policing 211

may not deprive individuals of life, liberty, or property without providing

due process of law. Unfortunately for Mrs. Bennis and for others simi-
larly situated, the procedural protections required to satisfy constitutional
minimums are minimal.7 Still, the fact that a particular procedure meets
the minimum requirements of procedural due process does not mean that
critics are without recourse. Critics often contend that an enforcement
operation like that in Bennis is inconsistent with substantive due process
even when the operation meets the constitutional requirements of proce-
dural due process. As Mary Cheh (1994: 25) has neatly summarized:
Substantive due process analysis ordinarily proceeds along two tracks. Almost all
laws touching social and economic matters are judged under a lenient rational-
basis test, while laws that interfere with certain intimate and personal rights, such
as child rearing, marriage and divorce, and use of contraceptives, are judged under
a vigorous strict-scrutiny test. The rational basis test is easily satisfied. Ordinary
social and economic regulation is presumed constitutional, and so long as the law
serves any permissible police power objective, it will be upheld. Legislatures are
free to decide, without court interference, how they will tax, regulate, and control
behavior.

Under these analyses a litigant must prove either that the forfeiture is so
unreasonable as to fail rational basis, or that there is a protected interest
at issue such that the more rigorous strict scrutiny test applies. Both of
these types of arguments fail more often than they succeed.
This review of Bennis and related constitutional doctrines might sug-
gest that it is futile for civil libertarians to adopt these constitutional
arguments to block third-party policing efforts. Such a conclusion would
be mistaken. Consider litigation against anti-gang loitering ordinances
and curfews, which some consider third-party policing tools, and which
have become increasingly popular in municipalities. Loitering and cur-
few ordinances are designed to keep teens from congregating at night
to attract the ire of rival gang members and pick fights, or standing on
corners to help friends hidden in alleys to sell drugs. By enforcing these
laws, it is thought that police can help adults simply by acting as additional
eyes and ears in the neighborhood. Some of these laws also are designed
to make parents more accountable by penalizing parents whose children
violate the ordinance. These laws have been met with resistance and con-
stitutional challenge, and in many cases, the civil libertarian critics have
been successful. Chicago’s first anti-gang loitering law was deemed
unconstitutionally vague, a fate shared by numerous “loitering with
intent” statutes around the country. Curfews in Washington, DC
(Hutchins v. District of Columbia), San Diego (Numez v. City of San Diego),
and other cities (e.g., City of Maquoketo v. Russell) have likewise been
deemed to abridge the due process rights of teens and their parents.
212 Tracey L. Meares

Ultimately the civil libertarian critique of government agencies involved

in third-party policing boils down to one basic point: a concern with the
scope of accountable government power. Civil libertarian critics believe,
with some justification, that the greater the government’s power, the more
difficult it is to control; therefore, we all are better off if government power
is limited as much as possible. One suspects that these critics would not
be so concerned if they could be persuaded government power in these
areas was utilized in a rational and transparent manner. Unfortunately, to
the extent that preferred strategies utilize civil justice and political mecha-
nisms as controls, as opposed to criminal justice mechanisms, it becomes
that much more difficult to keep track of government activity and to
limit it. Police discretion, always a powerful tool even when confined to
criminal justice processing, potentially grows to almost unrecognizable
proportions once let out of the criminal justice cage.

Racial equity critique

Some critics of third-party policing strategies have complained about a
specific problem that could be considered a subset of the critique detailed
above. Namely, these critics are worried that as third-party policing strate-
gies loosen the reins on governmental discretion regarding whom to
engage as the law is enforced, such discretion inevitably will be exer-
cised more often against poor and minority-race citizens than against
those who are not poor and/or of minority race.
Consider as an example Professor Dorothy Roberts’s (1999) race-
based critique of an anti-gang loitering law adopted by the City of Chicago
in 1992. The ordinance exhibited unique third-party policing features.
The Chicago City Council passed the ordinance to restrict gang-related
congregations in public ways (Chicago, Il., Code § 8-4-015 1992). The
ordinance was designed to respond to the grievances of citizens concerned
about commonly occurring criminal street gang activity in their neighbor-
hoods, such as drive-by shootings, fighting, and open-air drug dealing.
By loitering in alleyway entrances and on street corners, drug dealers
both solicited business and warned hidden compatriots of police patrols.
The ordinance empowered designated police officers in specified areas
to approach groups of three or more people loitering “with no appar-
ent purpose” provided that those officers had reasonable cause to believe
that at least one member of the group was a street gang member and ask the
group to move along. If the group refused, then the officer was entitled to
arrest the group. The third-party policing aspect is clear. Even those indi-
viduals whom the police had no reason to suspect were gang members,
were potentially subject to arrest merely because of the company those
 Third-party policing 213

individuals chose to keep (and where they chose to keep the company).
By subjecting non-gang members to liability, municipal regulators incen-
tivized them to help police constrain the behavior of their gang-involved
compatriots.
Professor Roberts’s (1999: 775) assessment of the law is blunt: “expan-
sive and ambiguous allocations of police discretion are likely to unjustly
burden members of unpopular or minority groups.” In support of this
claim, Roberts (1999: 785) points to the fact that in 1995 46.4 percent
of people arrested for vagrancy across US cities were black even though
blacks made up only 13 percent of the nation’s population. No doubt a
critic such as Roberts would also take data compiled by Justin Ready,
Lorraine Green Mazerolle, and Elyse Revere (1998), demonstrating that
91 percent of the tenants evicted from public housing developments in
Jersey City as part of a civil remedy program to achieve greater level of
order in those developments were black, as evidence of the unjust burden
suffered by minorities as third-party policing strategies are implemented
(Ready, Mazerolle, and Revere 1998).
An assumption that minorities are unfairly burdened by third-party
policing strategies that are designed to give law enforcers more flexibility
to address crime and disorder is based upon a fundamental premise –
the weak political power of minority groups. Consider a famous Yale
Law Journal article written in 1960 by William O. Douglas. In the piece
Douglas railed against the argument that anti-loitering laws were beyond
constitutional reproach. It was naı̈ve to trust contemporary communi-
ties to apply such laws evenhandedly, Douglas asserted, because those
arrested under such laws typically came “from minority groups” with
insufficient political clout “to protect themselves” and without “the pres-
tige to prevent an easy laying-on of hands by the police” (Douglas 1960:
13). When the court did ultimately deem traditionally worded loiter-
ing laws unconstitutionally vague in Papachristou v. City of Jacksonville,
Douglas wrote the court’s opinion.
Unlike the more general strategy of civil libertarians, advocates of con-
stitutional arguments designed to protect minority groups from over-
weening police power have achieved more success in the Supreme Court
and lower courts. In the 1960s the prevailing sense of the court was that
the coercive incidence of law enforcement, in both the North and South,
was concentrated most heavily on minority citizens, who by virtue of
their exclusion from the political process had no say about whether those
policies were just. The result was the systematic devaluation of both the
liberty of individual minority citizens and the well-being of minorities as
a group. By insinuating courts deeply into the process of criminal-law
enforcement, the federal constitutionalization of state police procedures
214 Tracey L. Meares

was intended to correct this imbalance (Klarman 1991). Similarly, in the

political context of the 1960s, law enforcement officials were accountable
only to representatives of the white majority. Indeed, for precisely this
reason, the police predictably used their discretion to harass and repress
minorities. Insisting that law-enforcement authority be exercised accord-
ing to hyper-precise rules was a device for impeding the responsiveness
of law enforcers to the demands of racist white political establishments.
Such rules also made it much easier for courts to detect and punish
racially motivated abuses of authority.
Thus, by pointing to racial imbalances in the enforcement of loitering
laws, curfews, public housing evictions, and nuisance abatement laws,
modern critics intentionally draw upon the particularized racial history
of such strategies – and the courts’ pointed disfavor of them. But, one
question is whether the costs of such strategies, even considering the
specific race-based arguments offered by critics such as Dorothy Roberts,
outweigh their benefits.

Costs and benefits

As noted above, police departments across the country have turned to
third-party policing strategies as seemingly low-cost alternatives to tradi-
tional criminal justice apparatus to prevent crime. If such strategies are
successful, the benefits are clear: reduced crime for (presumably) less
money, as civil justice mechanisms typically are less costly than criminal
justice ones. There is a dearth of available evaluations, but the initial stud-
ies are promising. For example, Jeffrey Grogger (2002) has found through
an empirical analysis of civil injunctions designed to reduce gang vio-
lence that in the first year after the injunctions are imposed violent crime
falls by 5 to 10 percent. Moreover, in a recent edited volume, Lorraine
Mazerolle and her colleagues (Mazerolle and Roehl 1998) collect stud-
ies of various civil remedy programs designed to control drug problems.
Most of the studies contained in the volume point to the effectiveness of
such programs at reducing levels of targeted drug offenses.
The case for third-party policing, then, might be made on this evidence.
Note, however, that many third-party policing strategies are place-based
programs. And, if such place-based programs caused crime reduction in
the targeted areas while simply diverting crime elsewhere, that would be
an obvious cost. Geographic displacement of crime is a serious concern
when place-oriented interventions are employed (Committee to Review
Research 2004: 240). Interestingly, the small body of research pertaining
directly to this issue with respect to third-party policing indicates the
opposite – a diffusion of crime control benefits (e.g., Mazerolle and Roehl
 Third-party policing 215

1998) to areas contiguous to the places targeted for third-party policing

interventions.
Still, the fact that a few studies indicate some benefit through reduc-
tion in crime does not resolve the accounting for costs and benefits in
favor of the benefit side of the ledger. While promising, the available
studies do not point to sizable results. Moreover, to be confident about
a cost-benefit analysis, one must ask in this context whether such crime
reduction benefits would be achieved through more traditional policing
methods – methods that are less likely to raise the hackles of civil liberties
proponents (Caulkins 1998). The results of such an analysis are difficult
to attain to say the least, but without such results, police departments and
other governmental entities ought to tread lightly before embarking on
strategies that present the kinds of grave risks to civil liberties discussed
earlier. Critics of third-party policing have a point when they assert that
third-party policing interventions cannot be justified on crime reduction,
even if substantial, alone.
Presumably the critics whose arguments are sound in racial equity con-
cerns would agree with this assessment. Urban minority residents face
more crime than other non-minority groups, so such residents would
likely benefit in the form of reduced crime, nuisance, and the like in their
communities. However, such residents may not believe that reductions
in say, observable open-air drug selling8 are worth the costs in terms of
civil liberties incursions. The case for benefits needs more.
An additional argument that can be made on the benefits side is that
third-party strategies potentially can help to change the social dynamics of
neighborhoods in ways that promote crime-reducing norms. One way to
do this is to point out that traditional criminal law enforcement methods
potentially can impede a community’s ability to resist and reduce crime.
Traditional law enforcement methods rely on criminal justice processing
and punishment – especially incarceration – to achieve crime reduction.
Scholars have articulated theories describing how mass incarceration
concentrated at the community or neighborhood level could hamper insti-
tutions of informal social control (Nagin 1998; Rose and Clear 1998).
Drawing on Shaw and McKay’s (1969) foundational work on the rela-
tionship between the social disorganization of neighborhoods and the
persistence of high crime rates at the community level, these scholars
have focused on various social processes, including (1) the prevalence,
strength, and interdependence of social networks; (2) the extent of col-
lective supervision by neighborhood residents and the level of personal
responsibility they assume for addressing neighborhood problems; and
(3) the rate of resident participation in voluntary and formal organiza-
tions (Sampson and Wilson 1995; Wilson 1996; Rose and Clear 1998).
216 Tracey L. Meares

The hypothesis is straightforward: When the processes of community

social organization are prevalent and strong, crime and delinquency
should be less prevalent, and vice versa. Burgeoning research suggests
that mass incarceration following from traditional law enforcement meth-
ods inhibits social processes that support crime reduction and prevention
(Lynch and Sabol 2004); thus, law enforcement that does not rely heavily
on incarcerative approaches, like third-party policing strategies, may be
less harmful than traditional law enforcement approaches to fragile urban
poor community structures and may indeed support those structures in
ways that lead to less crime.
Specifically, civil remedies can be democratizing in a sense. Such reme-
dies typically are not reactive; rather, they are proactive. Neighborhood
residents who have been (or feel they have been) underserved by policing
organizations historically often can turn to civil remedies as an alter-
native to traditional strategies. Indeed, an argument can be made that
third-party strategies are especially empowering to residents of disadvan-
taged neighborhoods in a way that traditional policing often is not. That
is, state-sponsored strategies that encourage individuals to work with one
another can help to sustain a healthy social organization dynamic that can
be harnessed in favor of crime reduction and resistance (Meares 2002).
Such forces can in themselves be supportive of neighborhood collective
efficacy, which itself is associated with lower crime at the neighborhood
level (Sampson, Raudenbush, and Earls 1997).
Finally, there may be normative concerns that can be arrayed against
the civil libertarians’ own arguments, whether those arguments are
grounded in individual libertarian concerns or take on a more group-
based, race-specific character. It is important to pay attention to where
the groundswell of support for many third-party policing strategies lies. In
no small number of cases, the support for these strategies comes from the
residents of high-crime neighborhoods who are themselves often mem-
bers of minority groups – the very people who face a heightened risk of
criminal victimization and who live with the destructive impact of crime
on the economic and social life of their communities and who feel the
pinch of these laws in a meaningful way.
There is often little room in either the general civil libertarian critique
or the more specifically racialized critique for the voice of this group of
people, but it should be clear that as a normative matter their voices
and votes ought to count for something. Of course, deciding that one
ought to listen to the people who are most affected by third-party policing
strategies does not guarantee a “right answer” to balancing issues inherent
in these debates. There is no perspective-free way to determine whether
the general structure of third-party approaches violates the Constitution.
 Third-party policing 217

Some individual or set of individuals, judging the question in the light of

her own experiences and values, must decide whether particular policies
embody a reasonable balance between liberty and order. The question is
who should decide.
To illustrate, consider an exchange from the City Council hearings on
the Chicago gang-loitering ordinance. At the hearings, dozens of inner-
city residents – from church leaders, to representatives of local neighbor-
hood associations, to ordinary citizens – testified in favor of the proposed
law. Harvey Grossman, Illinois director of the ACLU, testified against it:

I am a lawyer, and I spend a great deal of time doing nothing more than reviewing
ordinances and statutes, and it turns into a little bit of a long exam game. . . . We
pick apart the statute. We focus on [a] word or [a] phrase, and we try to say why
that phrase might or might not be constitutional. (Transcript of Meeting before
Chicago City Council Committee on Police and Fire 107. May, 18, 1992)

He then proceeded to “pick apart” the gang-loitering ordinance, demon-

strating the tension between it and various judicial precedents.
Alderman William Beavers, a council member who represented a poor
and high-crime minority district on Chicago’s south side, objected to this
bookish conception of how to appraise the constitutionality of the law. “I
don’t know if you are attuned to what’s going on in these neighborhoods,”
he told Grossman. “Maybe I need to take you out there and show you
what’s really going on” (ibid.: 119). Grossman replied that that wouldn’t
be necessary: “I think our ability to come together and to try to resolve
issues [like this] really doesn’t . . . depend on if I see what’s happening in
your neighborhood or you see what’s happening in my neighborhood.”
Rather, what it does depend on is “empathy or ability to understand
what’s happening to other people and, two, some commitment, some
intellectual integrity and some commitment to principle. And the prin-
ciple that I am suggesting to you is inviolate. It doesn’t change” (ibid.:
120).
The tough issues surrounding the new community policing obviously
aren’t an “exam game” for the median inner-city voter. The people
who experience both law enforcement and crime are people who think
empathy is motivated by “inviolate” “intellectual” “principles” and just
“doesn’t depend,” “really,” on “see[ing] what’s happening in [inner-city]
neighborhood[s].” The values at stake and the difficult tradeoffs that must
be made are not abstractions. They are real. This perhaps is a place where
critics and advocates can come together. Procedural justice scholars have
suggested that a reliance on procedures is useful when correct outcomes
are not obvious to disparate members of a group. Groups may not always
decide on an outcome, but they can almost always decide on procedures
218 Tracey L. Meares

that everyone finds satisfying (Lind and Tyler 1988). Thus, to the
extent that we achieve very little agreement on constitutional substance,
perhaps the best strategy to adopt as third-party policing goes forward is
to insure that those most affected are insinuated into the political process
and that their voices are heard.

Conclusion
As we progress into the twenty-first century, it is likely that non-traditional
approaches to policing will continue to include strategies that appear to
stray from criminal justice norms. This is to be expected in a world in
which community policing, with its reliance on police and citizen inter-
action, increasingly focuses upon problems that are not obviously con-
nected to crime control such as garbage clean-up, neighborhood marches,
and even community prayer vigils. The third-party policing strategies
reviewed in this volume clearly are congenial to this brave new world of
policing. But the question remains whether this new world adequately
recognizes the values that American citizens have always held dear. Civil
liberties proponents fear not, but perhaps such critics have not taken ade-
quate account of the instrumental benefits to be obtained through the
new policing. Instrumental benefits aside, both critics and proponents
of third-party policing seem to have overlooked the potential for polit-
ical engagement that the new policing strategies can offer to citizens –
especially to crime-beleaguered citizens of poor urban communities.
Whatever else may be said about these approaches in terms of crime
reduction, the normative benefits of expanded political participation for
those who have traditionally been shut out of governmental processes
may be worth the cost of these controversial programs.

   
1. Zimring and Hawkins (1997) explain this point by distinguishing the educa-
tive and deterrent effects of punishment. Likewise, Williams and Hawkins
(1992) explain the conceptual differences between the deterrent effect of legal
sanctions and social control through informal mechanisms.
2. A person is guilty of conspiracy with another person to commit a crime if with
the purpose of promoting or facilitating its commission he: (a) agrees with such
other person or persons that they or one of them will engage in conduct which
constitutes such crime or an attempt or solicitation to commit such crime; or
(b) agrees to aid such other person or persons in the planning or commission
of such crime (Model Penal Code, § 5.03).
3. A person is guilty of solicitation to commit a crime if with the purpose of pro-
moting or facilitating its commission he commands, encourages, or requests
another person to engage in specific conduct which would constitute such
 Third-party policing 219

crime or an attempt to commit such crime or which would establish his com-
plicity in its commission or attempted commission (Model Penal Code, § 5.02).
4. Note that one dissenter in Bennis, Justice Stevens, clearly disagrees with this
point (see Bennis at 1009, Stevens, J., dissenting). Justice Stevens declares that
the goal of deterrence is not served by punishing “a person who has taken all
reasonable steps to prevent an illegal act.”
5. In fact, the court, relying on substantive due process principles, held that
once a criminal is found “not guilty” of a crime, the State may not impose
punishment” (Foucha v. Louisiana).
6. Salerno vs. United States describes preventive detention as regulatory rather
than punitive and thus requiring more lenient procedural protection under
the Due Process Clause.
7. For example, see 18 USC. 981 (1988 & Supp. V 1993) for a discussion of
the government’s civil money laundering forfeiture provision or 21 U.S.C.
881 (1988 & Supp. V 1993) which authorizes the civil forfeiture of prop-
erty connected to narcotics activity. Forfeiture is authorized in more than 140
federal statutes and most states have one or more laws permitting forfeiture
(Steven L. Kessler [1994]. Civil and Criminal Forfeiture: Federal and State
Practice 2.01, at 2–1). Typically the government can effect its seizure with-
out notice to the owner and without giving the owner a prior opportunity to
object. See, for example, 19 USC. 1609–1615 (1988 & Supp. V 1993), the US
Customs Service’s procedures for seizure, forfeiture, and recovery of seized
property.
8. I should acknowledge here that the two noted aims do not necessary proceed
in step. That is, it is certainly possible to imagine a world in which there is
much less open-air drug selling and a simultaneous increase in consumption.



Braman, D. (2004). Doing time on the outside: The hidden effects on families and
communities. Ann Arbor: University of Michigan Press.
Caulkins, J. P. (1998). The cost effectiveness of civil remedies: The case of drug
control interventions. In L. G. Mazerolle and J. Roehl (eds.), Civil remedies
and crime prevention, 9. Monsey, NY: Criminal Justice Press.
Cheh, M. M. (1994). Can something this easy, quick, and profitable also be fair?
Runaway civil forfeiture stumbles on the Constitution. New York Law School
Law Review, 29, n1.
Committee to Review Research on Police Policy and Practices. (2004). Fair-
ness and effectiveness in policing: The evidence. Washington, DC: National
Academies Press.
Douglas, W. O. (1960). Vagrancy and arrest on suspicion. Yale Law Journal, 70,
n1: 1–14.
Doyle, D. P. and Luckenbill, D. F. (1991). Mobilizing law in response to collec-
tive problems: A test of Black’s Theory Of Law. Law & Society Review, 25,
103–116.
Grogger, J. (2002). The effects of civil gang injunctions on reported violent crime.
Journal of Law and Economics, 45, 69.
220 Tracey L. Meares

Klarman, M. J. (1991). The puzzling resistance to political process theory.

Virginia Law Review, 77, 747–766.
Lind, E. A. and Tyler, T. R. (1988). The social psychology of procedural justice. New
York: Plenum Press.
Lynch, J. P. and Sabol, W. J. (2004). Assessing the effects of mass incarceration
on informal social control in communities. Criminology and Public Policy, 3,
267–294.
Mazerolle, L. and Ransley, J. (2003). Third-party policing: Prospects, challenges
and implications for regulators. Paper presented at the Current Issues In
Regulations: Enforcement and Compliance Conference, Melbourne.
Mazerolle, L. G. and Roehl, J. (eds.). (1998). Civil remedies and crime prevention.
Monsey, NY: Criminal Justice Press.
Meares, T. L. (2002). Praying for community policing. California Law Review,
90, 1593–1634.
Nagin, D. (1998). Criminal deterrence research at the outset of the twenty-first
century. In M. Tonry (ed.), Crime and justice: A review of research. Chicago:
University of Chicago Press.
Ready, J., Mazerolle, L. G., and Revere, E. (1998). Getting evicted from public
housing: An analysis of the factors influencing eviction decisions in six public
housing sites. In L. G. Mazerolle and J. Roehl (eds.), Civil remedies and crime
prevention. Monsey, NY: Criminal Justice Press.
Roberts, D. (1999). Foreword: Race, vagueness, and the social meaning of order
maintenance policing: Supreme court issue. Journal of Criminology and Crim-
inal Law, 89, n3: 776–785.
Rose, D. and Clear, T. (1998). Incarceration, social capital and crime: Implica-
tions for social disorganization theory. Criminology, 36, 441–480.
Sampson, R. J. and Wilson, W. J. (1995). Toward a theory of race, crime and
urban inequality. In J. Hagan and R. Peterson (eds.), Crime and inequality.
Stanford: Stanford University Press.
Sampson, R. J., Raudenbush, S. W., and Earls, F. (1997). Neighborhoods and
violent crime: a multilevel study of collective efficacy. Science, 277, 918–925.
Shaw, C. R. and McKay, H. D. (rev. ed. 1969). Juvenile delinquency and urban
areas: A study of rates of delinquency in relationship to differential characteristics
of local communities in American cities. Chicago: University of Chicago Press.
Transcript of Meeting before Chicago City Council Committee on Police and
Fire 107 (1992, May 18).
Williams, R. and Hawkins, R. (1992). Wife assault, costs of arrest, and the deter-
rence process. Journal of Research in Crime and Delinquency, 29, n3: 292–294.
Wilson, W. J. (1996). When work disappears: The world of the new urban poor. New
York: Knopf.
Zimring, F. E. and Hawkins, G. (1997). Crime is not the problem: Lethal violence
in America. New York: Oxford University Press.

CASES CITED
Bennis v. Michigan, 116 S. Ct. 994 (1996).
City of Maquoketa v. Russell, 484 N.W.2d 179 Iowa (1992).
Foucha v. Louisiana, 504 US 71, 95–98. (1992).
Hutchins v. District of Columbia, 942 F. Supp. 665 D.D.C. (1996).
 Third-party policing 221

Nunez v. City of San Diego, 114 F.3d 935 9th Cir. (1997).
Salerno vs. United States, 481 US 739 (1987).

STATUTES CITED
18 USC. 981 (1988 & Supp. V 1993).
19 USC. 1609–1615 (1988 & Supp. V 1993)
21 USC. 881 (1988 & Supp. V 1993)
Chicago, Il., Code § 8-4-015 (1992).
Part VI

Hot spots policing

12 Advocate
Hot spots policing as a model for police
innovation

David Weisburd and Anthony A. Braga

Looking at the major police innovations of the last decade, what is most
striking from a criminologist’s perspective is the extent to which new
programs and practices have been developed without reference to either
criminological theory or research evidence. Some institutional theorists
might argue that this is understandable given the limited ability of police
agencies to reliably demonstrate their successes, and the political envi-
ronments within which police agencies must operate (Meyer and Rowan
1977; Mastrofski and Ritti 2000; Willis, Mastrofski, and Weisburd 2004).
However, this reality is very much at odds with a model of policing that
would seek to draw new policies and practices from a solid research base
(Sherman 1998), and suggests an approach to policing that is based more
on intuition and luck than on research and experimentation. Recent stud-
ies of the adoption of police innovation reinforce this problematic por-
trait of American police innovation. Widely touted programs such as
Community Policing or Compstat have been widely diffused across the
landscape of American policing absent any reliable evidence that they
accomplish the goals that they set out to achieve at the outset (Weisburd,
Mastrofski, McNally et al. 2003; Weisburd and Eck 2004).
In this context, “hot spots policing” represents a particularly impor-
tant innovation on the American police scene. Its origins can be traced
to innovations in criminological theory, and it was subjected to careful
empirical study before it was diffused widely across American police agen-
cies. In the following pages we describe the origins of hot spots policing in
theory and basic research, and the research evidence that has been devel-
oped to support hot spots policing practices. We will argue that there is
strong theoretical justification for hot spots policing, and that evaluation
evidence provides a solid empirical basis for continued experimentation
and development of this approach.

From theory to practice: an evidence-based model

The idea of hot spots policing can be traced to recent critiques of tradi-
tional criminological theory. For most of the last century criminologists

225
226 David Weisburd and Anthony A. Braga

have focused their understanding of crime on individuals and commu-

nities (Nettler 1978; Sherman 1995). In the case of individuals, crim-
inologists have sought to understand why certain people as opposed to
others become criminals (e.g., see Hirschi 1969; Akers 1973; Gottfredson
and Hirschi 1990; Raine 1993), or to explain why certain offenders
become involved in criminal activity at different stages of the life course
or cease involvement at other stages (e.g., see Moffitt 1993; Sampson and
Laub 1993). In the case of communities, criminologists have often tried
to explain why certain types of crime or different levels of criminality are
found in some communities as contrasted with others (e.g., see Shaw and
McKay 1972; Sampson and Groves 1989; Bursik and Grasmick 1993;
Agnew 1999) or how community-level variables, such as relative depri-
vation, low socioeconomic status, or lack of economic opportunity may
affect individual criminality (e.g., see Cloward and Ohlin 1960; Wolfgang
and Ferracuti 1967; Merton 1968; Agnew 1992). In most cases, research
on communities has focused on the “macro” level, often studying states
(Loftin and Hill 1974), cities (Baumer, Lauritsen, Rosenfeld, and Wright
1998), and neighborhoods (Sampson 1985; Bursik and Grasmick 1993).
This is not to say that criminologists did not recognize that the opportu-
nities found at more micro levels of place can impact upon the occurrence
of crime. Edwin Sutherland, for example, whose main focus was upon the
learning processes that bring offenders to participate in criminal behav-
ior, noted in his classic criminology textbook that the immediate situation
influences crime in many ways. For example, “a thief may steal from a
fruit stand when the owner is not in sight but refrain when the owner is
in sight; a bank burglar may attack a bank which is poorly protected but
refrain from attacking a bank protected by watchmen and burglar alarms”
(Sutherland 1947: 5). Nonetheless, Sutherland, as other criminologists,
did not see micro crime places as a relevant focus of criminological study.
This was the case, in part, because crime opportunities provided by such
places were assumed to be so numerous as to make concentration on
specific places of little utility for theory or policy. In turn, criminologists
traditionally assumed that situational factors played a relatively minor
role in explaining crime as compared with the “driving force of crimi-
nal dispositions” (Clarke and Felson 1993: 4; Trasler 1993). Combining
an assumption of a wide array of criminal opportunities, and a view of
offenders that saw them as highly motivated to commit crime, it is under-
standable that criminologists paid little attention to the problem of the
development of crime at micro levels of place.
While the focus on individuals and communities has continued to play a
central role in criminological theory and practice, traditional theories and
approaches were subjected to substantial criticism beginning in the 1970s.
 Hot spots policing as a model 227

Starting with Robert Martinson’s critique of rehabilitation programs in

1974, a series of studies documented the failures of traditional crime pre-
vention initiatives (e.g., Sechrest, White, and Brown 1979; Whitehead
and Lab 1989). In policing, as well, there was substantial criticism of
traditional approaches. For example, there was no more visible approach
to crime prevention in policing in the 1970s, or one that involved greater
cost, than preventive patrol in cars. The idea that police presence spread
widely across the urban landscape was an important method for pre-
venting crime and increasing citizens’ feelings of safety was a bedrock
assumption of American policing. But in a major evaluation of preventive
patrol in Kansas City, Missouri, the Police Foundation concluded that
increasing or decreasing the intensity of preventive patrol did not affect
either crime, service delivery to citizens, or citizens’ feelings of security
(Kelling, Pate, Dieckman, and Brown 1974). Similarly, rapid response
to emergency calls to the police was considered to be a crucial compo-
nent of police effectiveness. Yet in another large-scale study, Spelman and
Brown (1984) concluded that improvement in police response times has
no appreciable impact on the apprehension or arrest of offenders.
These and other studies in the 1970s and 1980s led scholars to chal-
lenge the fundamental premise of whether the police could have a signif-
icant impact on crime (see also Levine 1975; Greenwood, Petersilia, and
Chaiken 1977). By 1994, the distinguished police scholar David Bayley
was able to write:

The Police do not prevent crime. This is one of the best-kept secrets of modern
life. Experts know it, the police know it, but the public does not know it. Yet
the police pretend that they are society’s best defence against crime . . . This
is a myth. First, repeated analysis has consistently failed to find any connection
between the number of police officers and crime rates. Secondly, the primary
strategies adopted by modern police have been shown to have little or no effect
on crime. (1994: 3)

A number of scholars argued that the failures of traditional crime pre-

vention could be found in the inadequacies in program development
and research design in prior studies (e.g., Farrington, Ohlin, and Wilson
1986; Goldstein 1990). Other reviews stressed that there are examples
of successful offender-focused crime prevention efforts, which can pro-
vide guidance for the development of more effective prevention policies
(Farrington 1983; Lipsey 1992). Nonetheless, even those scholars who
looked to improve such policies, came to recognize the difficulties inher-
ent in trying to do something about criminality (Visher and Weisburd
1998). Summarizing the overall standing of what they define as traditional
“offender-centred” crime prevention, Patricia and Paul Brantingham
228 David Weisburd and Anthony A. Braga

wrote: “If traditional approaches worked well, of course, there would

be little pressure to find new forms of crime prevention. If traditional
approaches worked well, few people would possess criminal motivation
and fewer still would actually commit crimes” (1990: 19).
One influential critique of traditional criminological approaches to
understanding crime that had a strong influence on the development
of hot spots policing was introduced by Cohen and Felson (1979). They
argued that the emphasis placed on individual motivation in crimino-
logical theory failed to recognize the importance of other elements of
the crime equation. They noted that for criminal events to occur there
is need not only of a criminal, but also of a suitable target and the
absence of a capable guardian. They showed that crime rates could
be affected by changing the nature of targets or of guardianship, irre-
spective of the nature of criminal motivations. Cohen’s and Felson’s sug-
gestion that crime could be affected without reference to the motivations
of individual offenders was a truly radical idea in criminological circles
in 1979. The routine activities perspective they presented established the
context of crime as an important focus of study.
Drawing upon similar themes, British scholars led by Ronald Clarke
began to explore the theoretical and practical possibilities of situational
crime prevention (Clarke 1983; Cornish and Clarke 1986; Clarke 1992;
1995). Their focus was on criminal contexts and the possibilities for
reducing the opportunities for crime in very specific situations. Their
approach, like that of Cohen and Felson, turned traditional crime pre-
vention theory on its head. At the center of their crime equation was
opportunity. And they sought to change opportunity rather than reform
offenders. In situational crime prevention, more often than not “opportu-
nity made the thief” (Felson and Clarke 1998). This was in sharp contrast
to the traditional view that the thief simply took advantage of a very large
number of potential opportunities. Importantly, in a series of case stud-
ies situational crime prevention advocates showed that reducing criminal
opportunities in very specific contexts can lead to crime reduction and
prevention (Clarke 1992; 1995).
One natural outgrowth of these perspectives was that the specific places
where crime occurs would become an important focus for crime preven-
tion researchers (Eck and Weisburd 1995; Taylor 1997). While concern
with the relationship between crime and place goes back to the founding
generations of modern criminology (Guerry 1833; Quetelet 1842), the
“micro” approach to places emerged only in the last few decades (e.g.,
see Brantingham and Brantingham 1975; Duffala 1976; Mayhew, Clarke,
Sturman, and Hough 1976; Rengert 1980; Brantingham and Branting-
ham 1981; Rengert 1981; LeBeau 1987; Hunter 1988).1 Places in this
 Hot spots policing as a model 229

micro context are specific locations within the larger social environments
of communities and neighborhoods (Eck and Weisburd 1995). They
are sometimes defined as buildings or addresses (see Sherman, Gartin,
and Buerger 1989a; Green 1996), sometimes as block faces, or street
segments (see Sherman and Weisburd 1995; Taylor 1997), and some-
times as clusters of addresses, block faces, or street segments (see Block,
Dabdoub, and Fregly 1995; Weisburd and Green 1995).
In the mid to late 1980s a group of criminologists began to examine
the distribution of crime at micro places. Their findings were to radi-
cally alter the way many criminologists understood the crime equation,
drawing them into a new area of inquiry that was to have important impli-
cations for police practice. Perhaps the most influential of these studies
was conducted by Lawrence Sherman and his colleagues (Sherman et al.
1989b). Looking at crime addresses in the city of Minneapolis they found
a concentration of crime at places that was startling. Only 3 percent of
the addresses in Minneapolis accounted for 50 percent of the crime calls
to the police. Similar results were reported in a series of other studies in
different locations and using different methodologies, each suggesting a
very high concentration of crime in micro places (e.g., see Pierce, Spaar,
and Briggs 1986; Weisburd, Maher, and Sherman 1992; Weisburd and
Green 1994; Weisburd, Bushway, Lum, and Yang 2004). A recent study
by Weisburd et al. (2004) shows moreover that the concentration of crime
in hot spots is fairly stable across time.
Importantly, such concentration of crime at discrete places does not
necessarily follow traditional ideas about crime and communities. There
were often discrete places free of crime in neighborhoods that were
considered troubled, and crime hot spots in neighborhoods that were
seen generally as advantaged and not crime prone (Weisburd and Green
1994). This empirical research thus reinforced theoretical perspectives
that emphasized the importance of crime places, and suggested a focus
upon small areas, often encompassing only one or a few city blocks that
could be defined as hot spots of crime.

The emergence of hot spots policing

These emerging theoretical paradigms and empirical findings led Sher-
man and Weisburd (1995) to explore the practical implications of the
hot spots approach for policing. With cooperation from the Minneapo-
lis Police Department they developed a large experimental field study of
“police patrol in crime hot spots.” They sought to challenge the con-
clusions of the Kansas City Preventive Patrol Experiment noted earlier,
then well established, that police patrol has little value in preventing or
230 David Weisburd and Anthony A. Braga

controlling crime. But they also sought to show that the focus of police
efforts on crime hot spots presented a new and promising approach for
police practice.
The idea of focusing police patrol on crime hot spots represented a
direct application of the empirical findings regarding the concentration
of crime in micro places. The Kansas City Preventive Patrol Experiment
had looked at the effects of police patrol in large police beats. However,
if “only 3 percent of the addresses in a city produce more than half of all
the requests for police response, if no police are dispatched to 40 percent
of the addresses and intersections in a city over one year, and, if among
the 60 percent with any requests the majority register only one request a
year (Sherman et al. 1989b), then concentrating police in a few locations
makes more sense that spreading them evenly through a beat” (Sherman
and Weisburd 1995: 629).
Nonetheless, the application of these findings to police practice raised
significant questions about the overall crime control benefits of a hot spots
approach. How would one know if crime prevention benefits gained at hot
spots would not simply be displaced to other areas close by? Sherman and
Weisburd noted that displacement was a potential but not necessarily cer-
tain occurrence. However, they argued that the first task for researchers
was to establish that there would be any deterrent effect of police presence
at the hot spots themselves:
The main argument against directing extra resources to the hot spots is that
it would simply displace crime problems from one address to another without
achieving any overall or lasting reduction in crime. The premise of this argument
is that a fixed supply of criminals is seeking outlets for the fixed number of crimes
they are predestined to commit. Although that argument may fit some public drug
markets, it does not fit all crime or even all vice . . . In any case, displacement
is merely a rival theory explaining why crime declines at a specific hot spot, if it
declines. The first step is to see whether crime can be reduced at those spots at
all, with a research design capable of giving a fair answer to that question. (1995:
629)

The results of the Minneapolis Experiment stood in sharp distinction to

those of the earlier Kansas City study. The study used a rigorous exper-
imental design including randomization of 110 crime hot spots of about
a city block to treatment and control conditions. The treatment sites
received on average between two and three times as much preventive
patrol as the control sites. For the eight months in which the study was
properly implemented, Sherman and Weisburd found a significant rela-
tive improvement in the experimental as compared to control hot spots
both in terms of crime calls to the police and observations of disorder.
Indeed, the effects of the program on crime, as measured by the difference
 Hot spots policing as a model 231

300

250

200 Control
Experiment
150

100

50

-50

-100
Dec. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov.
Month

Figure 12.1 The Minneapolis hot spots experiment (Lines represent

absolute differences between baseline and experimental years in total
crime calls per month)
Source: Sherman and Weisburd, 1995

between crime calls in the pre-experimental and experimental years, was

found to be stable across the eight-month period in which the program
was properly implemented (see Figure 12.1). Crime, or at least crime
calls and disorder, appeared to be prevented in the treatment as opposed
to the control locations. Sherman and Weisburd (1995: 645) concluded
that their results show “clear, if modest, general deterrent effects of sub-
stantial increases in police presence in crime hot spots.” They noted that
it was time for “criminologists to stop saying ‘there is no evidence’ that
police patrol can affect crime” (1995: 647).

The empirical evidence for hot spots policing

The Minneapolis Hot Spots Experiment led to a series of federal govern-
ment supported studies of hot spots policing. These findings in turn pro-
vided what is perhaps the strongest weight of empirical evidence for the
introduction of any policing practice (see Committe to Review Research
2004; Weisburd and Eck 2004). Moreover, subsequent studies began to
examine the displacement impacts of policing crime hot spots. For the
most part this was in terms of immediate spatial displacement, or dis-
placement to areas close to the targeted areas. Nonetheless, the findings
reinforce the utility of hot spots approaches.
In a systematic review of the research evidence, Braga (2001) identified
nine evaluations of focused police interventions at crime hot spots. The
232 David Weisburd and Anthony A. Braga

effects of problem-oriented policing initiatives, comprised mostly of tra-

ditional tactics with limited situational responses, were evaluated in the
Minneapolis Repeat Call Address Policing (RECAP) Program (Sherman
et al. 1989a), the Jersey City Drug Markets Analysis Program (Weisburd
and Green 1995); the Jersey City Problem-Oriented Policing at
Violent Places Study (Braga, Weisburd, Waring et al. 1999), the St. Louis
Problem-Oriented Policing at Drug Market Locations Study (Hope
1994), and the Beenleigh (Australia) Calls for Service Project (Crimi-
nal Justice Commission 1998). The evaluation of the Houston Targeted
Beat Program examined the effects of three types of treatments applied in
different target areas. These interventions included high-visibility patrol,
zero tolerance disorder policing, and enforcement problem-oriented
policing (Caeti 1999). The Kansas City Gun Project examined the gun
violence prevention effects of proactive patrol and intensive enforcement
via safety frisks during traffic stops, plain view searches and seizures,
and searches incident to arrests on other charges (Sherman and Rogan
1995a). The Minneapolis Hot Spots Patrol Program, as described ear-
lier, evaluated the effects of increased levels of preventive patrol on crime.
The Kansas City Crack House Police Raids Program (Sherman and
Rogan 1995b) evaluated the effects of court-authorized raids on crack
houses.
Noteworthy crime reductions were reported in seven of the nine
selected studies. Five studies examined whether focused police efforts
were associated with crime displacement or diffusion of crime control
benefits. None of the five studies reported substantial immediate spa-
tial displacement of crime into areas surrounding the targeted locations.
Four studies suggested a possible “diffusion of crime control benefits”
(Clarke and Weisburd 1994) effect associated with the focused police
interventions. In these studies, areas near to the target sites, but which
did not receive special police attention, also improved in terms of mea-
sures used in the studies. Table 12.1 summarizes the findings of that
review.
Based on these studies, a recent review of police practices and policies
by a National Academy of Sciences panel concluded that the strongest
evidence presently available in support of any policing approach is found
for hot spots policing:
(S)tudies that focused police resources on crime hot spots provide the strongest
collective evidence of police effectiveness that is now available. On the basis
of a series of randomized experimental studies, we conclude that the practice
described as hot-spots policing is effective in reducing crime and disorder and
can achieve these reductions without significant displacement of crime control
 Hot spots policing as a model 233

Table 12.1 Results of hot spots policing evaluations

Displacement /
Study Crime Outcomes Other Outcomes Diffusion

Minneapolis (MN) No Effect None Not measured

RECAP,
Sherman et al.
(1989a)
Minneapolis (MN) Reductions in total Observational data Not measured
Hot Spots, citizen calls for revealed
Sherman and service reductions in
Weisburd (1995) crime and
disorder
Jersey City (NJ) Reductions in citizen None Little evidence of
DMAP, calls for disorder displacement;
Weisburd and offenses analyses suggest
Green (1995) modest diffusion
of crime control
benefits
Jersey City (NJ) Reductions in total Observational data Little evidence of
POP at Violent calls for service and revealed immediate spatial
Places, Braga total crime reductions in displacement.
et al. (1999) incidents. All social and Possible diffusion
subcategories of physical disorder of crime control
crime experienced benefits
varying reductions
St. Louis (MO) All 3 drug locations None One location
POP in 3 Drug experienced varying experienced
Areas, Hope reductions in total significant
(1994) calls displacement
Kansas City (MO) Modest decreases in None Not measured
Crack House citizen calls and
Raids, Sherman offense reports that
and Rogan decayed in two
(1995a) weeks
Kansas City (MO) Increase in guns Community survey No significant
Gun Project, seized by the police revealed crime
Sherman and followed by favorable opinion displacement
Rogan (1995b) decrease in gun of police efforts Diffusion effects
crimes reported
Houston (TX ) Aggregated None No evidence of
Targeted Beat experimental beats displacement
Program, Caeti experienced Diffusion effects
(1999) significant crime reported
reductions. Specific
beats reported
mixed results
(cont.)
234 David Weisburd and Anthony A. Braga

Table 12.1 (cont.)

Displacement /
Study Crime Outcomes Other Outcomes Diffusion

Beenleigh (AUS ) No noteworthy None Not measured

Calls for Service differences in total
Project, Criminal number of calls
Justice between Beenleigh
Commission and Brown Plains
(1998) areas. Noteworthy
reductions in calls
reported in majority
of case studies

Source: Adapted from Braga (2001).

benefits. Indeed, the research evidence suggests that the diffusion of crime control
benefits to areas surrounding treated hot spots is stronger than any displacement
outcome (Committee to Review Research 2004: 250).

Some caveats regarding the outcomes of hot spots policing

While the evidence for the effectiveness of hot spots policing is convinc-
ing, we think it important to note that there are still significant gaps in
our knowledge about the effects of these interventions. For example, we
know little of which specific hot spots strategies work best in which spe-
cific types of situations. While there is strong evidence that focusing on
hot spots reduces crime and disorder, research has not yet distinguished
the types of hot spots strategies that lead to the strongest crime prevention
benefits (Braga 2001). The National Academy of Sciences review sug-
gests that the most generalized strategies, for example, preventive patrol
(Sherman and Weisburd 1995) and drug raids (Sherman and Rogan
1995a), are likely to have less impact than approaches that include more
problem-solving elements, such as working with landlords (Eck and
Wartell 1996; Green-Mazerole and Roehl 1998). However, our under-
standing of the effects of hot spots policing remains very general. If we are
to maximize the crime prevention effects of hot spots approaches we need
to examine carefully the interaction of different strategies with different
hot spots settings. This effort would demand a large group of studies. But
given the promise of hot spots policing, such an investment in research
in this area seems appropriate.
 Hot spots policing as a model 235

Also, too little attention has been paid to the potential harmful effects
of hot spots approaches. Police effectiveness studies have traditionally
overlooked the effects of policing practices upon citizen perceptions of
police legitimacy (Tyler 2000; 2001). Does the concentration of police
enforcement in specific hot spots lead citizens to question the fairness of
police practices? There is some evidence that residents of areas that are
subject to hot spots policing welcome the concentration of police efforts in
problem places (see Shaw 1995). Nonetheless, focused aggressive police
enforcement strategies have been criticized as resulting in increased citi-
zen complaints about police misconduct and abuse of force in New York
City (Greene 1999). As in the case of understanding the effectiveness of
police strategies, the potential impacts of hot spots policing on legitimacy
may depend in good part on the types of strategies used and the context of
the hot spots affected. But, whatever the impact, we need to know more
about the effects of hot spots policing approaches on the communities
that the police serve.
Finally, the research overall strongly supports the position that hot
spots policing can have a meaningful effect on crime without simply dis-
placing crime control benefits to areas nearby. However, a recent study
which reinforces the findings that there is little immediate spatial dis-
placement of crime as a result of hot spots policing approaches, identifies
other displacement outcomes that may occur in focused policing efforts
(Weisburd, Wyckoff, Ready et al. 2004). Offenders interviewed in the
study described factors that inhibited spatial displacement, including the
importance of familiar territory to offenders, and the social organization
of illicit activities at hot spots which often precluded easy movement to
other areas that offer crime opportunities. Prostitutes, for example, were
found to work near to their homes, and described being uncomfortable
moving to other areas where different types of people worked and differ-
ent types of clients were found. Prostitutes and drug dealers in the study
described the importance of the familiarity of a place to their clients, and
some offenders talked of the dangers of encroaching on the territory of
offenders in other hot spots.
Overall, a number of factors seemed to discourage spatial displace-
ment in the study. Nonetheless, Weisburd et al. (2004) find that offend-
ers will often try other modes of adaptation to police interventions, the
most common being a change in the methods of committing illegal acts.
For example, prostitutes and drug dealers may begin to make “appoint-
ments” with their customers, or move their activities indoors in order
to avoid heightened police activity on the street. While the net gain in
crime prevention may still be large for hot spots efforts, these findings
236 David Weisburd and Anthony A. Braga

suggest the importance of continued investigation of possible non-spatial

displacement outcomes in hot spots policing.

The link between research and practice

We have so far shown that hot spots policing emerged from theoretical
developments in criminology, and basic research indicating a very high
concentration of crime at hot spots. Hot spots policing was also subjected
to substantial experimental evaluation during the early 1990s that showed
that it could be an effective policing strategy to prevent crime and disor-
der. Importantly, it is also the case that this approach had diffused widely
in American policing by the turn of the last century. In a 2001 Police
Foundation study more than seven in ten police departments with more
than 100 sworn officers reported using crime mapping to identify “crime
hot spots” (Weisburd, Mastrofski, McNally, and Greenspan 2001). But
the fact that the research literature is supportive of hot spots policing and
that police have implemented this approach widely, does not necessar-
ily mean that research strongly impacted police practice. For example,
it may be that the police adopted hot spots policing independently and
later evidence simply confirmed that the strategy was useful.
While the Minneapolis Hot Spots Experiment is the first example we
know of a successful program in which micro-level crime hot spots were
systematically identified on a large scale for the purpose of police inter-
vention, it is not the first example of police use of crime mapping to
identify crime problems. Police officers have long recognized the impor-
tance of place in crime. Hand-developed pin maps have been widely used
in police agencies for over half a century (Weisburd and McEwen 1997).
And one can find isolated examples of what today we would define as
hot spots policing in earlier periods (e.g., see Weiss 2001). Moreover,
during the 1970s, crime analysts looked for patterns in crime by plotting
the locations and times at which crimes were committed to direct patrol
officers to the most likely targets (Reinier 1977), and cutting-edge crime
analysts were experimenting with computerized crime mapping before
the hot spots studies were well known (Weisburd and Lum 2005).
Nonetheless, a recent study by Weisburd and Lum (2005) suggests
that the timing of the wide-scale implementation of hot spots policing
follows very closely the basic and applied research we have reviewed.
The study examined the diffusion of computerized crime mapping in
police agencies using data from the National Institute of Justice Crime
Mapping Laboratory (Mamalian, LaVigne, and Groff 1999), and a small
pilot study of 92 police agencies of over 100 sworn officers conducted
by the researchers. They found that computerized crime mapping in
 Hot spots policing as a model 237

Weisburd and Lum Pilot Study Cumulative Adoption Curve

60

50
Number of Departments

40

30

20

10

0
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
Year

1997 CMRC Cumulative Adoption Curve

200
180
Number of Departments

160
140
120
100
80
60
40
20
0
1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

Year

Figure 12.2 Cumulative distribution of computerized crime mapping

adopted from the Weisburd et al. Pilot Study and the CMRC study
Source: Weisburd and Lum 2005

larger police agencies first began to emerge in policing in the late 1980s
and early 1990s. Adoption began to grow steeply in the mid-1990s
with the number of adopters increasing at a large rate after 1995 (see
Figure 12.2).
238 David Weisburd and Anthony A. Braga

Weisburd and Lum (2005) make a direct link in their survey between
the diffusion of innovation in crime mapping and the adoption of hot
spots policing. When they asked why departments developed a crime
mapping capability, nearly half of those surveyed responded that crime
mapping was adopted to facilitate hot spots policing. Of other categories
of responses, many were likely related to hot spots approaches, though
respondents gave more general replies such as “crime mapping was ini-
tially developed in response to a specific police strategy.” Moreover, they
found that 80 percent of the departments in their sample that have a
computerized crime mapping capability conduct computerized hot spots
analysis, and two thirds of departments that have computerized crime
mapping capabilities use hot spots policing as a policing tactic.
If we can make the link between hot spots policing and computerized
crime mapping suggested by Weisburd and Lum, then the data suggest
that hot spots policing emerged as an important police strategy precisely
during the period that evaluation findings were being widely dissemi-
nated. Results of the Minneapolis Hot Spots experiment, for example,
though first published in an academic journal in 1995, were the focus
of a plenary panel at the Academy of Criminal Justice Sciences in 1990
that was chaired by the then Director of the National Institute of Justice,
James K. Stewart. Overall, the emergence of crime mapping and hot
spots policing follow closely the development of strong research evidence
regarding the hot spots approach. Of course, these data do not confirm
with certainty a causal link between research and practice in the adoption
of hot spots policing, but they suggest that hot spots policing approaches
began to be widely implemented after research studies began to show
their effectiveness.

Conclusion
In policing, most innovation has been developed using what might be
termed a “clinical experience model.” In such a model, research may
play a role, but the adoption of innovation is determined primarily by
the experiences of practitioners and often has little to do with research
evidence. Such models often have a weak theoretical basis and it is not
uncommon to discover that they have little crime prevention value once
they are subjected to serious empirical investigation. Given the impor-
tance of policing for public safety, it seems unreasonable that policing
should continue to rely on such a model for the development and diffu-
sion of innovation.
Our discussion of hot spots policing suggests an alternative model for
police innovation. Hot spots policing was consistent with developing
theoretical insights in criminology and was supported by basic
 Hot spots policing as a model 239

criminological research on crime and place. Accordingly, before hot spots

policing was to emerge as a coherent strategic approach there was strong
theoretical justification and empirical support for testing the value of this
strategy. Hot spots policing was subjected to rigorous empirical inves-
tigation before it was widely diffused and adopted by American police
agencies. In this sense, hot spots policing suggests that “evidence-based
policing,” as Lawrence Sherman (1998) has called it, can form the basis
for important police innovation. We think, overall, that hot spots policing
is an exception to the rule of how police innovation has developed in the
United States. Nonetheless, it provides a model for how policing innova-
tion might be developed even if, to date, it is the exception rather than
the rule.
Hot spots policing is a model for the integration of research in the
world of policing, and this integration has produced what is, accord-
ing to empirical evidence, the most effective police innovation of the last
decade. However, as we noted earlier, there is still much to be learned. For
example, our knowledge is too general and must be focused more on how
specific policing strategies affect specific types of hot spots. The research
to date has also ignored many of the potential social consequences of
hot spots policing. While it is clear that crime prevention benefits can
be gained from hot spots approaches, we need to know more about how
they affect the lives of people who live in areas that are targeted. Finally,
we need to focus more carefully on problems of displacement and dif-
fusion. Spatial displacement appears to be a much less serious threat to
the gains of hot spots policing than had been originally thought, and
indeed the evidence suggests that diffusion of crime control benefits to
areas near targeted places is more common. Nonetheless, we don’t know
enough about how other forms of displacement, such as changes in meth-
ods of crime commission, affect the crime control benefits of hot spots
approaches.

  
1. It should be noted that a few early criminologists did examine the “micro”
idea of place as discussed here (see Shaw and Myers 1929). However, interest
in micro places was not sustained and did not lead to significant theoretical or
empirical inquiry.


Agnew, R. (1992). Foundation for a general strain theory of crime and
delinquency. Criminology, 30, 47–84.
(1999). A general strain theory of community differences in crime rates. Journal
of Research in Crime and Delinquency, 36, 123–155.
240 David Weisburd and Anthony A. Braga

Akers, R. (1973). Deviant behavior: A social learning approach. Belmont, CA:

Wadsworth Publishing.
Baumer, E., Lauritsen, J., Rosenfeld, R., and Wright, R. (1998). The influence of
crack cocaine on robbery, burglary, and homicide rates: A cross-city, longitu-
dinal analysis. Journal of Research in Crime and Delinquency, 35 (3), 316–340.
Bayley, D. (1994). Police for the future. New York: Oxford University Press.
Block, C., Dabdoub, M., and Fregly, S. (eds.). (1995). Crime analysis through
computer mapping. Washington, DC: Police Executive Research Forum.
Braga, A. (2001). The effects of hot spots policing on crime. Annals of the American
Academy of Political and Social Science, 578, 104–125.
Braga, A., Weisburd, D., Waring, E., Mazerolle, L. G., Spelman, W., and Gajew-
ski, F. (1999). Problem oriented policing in violent crime places: A random-
ized controlled experiment. Criminology, 37 (3), 541–580.
Brantingham, P. J. and Brantingham, P. L. (1975). Residential burglary and urban
form. Urban Studies, 12 (3), 273–284.
(1981). Notes on the geometry of crime. In P. J. Brantingham and P. L.
Brantingham (eds.), Environmental criminology. Beverly Hills, CA: Sage
Publications.
(1990). Situational crime prevention in practice. Canadian Journal of Criminol-
ogy, January, 17–40.
Bursik, R. J., Jr. and Grasmick, H. G. (1993). Neighborhoods and crime. San
Francisco: Lexington.
Caeti, T. (1999). Houston’s targeted beat program: A quasi-experimental test of police
patrol strategies. Doctoral dissertation, Sam Houston State University. Ann
Arbor, MI: UMI.
Clarke, R. V. (1983). Situational crime prevention: Its theoretical basis and
practical scope. In M. Tonry and N. Morris (eds.), Crime and justice: An
annual review of research, 4 (pp. 225–256). Chicago: University of Chicago
Press.
(1992). Situational crime prevention: Successful case studies. Albany, NY: Harrow
and Heston.
(1995). Situational crime prevention: Achievements and challenges. In M.
Tonry and D. Farrington (eds.), Building a safer society: Strategic approaches
to crime prevention, crime and justice: A review of research, 19 (pp. 91–150).
Chicago: Chicago University Press.
Clarke, R. V. and Felson, M. (eds.). (1993). Routine activity and rational choice.
Advances in Criminological Theory, 5, 1–14. New Brunswick, NJ: Transaction
Press.
Clarke, R. V. and Weisburd, D. (1994). Diffusion of crime control benefits: Obser-
vations on the reverse of displacement. In R. V. Clarke (ed.), Crime prevention
studies, 2 (pp. 165–183). Monsey, NY: Criminal Justice Press.
Cloward, R. and Ohlin, L. (1960). Delinquency and opportunity. Glencoe, IL: Free
Press.
Cohen, L. E. and Felson, M. (1979). Social change and crime rate trends: A
routine activity approach. American Sociological Review, 44, 588–605.
Committee to Review Research on Police Policy and Practices. (2004). Fair-
ness and effectiveness in policing: The evidence. Washington, DC: National
Academies Press.
 Hot spots policing as a model 241

Cornish, D. B. and Clarke, R. V. (1986). The reasoning criminal: Rational choice

perspectives on offending. New York: Springer-Verlag.
Criminal Justice Commission. (1998). Beenleigh calls for service project: Evaluation
report. Brisbane, Queensland: Criminal Justice Commission.
Duffala, D. C. (1976). Convenience stores, robbery, and physical environmental
features. American Behavioral Scientist, 20, 227–246.
Eck, J. E. and Spelman, W. (1987). Problem solving: Problem-oriented policing in
Newport News. Washington, DC: Police Executive Research Forum.
Eck, J. E. and Wartell, J. (1996). Reducing crime and drug dealing by improving
place management: A randomized experiment. Report to the San Diego Police
Department. Washington, DC: Crime Control Institute.
Eck, J. E. and Weisburd, D. (eds.). (1995). Crime and place: Crime prevention
studies, 4. Monsey, NY: Criminal Justice Press.
Farrington, D. P. (1983). Offending from 10 to 25 years of age. In K. Van
Dusen and S. A. Mednick (eds.), Prospective studies of crime and delinquency
(pp. 17–38). Boston: Kluwer-Nijhoff.
Farrington, D. P., Ohlin, L., and Wilson, J. Q. (1986). Understanding and control-
ling crime. New York: Springer-Veriag.
Felson, M. and Clarke, R. V. (1998). Opportunity makes the thief: Practical
theory for crime prevention. Police Research Series: Policing and Reducing Crime
Unit, Paper 98. London: Home Office.
Franks, W. D. (1980). Montpelier, Vermont’s directed patrol experiment. Police
Chief , January, 24–26.
Gay, W. G., Beall, T. M., and Bowers, R. A. (1984). A four-site assessment of the
integrated criminal apprehension program. Washington, DC: University City
Science Center.
Gay, W. G., Schell, T., and Schack, S. (1977). Prescriptive package: Improving
patrol productivity, volume I – routine patrol. Washington, DC: US Government
Printing Office.
Goldstein, H. (1990). Problem-oriented policing. New York: McGraw-Hill.
Gottfredson, M. R and Hirschi, T. (1990). A general theory of crime. Stanford,
CA: Stanford University Press.
Green, L. (1996). Policing places with drug problems. Thousand Oaks, CA: Sage
Publications.
Green-Mazerolle, L. and Roehl, J. (1998). Controlling drugs and social disorder
using civil remedies: Final report of a randomized field experiment in Oakland,
California. Washington, DC: National Institute of Justice.
Greene, J. A. (1999). Zero tolerance: A case study of police practices and policies
in New York City. Crime and Delinquency, 45, 171–181.
Greenwood, P. J., Petersilia, J., and Chaiken, J. (1977). The criminal investigation
process. Lexington: D.C. Heath.
Guerry, A. M. (1833). Essai sur la Statistique morale de la France. Paris:
Crochard.
Hirschi, T. (1969). Causes of Delinquency. Berkeley, CA: University of California
Press.
Hope, T. (1994). Problem-oriented policing and drug market locations: Three
case studies. In R. V. Clarke (eds.), Crime prevention studies, 2 (pp. 5–31).
Monsey, NY: Criminal Justice Press.
242 David Weisburd and Anthony A. Braga

Hunter, R. D. (1988). Environmental characteristics of convenience store robberies in

the state of Florida. Paper presented at the annual meeting of the American
Society of Criminology, Chicago.
Kelling, G., Pate, A. M., Dieckman, D., and Brown, C. E. (1974). The Kansas City
preventive patrol experiment: Summary report. Washington, DC: The Police
Foundation.
LeBeau, J. (1987). The methods and measures of centrography and the spatial
dynamics of rape. Journal of Quantitative Criminology, 3, 125–141.
Levine, J. P. (1975). Ineffectiveness of adding police to prevent crime. Public
Policy, 23, 523–545.
Lipsey, M. W. (1992). Juvenile delinquency treatment: A meta-analytic inquiry
into the variability of effects. In T. D. Cook, H. Cooper, D. S. Cordray,
H. Hartman, L. V. Hedges, R. J. Light, T. A. Louis, and F. Mosteller
(eds.), Meta-analysis for explanation: A casebook (pp. 83–127). New York: Sage
Publications.
Lipton, D., Martinson, R., and Wilks, J. (1975). The effectiveness of correc-
tional treatment: A survey of treatment evaluation studies. New York: Praeger
Publishers.
Loftin, C. and Hill, R. (1974). Regional subculture and homicide: An examina-
tion of the Gastil-Hackney thesis. American Sociological Review, 39, 714–724.
Mamalian, C., LaVigne, N., and Groff, E. (1999). The use of computer-
ized crime mapping by law enforcement: Survey results. NIJ Research Pre-
view. Washington, DC: National Institute of Justice. Available online at
http://www.ncjrs.org/pdffiles1/fs000237.
Martinson, R. (1974). What works? – Questions and answers about prison reform.
The Public Interest, 35, 22–54.
Mastrofski, S. and Ritti, R. (2000). Making sense of community policing: A
theoretical perspective. Police Practice and Research Journal, 1 (2), 183–210.
Mayhew, P., Clarke, R. V., Sturman, A., and Hough, M. (1976). Crime as oppor-
tunity. (Home Office Research Study 34). London: H.M. Stationery Office.
Merton, R. K. (1968). Social structure and anomie. American Sociological Review,
3, 672–682.
Meyer, J. and Rowan, B. (1977). Formal structure as myth. American Journal of
Sociology, 83, 340–363.
Moffitt, T. (1993). Adolescence-limited and life-course persistent antisocial
behavior: A developmental taxonomy. Psychological Review, 4, 674.
Nettler, G. (1978). Explaining crime. 2nd ed. New York: McGraw Hill.
Pierce, G. L., Spaar, S., and Briggs, L. R. (1986). The character of police work:
Strategic and tactical implications. Boston, MA: Center for Applied Social
Research, Northeastern University.
Quetelet, A. J. (1842). A treatise of man. Gainesville, FL: Scholar’s Facsimiles and
Reprints (1969 ed.).
Raine, A. (1993). The psychopathy of crime. New York: Academic Press.
Reinier, G. H. (1977). Crime analysis in support of patrol. National Evaluation
Program: Phase I Report. Washington, DC: US Government Printing Office.
Rengert, G. F. (1980). Theory and practice in urban police response. In D.
Georges-Abeyie and K. Harries (eds.), Crime: A spatial perspective. New York:
Columbia University Press.
 Hot spots policing as a model 243

(1981). Burglary in Philadelphia: A critique of an opportunity structure model.

In P. J. Brantingham and P. L. Brantingham (eds.), Environmental criminology
(pp. 189–201). Beverly Hills, CA: Sage Publications.
Sampson, R. J. (1985). Neighborhood and crime: The structural determinants
of personal victimization. Journal of Research in Crime and Delinquency, 22
(1), 7–40.
Sampson, R. J. and Groves, W. B. (1989). Community structure and crime:
Testing social disorganization theory. American Journal of Sociology, 94,
774–802.
Sampson, R. J. and Laub, J. H. (1993). Crime in the making: Pathways and turning
points through life. Cambridge, MA: Harvard University Press.
Sechrest, L. B., White, S. O., and Brown, E. O. (1979). The rehabilitation of
criminal offenders: Problems and prospects. Washington, DC: National Academy
of Sciences.
Shaw, J. (1995). Community policing against guns: Public opinion of the Kansas
City gun experiment. Justice Quarterly, 12, 695–710.
Shaw, C. R. and McKay, H. (1972). Delinquency and urban areas. Chicago:
University of Chicago Press.
Shaw, C.R. and Myers, E. D. (1929). The juvenile delinquent. In Illinois Crime
Survey. Chicago: Illinois Association for Criminal Justice.
Sherman, L. W. (1995). Hot spots of crime and criminal careers of places. In
J. E. Eck and D. Weisburd (eds.), Crime Prevention Studies, 4 (pp. 35–52).
Monsey, NY: Criminal Justice Press.
(1998). Evidence-based policing. Ideas in American policing series. Washington,
DC: The Police Foundation.
Sherman, L. W. and Rogan, D. (1995a). Effects of gun seizures on gun violence:
Hot spots patrol in Kansas City. Justice Quarterly, 12, 673–694.
(1995b). Deterrent effects of police raids on crack houses: A randomized con-
trolled experiment. Justice Quarterly, 12, 755–782.
Sherman, L. W. and Weisburd, D. (1995). General deterrent effects of police
patrol in crime “hot-spots”: A randomized controlled trial. Justice Quarterly,
12, 626–648.
Sherman, L. W., Gartin, P. R., and Buerger, M. E. (1989a). Repeat call address
policing: The Minneapolis RECAP experiment. Final report to the National
Institute of Justice. Washington, DC: Crime Control Institute.
(1989b). Hot spots of predatory crime: Routine activities and the criminology
of place. Criminology, 27, 27–56.
Spelman, W. and Brown, D. K. (1984). Calling the police: Citizen reporting of
serious crime. Washington: US Government Printing Office.
Sutherland, E. (1947). Principals of criminology. Chicago: J. B. Lippincott
Co.
Taylor, R. (1997). Social order and disorder of street blocks and neighborhoods:
Ecology, microecology, and the systemic model of social disorganization.
Journal of Research in Crime and Delinquency, 34 (1), 113–155.
Trasler, G. (1993). Conscience, opportunity, rational choice, and crime. In R.
V. Clarke and M. Felson (eds.), Routine activity and rational choice: Advances
in criminological theory, 5 (pp. 305–322). New Brunswick, NJ: Transaction
Publishers.
244 David Weisburd and Anthony A. Braga

Tyler, T. (2000). Social justice: Outcomes and procedures. International Journal

of Psychology, 35 (2), 117–125.
(2001). Public trust and confidence in legal authorities: What do majority
and minority groups members want from the law and legal institutions?
Behavioral Sciences and the Law, 19 (2), 215–235.
Visher, C. and Weisburd, D. (1998). Identifying what works: Recent trends in
crime prevention strategies. Crime, Law and Social Change, 28, 223–242.
Weisburd, D. and Green, L. (Green-Mazerolle) (1994). Defining the drug market:
The case of the Jersey City DMA system. In D. L., MacKenzie and C.D.
Uchida (eds.), Drugs and crime: Evaluating public policy initiatives. Newbury
Park, CA: Sage Publications.
(1995). Policing drug hot-spots: The Jersey City drug market analysis experi-
ment. Justice Quarterly, 12, 711–735.
Weisburd, D. and McEwen, T. (eds.). (1997). Crime mapping and crime prevention:
Crime prevention studies, 8. Monsey, NY: Criminal Justice Press.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder,
and fear? The Annals of the American Academy of Political and Social Science,
593, 42–65.
Weisburd, D. and Lum, C. (2005). The diffusion of computerized crime mapping
policing: Linking research and practice. In Police Practice and Research, 6 (5),
433–448.
Weisburd, D., Maher, L., and Sherman L.W. (1992). Contrasting crime gen-
eral and crime specific theory: The case of hot-spots of crime. Advances in
criminological theory, 4 (pp. 45–70). New Brunswick, NJ: Transaction Press.
Weisburd, D., Mastrofski, S., McNally, A. M., and Greenspan, R. (2001). Comp-
stat and organizational change: Findings from a national survey. Washington
DC: The Police Foundation.
Weisburd, D., Bushway, S., Lum, C., and Yang, S. M. (2004). Trajectories of
crime at places: A longitudinal study of street segments in the city of Seattle.
Criminology, 42 (2), 283–321.
Weisburd, D., Mastrofski, S. D., McNally, A. M., Greenspan, R., and Willis, J. J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2 (3), 421–456.
Weisburd, D., Wyckoff, L. A., Ready, J., Eck, J., Hinkle, J., and Gajewski, F.
(2004). Does crime just move around the corner? A study of displacement and
diffusion in Jersey City, NJ. Washington, DC: The Police Foundation.
Weiss, A. (2001). The police and road safety. Paper prepared for The National
Research Council Committee to Review Research on Police Policy and
Practice.
Whitehead, J. T. and Lab, S. P. (1989). A meta-analysis of juvenile correctional
treatment. Journal of Research in Crime and Delinquency, 26, 276–295.
Willis, J. J., Mastrofski, S. D., and Weisburd, D. (2004). Compstat and
bureaucracy: A case study of challenges and opportunities for change. Justice
Quarterly, 21 (3), 463–496.
Wolfgang, M. E. and Ferracuti, F. (1967). The subculture of violence: Toward an
integrated theory in criminology. New York: Tavistock.
13 Critic
The limits of hot spots policing

Dennis P. Rosenbaum

Introduction
This author is a strong advocate of using sophisticated information tech-
nology and the latest research findings to guide decisionmaking in police
organizations. Hence, this article begins with a brief acknowledgment of
the potential benefits of hot spot policing in theory, followed by a serious
critique. The thesis of this chapter is that, while the concept of hot spots
policing is attractive, we should be disappointed in how scholars have
narrowly defined it in theory and research and how police organizations
have narrowly practiced it. This approach has failed to embody the fun-
damental principles of either problem-oriented policing or community
policing, which many scholars believe represent the basic pillars of “good
policing” in the twenty-first century.

Acknowledging the benefits

The concept of hot spots is indisputable as a criminological phenomenon
and suggests the need for focused responses. From the very beginning
of criminological inquiries in nineteenth century France, scholars noted
that criminal activity is not randomly distributed, but rather varies by
geographic area such as regions, states, and communities (see Eck and
Weisburd 1995). More recent micro-level analyses have focused on
sizable variations between and within urban neighborhoods. Hence,
a sensible policy implication is to recommend the concentration of
more resources in these high-crime areas, including police resources
(Sherman, Gottfredson, MacKenzie et al. 1997). The most important
question, however, is not whether we should assign more resources to
problem areas, but rather, what resources should be deployed and how
should they be deployed?
Also, the concept of data-driven policing is difficult to dispute. Rely-
ing on information to make decisions about tactics, strategies, and pro-
grammatic interventions, assuming the data are accurate and complete, is

245
246 Dennis P. Rosenbaum

preferred to “cursing the dark” or making decisions primarily on the basis

of personal whim, personal experience, opinions of friends, or politi-
cal pressure. On the basis of controlled evaluation research, hot spots
researchers have encouraged the police to focus their attention on geo-
graphic areas smaller than police beats, including addresses, with a high
concentration of activity or repeat calls for service (e.g., Sherman, Gartin,
and Buerger 1989; Sherman and Weisburd 1995). Building on these stud-
ies and the revolution in information technology, including mapping pro-
grams (Mamalian and LaVigne 1999), the field has witnessed the strate-
gic deployment of police personnel in response to geo-based patterns of
crime incidents and calls for service. New York’s Compstat program was
a trendsetter in the 1990s and Chicago’s CLEAR program epitomizes the
IT capabilities of law enforcement for the next decade (Skogan, Steiner,
Hartnett et al. 2002).

Short-term impact
Hot spots policing, when implemented under controlled experimental
conditions with researchers involved, appears to have some effects on
crime and disorder (for reviews, see Sherman 1997; Taylor 1998; Braga
2001; Weisburd and Eck 2004). But the qualifications on this conclu-
sion are extremely important: first, the effects on crime are small and
not as consistent as the effects on disorder. Second, and most impor-
tantly, the effects dissipate quickly. Sherman’s (1990) review of the police
crackdown literature indicates that any residual deterrence effect is weak
and likely to decay rapidly. The conclusion regarding drug market crack-
downs is that they are ineffective in controlling drug hot spots. In one
of the stronger experimental tests, crime dropped on targeted blocks
in Kansas City after raids of crack houses, but returned within seven
days, leading the authors to conclude, “Like aspirin for arthritis, the
painkiller does nothing to remedy the underlying condition” (Sherman
and Rogan 1995a: 777). For directed patrols of high-crime locations, the
best research in Minneapolis suggests a positive relationship between the
length of patrol presence at a hot spot and the length of the deterrent
effect, up to 15 minutes, after which time, the effect reverses (Koper
1995).
The implication of these findings are that (1) the enormous expense of
concentrating police resources, especially for drug market crackdowns,
is difficult to justify given the cost of sustaining the effect; (2) a more
sophisticated understanding of deterrence is required before police are
ready to implement high-impact “schedules of punishment”; and (3) the
absence of larger and more sustained effects suggests the need for a
 The limits of hot spots policing 247

more complete understanding of the criminogenic forces at work in hot

spots. Much more evaluation research is needed in this area, especially
using randomized longitudinal designs.

Problem definition
What follows is a critique of hot spots policing from a problem-oriented
and community policing framework, beginning with the definition of the
problem. Engaging in “good” hot spot policing is not feasible if the hot
spot itself cannot be easily identified or well defined. The definitional
problem is complex and involves both conceptual and operational issues.
At the conceptual level, we need to ask, how does a particular place
achieve the status of being a “hot spot”? And who decides – the police on
the street, administrators or politicians, or the community? Generally, the
police have decided that a hot spot is a place where there are too many
violent crimes, drug deals, or gangs. But why is the definition of “the
problem” so narrowly construed? Undoubtedly, urban neighborhoods
have hot spots of public fear of crime, hot spots of public hostility toward
the police, of slum landlords, of racial profiling, disorder, weak informal
social control, institutional disinvestments, and weak interagency part-
nership, to name just a few. Yet these problems are not treated as hot
spots because they are not police priorities, because they are not well
measured or understood, and because they do not fit within the tradi-
tional definition of the police function.
The notion of hot spots, even if expanded, still limits the definition of
the problem to geographically linked phenomena. This approach over-
looks a number of serious crime-related problems that are not structured
in this way. Terrorism, computer crime, economic and international
crime, and even drug trafficking are examples of serious problems that
stretch beyond small geographic or neighborhood boundaries. Even gang
homicide, which is associated with geography, is best understood in terms
of the social structure of the gang and social interactions rather than loca-
tion (Papachristos 2003). In fact, the Boston Gun Project was a big suc-
cess because it focused on disrupting conflicts between gangs rather than
on the places where shots were fired (Braga, Kennedy, Piehl, and War-
ing 2001). Hence, place-based conceptions can sometimes restrict our
ability to understand and respond effectively to serious crime problems,
even those that cluster in space and time.
Even if we accept the traditional concept of hot spots, we still face
serious problems trying to identify and operationally define them. First,
the judgments of individual officers about hot spots can be inaccurate
if not supported by computer analysis of larger samples. Second, when
248 Dennis P. Rosenbaum

community input is sought, the police and local residents often disagree
when evaluating and prioritizing neighborhood problems (Skogan et al.
2002). Third, when GIS mapping software is employed, the hot spot
boundaries can be “fuzzy” (Taylor 1998). Circles can be imposed over
data plots, but in reality, these are somewhat arbitrary cut-offs. Where
does one hot spot end and another begin? In high-crime neighborhoods
this can be a serious problem. If a hot spot is enlarged, then what is the
benefit of a focused deployment scheme? If the hot spot is circumscribed
to a small area, the risk of making a false positive identification has been
increased. One can question whether small hot spots that come and go
quickly are sufficiently stable to warrant this label.
Finally, the selection of places as hot spots on the basis of extreme scores
(e.g., a spike in violence during the past month) can lead police managers
and researchers to draw false conclusions about the effectiveness of hot
spots policing. “Regression to the mean” – a statistical artifact that would
show up as a decline in the crime rate regardless of policing efforts – is
more likely under these circumstances, so a strong evaluation design is
needed to avoid making false causal inferences about the effectiveness of
intensive directed policing (see Shadish, Cook, and Campbell 2003).

Weak problem analysis

Identifying a hot spot is not the same as understanding it. The analysis
phase of problem-oriented policing is often lacking in hot spots policing.
Too often the data analysis team is satisfied with colorful crime maps as
the final product. Rarely do we see a detailed analysis of the character-
istics of the hot spot and the nature of the problem. How much can we
really learn about the problem from the spatial distribution of calls about
drug transactions, crime incidents, or arrests? A thorough and compre-
hensive analysis of the hot spot would require that these data be placed in
the larger environmental context. Knowing the physical and social milieu
is critical for understanding the factors that facilitate and constrain hot
spot behaviors. Census, housing, and survey data can be used to trian-
gulate police data. Interviews and observations of users of the hot spot
environment are essential (see Rosenbaum and Lavrakas 1995). The real
problems are hidden behind the calls for service or arrest data. The real
story is more complex, more dynamic, and more difficult to summarize.
Without digging deeper, the police responses will be standardized and
superficial, thus resulting in either short-term impact or no impact at all.
For the modern high-tech police organization, the information man-
agers believe that the most sophisticated type of hot spot analysis involves
using real-time data to engage in “crime forecasting” and deployment.
 The limits of hot spots policing 249

Analyses of monthly or weekly data are used to identify emerging hot spots
and deploy officers in a proactive, preventative manner. But statisticians
will caution against making this type of prediction because the estimates
are unstable within small geographic areas using small amounts of data
(Spelman 1995). Yet police commanders today can be chided for a single
shooting in a hot spot area.

Narrow and predictable response options

After the police (and hopefully, the community) identify and analyze the
hot spot, they are still facing the problem of what to do about it. On
this topic, only rarely have the police followed the guidance of Herman
Goldstein (1990: 102), the father of problem-oriented policing: “This
requires a process both broad and uninhibited – broad in that it breaks out
of the rigid mindset of the past, and uninhibited in that it explores sensible
responses without regard, at least initially, to potential impediments to
adopting them.” Instead, police departments have turned to what they
have done for years: patrols, sweeps, stakeouts, buy and busts, reverse
stings, etc. The responses are narrow and predictable – surveillance, stop
and frisk, question, and arrest – regardless of the nature and causes of
the problem. Goldstein (1990) encouraged a systematic inquiry into the
nature of these problems and the creation of strategic responses that
would likely be effective in reducing or eliminating them. Frankly, I do
not see this happening in American law enforcement. Problem analysis
tends to be superficial, non-existent, or based on a limited set of specific
criteria. Responses tend to be prepackaged, cookie-cutter reactions rather
than tailored, researched strategic plans for solving or eliminating the
problem over the long haul.
The simple fact that crime is concentrated in identifiable locations does
not justify the geographic concentration of limited police resources unless
the police have a compelling plan for dealing with the problem. What is
the plan? What is the theory of action behind the plan? Why should it
work? How long will it work? What are the potential adverse effects of
the plan? How is “success” defined and when is a problem “solved”?
Arguably, short-term reductions in crime and disorder are overrated and
give the false impression that the problem has been solved. This is a very
expensive way to run an organization.

The deterrence model

One major reason that hot spots policing tactics are not likely to have
a sustained impact on serious crime is because they are not based on
250 Dennis P. Rosenbaum

a compelling understanding of criminality or the larger hot spot envi-

ronment. The importance of place (hot spots) in crime causation and
prevention can be viewed through the lenses of rational choice theory
(Cornish and Clarke 1986), routine activity theory (Cohen and Felson
1979), crime pattern theory (Brantingham and Brantingham 1993),
or social disorganization theory (Bursik and Grasmick 1993; Sampson
2002). But hot spots policing, in practice, is not so sophisticated and
reflects a basic deterrence model. Directed patrols, undercover intelli-
gence gathering, visible surveillance systems (such as cameras), and var-
ious types of aggressive enforcement in the target areas are hypothesized
to deter offending by increasing the actual and perceived risk of detec-
tion, apprehension, and punishment. The preventative focus of hot spots
policing tends to be the potential offender’s fear of punishment and not
the other elements of criminal opportunity in the environment (e.g., vic-
tims, witnesses, social control agents, physical features) that influence
offender motivation. Repeat offenders often do not fear punishment and
have become increasingly sophisticated in avoiding it. Also, police sanc-
tioning of gang members can be viewed as a “badge of honor” (Klein
1999). Suffice it to say that subjective assessments of risk by potential
offenders (the key element!) are based on a variety of factors, including
the perceived certainty, severity, and speed of punishment (Tittle and
Paternoster 2000). Unfortunately, the criminal justice system has failed
to deliver on these threats in a consistent manner and has not demon-
strated an understanding of the complexity of the deterrence processes.
Deterrence theories are based on the assumption that people are ratio-
nal and accurate processors of information, but research suggests that we
are not good at estimating probabilities of events, and we often ignore or
misperceive information presented to us (e.g., Kahneman and Tversky
1973; Nisbett and Ross 1980; Kahneman, Slovic, and Tversky 1982).
This is not to say that people are always irrational. In general, individuals
have their own rationality and logic that is not always evident to those
empowered to administer formal sanctions. Selling drugs or carrying a
gun on the streets may seem “irrational” to the average person, but these
behaviors can seem very sensible to the person who faces daily threats of
bodily injury or theft of contraband, sees no good employment opportu-
nities, and has nothing to lose from another arrest.

Community and problem-oriented theories

Hot spots policing, in practice, is like old wine in new bottles. Police
continue to do what they do best – undercover and visible enforcement
activities – but with greater efficiency and focus on specific locations.
 The limits of hot spots policing 251

Police administrators ask themselves: How can we deploy more police

officers (and the right officers) to these locations to make more arrests
and seize more contraband? The real question should be, What are the
best strategies to combat crime, disorder, and quality of life in these hot
spots given all that we know about these problems, these locations, and the
many resources that can be leveraged (including non-police resources)?
Theories of community policing (e.g., Skogan 1990; Greene 2000;
Skogan 2003), problem-oriented policing (Goldstein 1990), and commu-
nity crime prevention (Tonry and Farrington 1995; Rosenbaum, Lurigio,
and Davis 1998) have, to all intents and purposes, been ignored in hot
spots policing, despite advocates’ claims to the contrary. These models all
suggest that police would be unwise to limit their focus to pictures of where
crime is occurring, and should give more attention to why it is occurring.
Hot spots policing somehow leaves the impression that our knowledge
of hot spots is limited to GIS plots of crime incidents. In fact, we know
a lot more about hot spots, and this knowledge could be exploited to
develop lasting solutions to neighborhood problems. Criminologists have
examined the role of the physical environment, families, peers, schools,
neighborhood resources, housing policies, labor markets, social organiza-
tion processes, public attitudes, gun markets, offender re-entry problems,
and many other factors that contribute to crime, disorder, and deviance
within specific geographies (see Tonry and Farrington 1995; Tittle and
Paternoster 2000; Wilson and Petersilia 2002). Our knowledge of how to
prevent crime is substantial! The only question here is whether the police
have any role in converting this knowledge into practice. The argument
here is that they do.

Police as experts on crime

Over the years, police organizations have acquired an image of “effec-
tive crime fighters” and “experts” on crime in general, thus receiving the
lion’s share of taxpayer dollars for public safety programs. To justify this
reputation, law enforcement must demonstrate a deeper knowledge of the
forces that contribute to crime and the quality of urban life. Educated
people can disagree about whether the police have any responsibility to
reach beyond short-term strategies to address the underlying, chronic
causes of crime. (Most police officers would say, “that’s not our job.”) I
would argue that law enforcement leaders can play an important role by
(1) working with criminologists to educate policymakers and the public
about what can, and should be, done to prevent crime; (2) creating and
leading multiagency partnerships that have a higher probability of yield-
ing a sustainable impact on crime; and (3) focusing on comprehensive
252 Dennis P. Rosenbaum

strategies that attack crime at all levels (Rosenbaum 2002; Schuck and
Rosenbaum forthcoming). Unfortunately, hot spots policing can rein-
force the persistent public misperception that police can solve the crime
problem alone, without the help of the communities they serve, other
government agencies, or the private sector.
A close adherence to the principles of problem-oriented policing would,
I believe, lead most police administrators to question current stand-alone
hot spots practices. The urban problems of youth violence cannot be
explained by simple statements such as, “kids make bad choices and they
need to pay the price” or “our job is to take the scumbags and gang
bangers off the street.” The solutions are much more complex and,
furthermore, most convicted offenders will return to the neighborhood
within a year or two. The goal of problem-oriented policing is to eliminate
the problem, not to increase the efficiency of the response. Furthermore,
a major goal of community policing is to achieve fair and equitable polic-
ing through community input and feedback, not just to achieve efficient
policing (Eck and Rosenbaum 1994). Hot spots policing, as currently
practiced, may be at odds with these goals for reasons noted below.

Adverse effects of hot spots policing

Arguably, hot spots policing is not only ineffective at solving persistent
problems, but is potentially harmful to both targeted and non-targeted
communities. Some of the potential untoward effects of hot spots policing
are discussed in this section.

Displacement effects
One of the big question marks surrounding hot spots policing is about
possible displacement effects. If hot spots policing is simply altering crim-
inal activity by location, time, modus operandi, or type of offense, rather
than preventing it, then the collective benefits to the larger community
are non-existent. The problem of displacement, if occurring, may indi-
cate that non-targeted residents or locations are suffering more because
additional criminal activity is being pushed into their environment.
Criminologists are divided on this issue and the evidence is mixed.
Most evaluations do not define and measure displacement adequately if
at all, and most research designs are biased in favor of the null hypoth-
esis (Weisburd and Green 1995a). Some studies suggest there is evi-
dence of diffusion of crime control benefits to nearby areas (see Sherman
and Rogan 1995b; Weisburd and Green 1995b; Green-Mazerolle and
Roehl 1998; Braga et al. 1999), but some of these effects may reflect a
 The limits of hot spots policing 253

mis-specification of the target area boundaries. On the other side of the

fence, there is considerable evidence of spatial displacement of calls or
crime incidents as a result of police crackdowns, especially during drug
enforcement (e.g., Kleiman 1988; Potter, Gaines, and Holbrook 1990;
Smith, Sviridoff, Sadd et al. 1992; Uchida, Forst, and Annan 1992;
Kennedy 1993; Hope 1994; Sherman and Rogan 1995a; Braga et al.
1999; Maher and Dixon 2001). A meta-analysis by Sherman (1997),
however, suggests that displacement effects are not as large as crime pre-
vention effects, but the research in this field is inconclusive. Clearly, there
are many types of displacement that have not been measured.

Police–community relations
Hot spots policing, because it has been operationally defined as aggres-
sive enforcement in specific areas, runs the risk of weakening police–
community relations (see Kleiman 1988; Worden, Bynum, and Frank
1994; Sherman 1997; Rosenbaum et al. 1998). Hot spots policing can
easily become zero tolerance policing and broken windows policing since
these are models that police find easy to adopt. These tactics can drive
a wedge between the police and the community, as the latter can begin
to feel like targets rather than partners. Because the police have chosen
to focus on removing the “bad element” and serving as the “thin blue
line” between “good” and “bad” residents, these strategies can pit one
segment of the community against another, as the “good” residents are
asked to serve as the informants and the “eyes and ears” of police. Par-
ents, siblings, and friends of gang members and drug dealers can feel a
divided loyalty and be caught in the crossfire.
The success of police organizations depends largely on the cooperation
of the citizenry, but the legitimacy of the institution is compromised when
the public’s trust and confidence in the police is undermined. The con-
sequences are enormous, ranging from lawsuits to a declining willingness
to obey the law (e.g., Tyler 1990; 2001). What determines public trust in
the police? The answer is complex (see Weitzer and Tuch in press). Cer-
tainly, perceptions of police effectiveness in lowering crime rates is one
factor that affects attitudes about the police, but research indicates that it
is not as important as procedural justice during the exercise of authority
(Skogan in press; Tyler in press). That is, positive attitudes about the
police drop when citizens feel that they have been treated unfairly, dis-
respected, not listened to, or physically abused during encounters with
the police. Minorities, who are much more likely than non-minorities to
live in hot spot areas, express these sentiments at much higher rates than
254 Dennis P. Rosenbaum

whites (Rosenbaum, Hawkins, Costello, Skogan et al. 2005; Skogan in

press).
We must acknowledge that residents of high-crime areas are very
ambivalent about aggressive enforcement. Many demand it and are
pleased to see the police restore order. In fact, surveys indicate that many
are willing to give up their civil liberties to achieve this sense of security,
however fleeting (Rosenbaum 1993). Residents will insist on aggressive
enforcement up to the point where it directly affects them, their family,
or their friends, who frequently end up in jail and prison or report being
mistreated by the police. How much of this policing they will tolerate
remains to be seen. Like other Americans, however, hot spots residents
should have the opportunity to experience both safety and liberty and not
be required to choose one or the other. This topic demands much more
careful policy deliberation and research.

Abusive policing
Hot spots policing, as practiced, runs the risk of becoming abusive and/or
corrupt policing. When officers feel the pressure to make arrests, seize
drugs, and seize guns, some will be inclined to cut corners and, as a result,
every officer’s credibility is compromised. Complaints about excessive
force and police corruption are not uncommon in hot spots neighbor-
hoods where police are sometimes viewed as an “occupying force.” The
issue of due process is critical to the criminal justice system. A twenty-
year prosecutor recently summarized the problem for me in this way:
“Winking at questionable stops and arrests may serve to get major waves
of contraband off the streets for a while but it eventually begins to draw
the ire of the judiciary and threatens the credibility of both the police and
that of the prosecuting authorities” (Andreou, personal communication,
2003).

Stigma from labeling

There is a real risk that neighborhoods and smaller areas identified as
hot spots will acquire a more negative image as a result of the labeling
process. In some cases, the stigma of being a hot spot of crime may
stimulate greater fear of crime among residents (both inside and outside
the neighborhood) and eventually lower property values. Granted, some
hot spots already have a bad reputation, but others on the margins could
be further damaged by the labeling process. Yet police face a “Catch
22.” The adverse effects of labeling could be minimized by keeping the
identity of the location confidential and internal to the police, but that
 The limits of hot spots policing 255

would prevent the department from soliciting community support for

aggressive police interventions.

Policing bias by race and class

We cannot avoid the fact that policing tactics and strategies vary by race
and social class. These differences are due, in part, to data-driven deploy-
ment of police officers on the basis of crime hot spots. Violence and illicit
drug activity are most visible in low-income, minority communities and,
therefore, these areas receive a disproportionate share of police attention
in response to public demand. By definition, the pressure on the police
to lower crime rates is greatest in these hot spot areas.
The problem is that civil liberties are more easily jeopardized in low-
income and minority neighborhoods where residents feel disenfranchised
and do not have easy access to legal remedies when feeling mistreated.
Minority communities are the primary focus of not only drug and gun
enforcement, but also minor disorders (e.g., hanging out, public drink-
ing) and various types of traffic enforcement activities (e.g., road checks
for seat belt usage and drinking). Following the lead of New York City
and the “broken windows” model, a commonly employed strategy is to
use these police-initiated contacts (often involving minor infractions) as
a tool to identify weapons, drugs, guns, and persons with outstanding
warrants. One problem is that the “hit rate” is extremely low, so the vast
majority of persons who are inconvenienced (if not offended) by these
stops are innocent persons of color and limited means.
Beyond inconvenience is the troubling question of police abuse, which,
as it turns out, is not randomly distributed. Looking at twenty years of
data in New York, one study concluded that police misconduct seems to
be attracted to neighborhoods with structural disadvantage, population
mobility, and increases in the Latino population, among other factors
(Kane 2002). So, neighborhoods that suffer the most from crime are also
areas where the police are more likely to violate the rules of conduct.
Finally, there is the macro-level question regarding the long-term con-
sequences of applying greater enforcement resources to minority commu-
nities. The bottom line, after the criminal justice system has completed its
work, is called “disproportionate minority confinement.” Minorities are
being confined and incarcerated at much higher rates than non-minorities
(for a review, see Pope, Lovell, and Hsia 2003), which will continue to
be a hotly debated political issue. This begs the question of whether we
can solve these neighborhood problems differently.
In sum, the style and consequences of policing in low-income minority
communities are often different than in middle-class neighborhoods and
256 Dennis P. Rosenbaum

the application of enforcement tactics in specific locations is increasingly

driven by hot spot analyses.

Direct and opportunity costs

Intensifying police resources in hot spots can be expensive in terms of
added personnel costs and additional activity for the criminal justice sys-
tem. The options are to increase the police budget, borrow resources from
other neighborhoods, or discontinue the program after a short period.
Intensive policing is simply not affordable in many cases and, therefore,
not sustainable in the absence of careful planning about deployment.
A bigger concern is the opportunity costs associated with hot spots
policing. In other words, what is the police department not doing because
it is involved in hot spots policing? First, there is a tendency to reduce
police resources in low and moderate crime neighborhoods by creating
special teams, transferring personnel, or redrawing beat boundaries. Low
crime communities may feel shortchanged, but moderate crime commu-
nities may experience crime increases. Second, there may be a tendency
to reduce police resources for community policing and joint problem-
solving activities.
Opportunity costs come in many sizes and shapes. One could argue
that hot spots policing discourages an intelligent public policy debate
about long-term solutions to our crime problems. As noted earlier, this
type of policing is designed to leave the public with the impression that
crime is under control, thus contributing to the image of law enforcement
agencies as effective crime-fighting machines.

Unanticipated effects on crime

Arrest (and for some, conviction and incarceration) does not always have
the desired deterrent effects, and may even produce boomerang effects. A
couple of examples will suffice. First, there is some evidence that arrest-
ing inner-city youths can have a criminogenic effect of increasing, rather
than reducing, the probability of recidivism (Klein 1986). Second, there
is evidence that having a criminal record reduces one’s probability of
finding gainful employment (Bushway 1996). The plight of 600,000 ex-
offenders who return from prison each year illustrates the many legal and
social obstacles they face to successful reintegration (Travis, Solomon,
and Waul 2001), thus explaining why most return to a life of crime. Also,
the extent to which law enforcement agencies use technology to track and
monitor ex-offenders by address is unprecedented.
In sum, our criminal justice system may be having criminogenic effects,
making crime more probable for those who are targeted, arrested, and
 The limits of hot spots policing 257

labeled (many for life) as “criminals.” This is especially troubling for

the large volume of incidents involving misdemeanor and non-violent
drug offenses and raises the question of whether aggressive enforcement
activity in high-crime neighborhoods is helping or hurting public safety
in the long run. Someday, we will take the time to answer these questions.

Collective efficacy
Neighborhoods with high rates of violent crime that are typically the target
of hot spots policing also suffer from social disorganization and a lack of
collective efficacy in solving problems (Sampson, Raudenbush, and Earls
1997). “Get tough” policies run the risk of further undermining social
control and a community’s capacity for self-regulation. By strengthening
the hand of the police to solve crime-related problems, the residents may
feel less empowered to solve neighborhood problems. Policymakers may
need to be reminded that community crime rates, in the final analysis,
are influenced more by the social ecology of communities than by for-
mal control mechanisms (Sampson 2002). Hence, within the community
policing framework, a primary mission of police organizations should be
to work with other organizations to design strategies that strengthen com-
munity capacity rather than to supplant or weaken the role of community.
But in fairness to the police, restoring order would seem to be a neces-
sary element in the process of restoring community capacity. The only
question is about the means of restoring order, the timeframe, and the
strategies that will accompany these police activities.

Rethinking hot spots policing

If police can rethink the concept of place, the notion of hot spots polic-
ing can be given new meaning. I would encourage police departments to
adopt a more holistic approach to problems that cluster in space and time,
including: (1) conducting an in-depth, comprehensive analysis of the hot
spot environment and the many factors that are responsible for making
the area “hot”; (2) exploring a wide range of alternative solutions that
might reasonably be expected to have either a short-term or long-term
impact on the problem (both are important!); (3) attempting to build real
partnerships among government, private sector, and community organi-
zations that have a stake in public safety; and (4) giving more attention to
the prevention of crime at the individual, family, and community levels.
The Chicago Police Department (CPD) is an interesting case study that
seeks to achieve some of these objectives. While investing in enforcement-
oriented hot spots policing, the CPD is also pursuing a more holistic
258 Dennis P. Rosenbaum

approach to hot spots. First, in selected hot spot locations, the CPD has
sought to expand the responsibility for crime control to third parties, such
as landlords and business owners, using both rewards and sanctions to
clean up entire blocks. Second, the superintendent’s office has also taken
a leadership role in building a coalition of organizations to address hot
spots of violence in selected areas of the city and to provide feedback on
police performance. Third, Chicago has a nationally recognized commu-
nity policing effort (CAPS) that allows community concerns to be aired
in monthly beat meetings. As part of this process, the CPD is working
with the University of Illinois at Chicago to develop a geo-based Internet
survey that will “measure what matters” to the public, including indica-
tors of hot spots of disorder, fear, and strained police–community rela-
tions, among other factors. These new data are expected to result in new
definitions of local problems, a stronger community policing/problem-
solving process, and greater accountability for all parties (Rosenbaum
2004). Thus, police departments can supplement traditional hot spots
enforcement with community policing and/or problem-oriented polic-
ing approaches (for other examples, see Weisburd and Green 1995b;
Eck and Wartell 1996; Green-Mazerolle and Roehl 1998; Braga et al.
1999).
In this chapter I have suggested that the concept of hot spots, despite
its theoretical and empirical attractiveness, can create problems that were
unintended. Police administrators and researchers need to exercise cau-
tion when labeling and responding to locations in a narrow manner, given
the political and social sensitivities associated with hot spots. Aggressive
enforcement has its place in the arsenal of urban policing and is essential
for providing short-term relief to distressed areas, but it should not be
used as a stand-alone strategy or as society’s primary answer to crime.
Also, it requires close supervision of police officers and regular feedback
from the community members to prevent abuses. Most importantly, given
the central role of urban law enforcement agencies in public policy anal-
ysis, the argument put forth here is that the majority of tax-based police
resources should be devoted to strategies that (1) show promise for long-
term impact on rates of community crime and the quality of community
life; (2) reflect the equitable distribution of police resources based on
human need; and (3) embody the police organization’s commitment to
the fair and impartial treatment of all segments of society.
In essence, hot spots policing, as currently practiced, does not ade-
quately address the historical, economic, political, social, and cultural
dimensions of these target areas, all of which contribute to crime rates.
The solution is not as simple as getting local residents, under the
threat of arrest, to make better choices or accept responsibility for their
 The limits of hot spots policing 259

behavior. These assumptions are overly simplistic and do not acknowl-

edge the loss of hope and despair among these residents, their disconnec-
tion from conventional institutions, the stigma and rejection that result
from having a criminal record, the neighborhood’s loss of public and pri-
vate resources to support healthy families, and the fact that dozens of
other social problems (beyond crime) cluster in these hot spots. Hence,
real solutions in these hot spots will require sophisticated research and
intelligence, strategic planning, and comprehensive prevention programs
involving many entities working in concert. I believe that senior police
officials are capable of meeting this challenge by playing a lead role in
multiagency partnerships and accepting the full responsibility that comes
with being “the crime experts.”

  
The author would like to thank the following individuals for reviewing and
providing helpful comments on previous drafts of this manuscript: Anthony
Braga, Susan Hartnett, Amie Schuck, Cody Stephens, and David Weisburd.


Braga, A. A. (2001). The effects of hot spots policing on crime. Annals of American
Political and Social Science, 578, 104–125.
Braga, A. A., Kennedy, D. M., Piehl, A. M., and Waring, E. J. (2001). Reducing
gun violence: The Boston Gun Project’s Operation Ceasefire. Washington, DC:
Department of Justice, National Institute of Justice.
Braga, A., Weisburd, D., Waring, E., Green Mazerolle, L., Spelman, W., and
Gajewski, F. (1999). Problem-oriented policing in violent crime places: A
randomized controlled experiment. Criminology, 37 (3), 541–580.
Brantingham, P. L. and Brantingham, P. J. (1993). Environment, routine, and
situation: Toward a pattern theory of crime. In R. V. Clarke and M. Felson
(eds.), Routine activity and rational choice. Advances in Criminological Theory,
5. New Brunswick, NJ: Transaction Publications.
Bursik, R. J., Jr. and Grasmick, H. G. (1993). Neighborhoods and crime: The dimen-
sions of effective community control. New York: Lexington.
Bushway, S. (1996). The impact of a criminal history record on access to legiti-
mate employment. Unpublished doctoral dissertation. H. John Heinz School
of Public Policy and Management, Carnegie Mellon University.
Cohen, L. E. and Felson, M. (1979). Social change and crime rate trends: A
routine activity approach. American Sociological Review, 44, 588–605.
Cornish, D. and Clarke, R. V. (eds.) (1986). The reasoning criminal: Rational choice
perspectives on offending. New York: Springer-Verlag.
Eck, J. E. and Rosenbaum, D. P. (1994). The new police order: Effectiveness,
equity, and efficiency in community policing. In D. P. Rosenbaum (ed.), The
challenge of community policing: Testing the promises (pp. 3–23). Newbury Park,
CA: Sage Publications.
260 Dennis P. Rosenbaum

Eck, J. E. and Weisburd, D. (1995). Crime places in crime theory. In J. E. Eck

and D. Weisburd (eds.), Crime and place, crime prevention studies 4 (pp. 1–33).
Monsey, NY: Criminal Justice Press.
Eck, J. E. and Wartell, J. (1996). Reducing crime and drug dealing by improving
place management: A randomized experiment. Report to the San Diego Police
Department. Washington, DC: Crime Control Institute.
Goldstein, H. (1990). Problem-oriented policing. New York: McGraw-Hill.
Green, J. R. (2000). Community policing in America: Changing the nature, struc-
ture, and function of the police. In J. Horney, J. Martin, D. L. MacKenzie,
R. Peterson, and D. P. Rosenbaum (eds.), Policies, processes, and decisions of
the criminal justice system, 3, Criminal Justice 2000 series. Washington, DC:
US Department of Justice, National Institute of Justice.
Green-Mazerolle, L. and Roehl, J. (1998). Civil remedies and crime prevention.
Munsey, NJ: Criminal Justice Press.
Hope, T. (1994). Problem-oriented policing and drug market locations: Three
case studies. In R. V. Clarke (ed.), Crime prevention studies, 2. Monsey,
New York: Criminal Justice Press.
Kahneman, D., Slovic, P., and Tversky, A. (1982). Judgment under uncertainty:
Heuristics and biases. Cambridge: Cambridge University Press.
Kahneman, D. and Tversky, A. (1973). On the psychology of prediction. Psycho-
logical Review, 80, 237–251.
Kane, R. J. (2002). The social ecology of police misconduct. Criminology, 40 (4),
867–896.
Kennedy, D. (1993). Closing the market: Controlling the drug trade in Tampa,
Florida. NIJ Program Focus. Washington, DC: National Institute of Justice,
US Department of Justice.
Kleiman, M. (1988). Street-level drug enforcement: Examining the issues. Washing-
ton, DC: National Institute of Justice, US Department of Justice.
Klein, M. W. (1986). Labeling theory and delinquency policy: An empirical test.
Criminal Justice and Behavior, 13, 47–79.
(1999). Attempting gang control by suppression: The misuse of deterrence
principles. In L. K. Gaines and W. Cordner (eds.), Policing perspectives: An
anthology (ch. 18, pp. 269–282). Los Angeles, CA: Roxbury.
Koper, C. (1995). Just enough police presence: Reducing crime and disorderly
behavior by optimizing patrol time in crime hot spots. Unpublished paper.
Maher, L. and Dixon, D. (2001). The cost of crackdowns: Policing Cabramatta’s
heroin market. Current Issues in Criminal Justice, 13 (1), 5–22.
Mamalian, C. A. and LaVigne, N. G. (1999). The use of computerized mapping
by law enforcement. Washington, DC: National Institute of Justice, Office of
Justice Programs, US Department of Justice.
Nisbett, R. E. and Ross, L. (1980). Human inference: Strategies and shortcomings
of social judgment. Englewood Cliffs, NJ: Prentice-Hall.
Papachristos, A. V. (2003). The social structure of gang homicide in Chicago.
Paper presented at the Annual Conference of the American Society of Crim-
inology, November 21, Denver, Colorado.
Pope, C., Lovell, R., and Hsia, H. M. (2003). Disproportionate minority
confinement: A review of research literature from 1989 to 2001. Washington,
 The limits of hot spots policing 261

DC: Office of Juvenile Justice and Delinquency Prevention, Office of Justice

Programs, US Department of Justice.
Potter, G., Gaines, L., and Holbrook, B. (1990). Blowing smoke: An evaluation of
marijuana eradication in Kentucky. American Journal of Police, 9 (1), 97–116.
Rosenbaum, D. P. (1993). Civil liberties and aggressive enforcement: Balancing
the rights of individuals and society in the drug war. In R. C. Davis, A. J.
Lurigio, and D. P. Rosenbaum (eds.). Drugs and the community (pp. 55–82).
Springfield, IL: Charles C. Thomas.
(2002). Evaluating multi-agency anti-crime partnerships: Theory, design and
measurement issues. Crime Prevention Studies, 14, 171–225.
(2004). Community policing and web-based communication: Addressing the
new information imperative. In L. A. Fridell and M. A. Wycoff (eds.), Com-
munity policing: The past, present and future. Washington, DC: Police Executive
Research Forum.
Rosenbaum, D. P. and Lavrakas, P. J. (1995). Self-reports about place: The
application of survey and interview methods to the study of small areas. In
J. E. Eck and D. Weisburd (eds.), Crime and place: Crime prevention studies 4
(pp. 285–314). Monsey, NY: Criminal Justice Press.
Rosenbaum, D. P., Lurigio, A. J., and Davis, R. C. (1998). The prevention of crime:
Social and situational strategies. Belmont, CA: Wadsworth.
Rosenbaum, D. P., Hawkins, D. F., Costello, S. K., Skogan, W. G., Roman
Rivera, L., Vera, C., Rokita, R., Ring, M. K., Larson, T., and Munansangu,
M. (2005). Race and police: A matter of public trust. Report to the National
Institute of Justice, Office of Justice Programs, US Department of Justice.
Sampson, R. J. (2002). The community. In J. Q. Wilson and J. Petersilia (eds.),
Crime: Public policies for crime control (pp. 225–252). Oakland, CA: Institute
for Contemporary Studies Press.
Sampson, R. J., Raudenbush, S. W., and Earls, F. (1997). Neighborhoods and
violent crime: A multilevel study of collective efficacy. Science, 277, 918–924.
Schuck, A. M. and Rosenbaum, D. P. (forthcoming). Promoting safe and healthy
neighborhoods: What research tells us about intervention. In K. Fulbright-
Anderson (ed.), Community change: Theories, practices and evidence. Washing-
ton, DC: The Aspen Institute.
Shadish, W. R., Cook, T. D., and Campbell, D. T. (2002). Experimental and
quasi-experimental designs for generalized causal inference. Boston: Houghton
Mifflin.
Sherman, L.W. (1990). Police crackdowns: Initial and residual deterrence. In
M. Tonry and N. Morris (eds.), Crime and justice: A review of research, 12.
Chicago: University of Chicago Press.
(1997). Policing for crime prevention. In L. W. Sherman, D. Gottfredson,
D. MacKenzie, J. E. Eck, P. Reuter, and S. Bushway. Preventing crime: What
works, what doesn’t, what’s promising (ch. 8). Report to the National Institute of
Justice, Office of Justice Programs, US Department of Justice (NCJ 165366).
Sherman, L. W. and Rogan, D. P. (1995a). Deterrent effects of police raids on
crackdowns: A randomized, controlled experiment. Justice Quarterly 12 (4),
755–781.
262 Dennis P. Rosenbaum

(1995b). Effects of gun seizures on gun violence: “Hot spots” patrol in Kansas
City. Justice Quarterly 12 (4), 673–693.
Sherman, L. W. and Weisburd, D. (1995). General deterrent effects of police
patrol in crime “Hot spots”: A randomized, controlled trial. Justice Quarterly
12 (4), 625–648.
Sherman, L. W., Gartin, P. R., and Buerger, M. E. (1989). Hot spots of predatory
crime: Routine activities and the criminology of place. Criminology, 27, 27–
55.
Sherman, L. W., Gottfredson, D., MacKenzie, D., Eck, J., Reuter, P., and
Bushway, S. (1997). Preventing crime: What works, what doesn’t, what’s promis-
ing. Report to the National Institute of Justice, Office of Justice Programs,
US Department of Justice. (NCJ 165366).
Skogan, W. G. (1990). Disorder and decline: Crime and the spiral of decay in American
neighborhoods. New York: Free Press.
(ed.). (2003). Community policing: Can it work? Belmont, CA: Wadsworth.
(In press). Citizen satisfaction with police encounters. Police Quarterly.
Skogan, W. G., Steiner, L., Hartnett, S. M., DuBois, J., Bennis, J., Rottinghaus,
B., Young Kim, S., Van, K., and Rosenbaum, D. P. (2002). Community
policing in Chicago, years eight and nine: An evaluation of Chicago’s Alterna-
tive Policing Strategy and information technology initiative. Final report to the
Illinois Criminal Justice Information Authority. Evanston, IL: Northwestern
University, Institute for Policy Research. December.
Smith, M., Sviridoff, M., Sadd, S., Curtis, R., and Grinc, R. (1992). The neigh-
borhood effects of street-level drug enforcement. Tactical Narcotics Teams in New
York: An evaluation of TNT. New York: Vera Institute of Justice.
Spelman, W. (1995). Criminal careers of public places. In J. E. Eck and D.
Weisburd (eds.) Crime and place: Crime prevention studies, 4 (pp. 115–144).
Monsey, NY: Criminal Justice Press.
Taylor, R. B. (1998). Crime and small-scale places: What we know, what we can
prevent, and what else we need to know. Crime and place: Plenary papers of the
1997 conference on criminal justice research and evaluation. Washington, DC:
National Institute of Justice, Office of Justice Programs, US Department of
Justice. (NIJ 168618).
Tittle, C. R. and Paternoster, R. (2000). Social deviance and crime: An organiza-
tional and theoretical approach. Los Angeles, CA: Roxbury.
Tonry, M. and Farrington, D. P. (eds.). (1995). Building a safer society: Strategic
approaches to crime prevention. Crime and Justice: A Review of Research, 19.
Travis, J., Solomon, A. L., and Waul, M. (2001). From prison to home: The dimen-
sions and consequences of prisoner reentry. Washington, DC: The Urban Insti-
tute.
Tyler, T. R. (1990). Why people obey the law. New Haven: Yale University Press.
(2001). Trust and law-abiding behavior: Building better relationships between
the police, the courts, and the minority community. Boston University Law
Review 81, 361–406.
(In press). Policing in black and white: Ethnic group differences in trust and
confidence in the police. Police Quarterly.
Uchida, C., Forst, B., and Annan, S. (1992). Modern policing and the control of
illegal drugs: Testing new strategies in two American cities. Washington, DC:
 The limits of hot spots policing 263

National Institute of Justice, Office of Justice Programs, US Department of

Justice.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder,
and fear? Annals of the American Academy of Political and Social Science, 593,
42–65.
Weisburd, D. and Green, L. (1995a). Measuring immediate spatial displacement:
Methodological issues and problems. In J. E. Eck and D. Weisburd (eds.),
Crime and place: Crime prevention studies, 4 (pp. 349–361). Monsey, NY:
Criminal Justice Press.
(1995b). Policing drug hot spots: The Jersey City Drug Market Analysis Exper-
iment. Justice Quarterly, 12 (4), 711–735.
Weitzer, R. and Tuch, S. A. (In press). Determinants of public satisfaction with
the police. Police Quarterly.
Wilson, J. Q. and Petersilia, J. (eds.). (2002). Crime: Public policies for crime control.
Oakland, CA: Institute for Contemporary Studies Press.
Worden, R., Bynum, T., and Frank, J. (1994). Police crackdowns on drug abuse
and trafficking. In D. McKenzie and C. Uchida (eds.), Drugs and crime:
Evaluating public policy initiatives. Thousand Oaks, CA: Sage.
Part VII

Compstat
14 Advocate
Compstat’s innovation

Eli B. Silverman

Introduction
Compstat tributes are extensive. Compstat has been described as
“perhaps the single most important organizational/administrative inno-
vation in policing during the latter half of the 20th century” (Kelling
and Sousa 2001: 6). A Criminology and Public Policy Journal editor
recently termed Compstat “arguably one of the most significant strategic
innovations in policing in the last couple of decades” (Criminology and
Public Policy 2003: 419). The authors of a major study note that Comp-
stat “has already been recognized as a major innovation in Ameri-
can policing” (Weisburd, Mastrofski, McNally et al. 2003: 422). In
1996, Compstat was awarded the prestigious Innovations in American
Government Award from the Ford Foundation and the John F. Kennedy
School of Government at Harvard University. Former Mayor Giuliani
proclaims Compstat as his administration’s “crown jewel” (Giuliani
2002: 7).
Why the praise, what are they specifically praising and is this praise
warranted? These questions constitute the core of this chapter which
maintains that Compstat praise, criticism, and replication are frequently
based on a superficial understanding of its proper development, imple-
mentation, and many dimensions. The literature inadequately reflects
how Compstat’s successful implementation and maintenance is often
incomplete when it lacks substantial organizational revamping and proper
managerial preparation. This contributes to an insufficient appreciation
of Compstat’s array of attributes. In addition, there is often a lack of
understanding of how any particular Compstat may reflect the organi-
zational and managerial arrangements of an individual law enforcement
agency at any specific time.
Exploration of these points will center on: Compstat’s many facets, its
origins, the reasons for and the nature of its replication in numerous ver-
sions and venues, and its very positive strengths as well as its prospective
drawbacks.

267
268 Eli B. Silverman

Understanding Compstat
Compstat is most frequently understood by its most visible elements
today. These include: up-to-date computerized crime data, crime analy-
sis, and advanced crime mapping as the bases for regularized, interactive
crime strategy meetings which hold managers accountable for specific
crime strategies and solutions in their areas.

Familiar explanations
It is fair to say that the widespread diffusion of Compstat refers to these
most noticeable elements. Since Compstat was first unveiled by the New
York City Police Department (NYPD) in 1994, a Police Foundation 1999
survey for the National Institute of Justice (NIJ) revealed that a third of the
nation’s 515 largest police departments had implemented a Compstat-
like program by 2001 and 20 percent were planning to do so. The same
survey found that about 70 percent of police departments with Comp-
stat programs reported attending a NYPD Compstat meeting (Weisburd,
Mastrofski, McNally, and Greenspan 2001). My own research indicates
that very few of the over 250 outsiders who attended Compstat meetings
between 1994 and September 1997 were exposed to any Compstat ele-
ments other than the meetings (with the exception of a NYPD booklet
on Compstat). It is unlikely that the Police Foundation’s 70 percent dif-
fered much in their exposure to Compstat. (There may, in fact, be some
overlap between the two groups.)
This process is continuing. Gootman reported that 219 police agency
representatives visited NYPD Compstat meetings in 1998, 221 in 1999,
and 235 in the first ten months of 2000 (Gootman 2000: B1). Attendance
at a Compstat meeting, while a useful introduction, does not provide adequate
preparation for introducing and establishing Compstat. In fact, it may be mis-
leading because attendees often become mesmerized by the flashy over-
head display of multiple crime maps synchronized with technologically
advanced portrayals of computerized crime statistics.

The lure of Compstat

The first three years of the NYPD’s Compstat corresponded with dra-
matic declines in the city’s crime rate. According to the FBI’s Unified
Crime Reports, the city’s 12 percent decline in index crime in1994 (com-
pared to a national drop of less than 2 percent) grew to 16 percent in 1995
and yielded another 16 percent in 1996. These decreases accounted for
more than 60 percent of the national decline during this period.
 Compstat’s innovation 269

While these figures, of course, do not prove a causal relationship, they

received extraordinary law enforcement and national attention. The New
York Model (Compstat) was offered as the road to rapid crime reduction
(Gootman 2000: B1). A Time magazine 1996 observation is still appli-
cable today. “Compstat has become the Lourdes of policing, drawing
pilgrim cops from around the world . . . for a taste of New York’s magic”
(Pooley 1996: 55–56).
Too often, therefore, Compstat has been interpreted as primarily a meeting
with a statistical computer program which, when it generates accurate and timely
crime statistics, transforms a traditional bureaucracy into a flexible, adaptable
police agency geared to effective crime control strategies. In the vernacular, it is
only necessary to display computer-generated crime maps and pressure
commanders in order “to make the dots go away” (Maple 1999: 38).
This superficial approach is emblematic of the quick managerial fix
approach, thus contributing to the misunderstanding and misapplication
of Compstat.
This Compstat allure of crime reduction through technological
advancement is reflected in the in-depth study of the Lowell Police
Department Compstat:

What police department, however, would not want to adopt a program whose
clear purpose is to reduce crime through the implementation of a well-defined
set of technologies and procedures? The appeal of Compstat’s crime fighting
goal to the police increases the likelihood that it will endure. (Willis, Mastrofski,
Weisburd, and Greenspan 2003: 11)

The full Compstat

Compstat, however, is a far more complex product of changes in manage-
rial and organizational arrangements, including flattening, decentraliza-
tion, greater personnel authority, discretion and autonomy, geographic
managerial accountably, and enhanced problem solving. Based on the
New York experience, it is my view that Compstat cannot be a fully viable
entity if the above administrative, managerial, and operational activities
do not precede it.
It is worthwhile noting that the Police Foundation study’s “six key
elements” essential to Compstat contain similar components. They
are “mission clarification, internal accountability, geographic organi-
zation of operational command, organizational flexibility, data driven
problem identification and assessment and innovative problem solv-
ing tactics, and external information exchange” (Weisburd et al. 2003:
427).
270 Eli B. Silverman

It is equally noteworthy that the Police Foundation study found many

of these key elements lacking in many police Compstat programs. In
their comparison of Compstat and non-Compstat agencies, the study
concludes that the Compstat agencies “have opted for a model much
heavier on control than on empowerment” (Weisburd et al. 2003: 448).
Moreover, despite its virtues, the authors found that:
Compstat agencies were largely indistinguishable from non-Compstat agencies
on measures that gauged geographic organization of command, organizational
flexibility, the time availability of data, and the selection and implementation of
innovative strategies and tactics . . . Compstat departments are more reluctant to
relinquish power that would decentralize some key elements of decisionmaking
geographically . . . enhance flexibility, and risk going outside of the standard
tool kit of police tactics and strategies. The combined effect overall, whether or
not intended, is to reinforce a traditional bureaucratic model of command and
control. (Weisburd et al. 2003: 448)
But I find this conclusion less than surprising. The study’s Compstat
programs are self-designated. There is no evidence that these police agen-
cies underwent the self-diagnosis, reengineering, and organizational and
managerial overhaul processes that preceded the New York Compstat
experience.
Thus a more complete understanding of Compstat may be gained
through a review of the context and origins of New York’s Compstat
in order to fully appreciate its positive qualities for modern day policing.
Failure to grasp the differences between popular accounts and Comp-
stat’s actual origins can deflect attention away from Compstat’s merits
while activating undesirable features.

Immediate origins
In the immediate sense, contrary to a widely held view, Compstat was
not a preordained planned system of managerial supervision, account-
ability, and strategic policing. Its first meeting, as a matter of fact, was
almost serendipitous. Upon taking office in 1994, Commissioner Bratton
called for a weekly, one-on-one, current events briefing with a represen-
tative from each of the NYPD’s eight bureaus during the early months
of his administration. Deputy Commissioner Maple authorized the head
of the Patrol Bureau to discuss crime statistics with the commissioner.
A disturbing reality surfaced: the NYPD did not know its current crime
statistics; there was a reporting time lag of three to six months.
Maple, in conjunction with other key people pressed the precincts to
generate crime activity statistics on a weekly basis. During the second
week of February 1994, all precincts provided a hand count of the seven
major crimes for the first six weeks of 1993 compared to the same period
 Compstat’s innovation 271

in 1994. The Patrol Bureau’s staff computerized this crime activity and
assembled it into a document referred to as the “Compstat book.” The
first Compstat book included current data on a year-to-date basis for
crime complaints and arrests for every major felony category, as well as
gun arrests, compiled on citywide, patrol borough, and precinct levels.
Contrary to most accounts, the acronym Compstat is not short for
“computer statistics.” Compstat actually arose from a computer file,
“compare stats,” in which the data was originally stored. Compstat,
then, was simply short for “compare stats.” This distinction is not
trivial since the “computer statistics” interpretation frequently suggests
that an advanced statistical computer program is synonymous with effec-
tive crime control, further contributing to Compstat’s misapplication.
In fact, at the outset, Compstat was based on an elementary
database, created in a set of desktop office software called Smart Ware,
from Informix. This was later replaced, for a considerable period, by
Microsoft’s very basic FoxPro database for businesses to enter all the
statistics into files.
In New York, regularized Compstat meetings grew out of a need for a
mechanism to ensure precinct COs’ accountability and to improve perfor-
mance. In April 1994, the leadership was searching for ways to sharpen
the NYPD’s crime-fighting focus. At that time, for example, boroughs
held monthly field robbery meetings in which precinct COs and rob-
bery and anti-crime sergeants met with the borough staffs (to whom they
reported) to discuss robbery trends.
Top-level executives requested that the Brooklyn North patrol bor-
ough hold its monthly robbery meeting at headquarters – One Police
Plaza. After the Brooklyn borough CO’s overview of special conditions,
several precinct COs were called to the front of the room. Their pre-
sentations, although suitable for public community council meetings,
lacked in-depth analyses of complex crime problems. Dissatisfied, the
chief abruptly terminated the meeting and announced monthly headquar-
ters meetings for each borough. The Compstat process was launched;
there was no turning back.
Late in April 1994 the NYPD leadership decided to use these head-
quarters meetings to link the newly released drug and gun strategies with
the Compstat books. Tenacity was essential. The meetings, held twice
weekly, became mandatory. They began promptly at 7:00 a.m., a time
when there are likely to be few distractions.

Organization and managerial foundations

Compstat’s roots, however, are far deeper than the scenario described
above. Many managerial and organizational interventions laid the
272 Eli B. Silverman

foundation for Compstat. The groundwork for Compstat centered on

early developments prior to the Bratton administration (Silverman 1999:
21–66). Even during the Bratton administration which began in 1994,
all interventions took place before Compstat’s introduction. Early in 1994, the
NYPD began to redesign its organizational structure, employing manage-
ment strategies designed to re-engineer its business processes and create
a “flatter” organizational structure based on geographic decentralization,
teamwork, information sharing, and managerial accountability (Silver-
man and O’Connell 1999).
Information flow, for example, was eased by the elimination of the level
of division which, prior to 1994, was interposed between the precincts and
the boroughs. Each of the twenty-three divisions was responsible for from
two to four precincts. Since the divisions’ responsibilities were primarily
administrative, the new organizational arrangement smoothed relation-
ships between newly beefed up precinct and borough responsibilities and
capabilities.
The rapid redesign of the department’s organizational architecture was
based upon the concept of continuous performance improvement charac-
terized by clearly delineated objective standards, benchmarking, sharing
of “best practice,” and the development and analysis of timely and accu-
rate information to manage change.
Modern management provided the orchestral score; “reengineering”
was its name. Contemporary management literature explains that reengi-
neering requires “radical change,” a “starting over” throughout the
entire organization, nothing less than a “reinvention of how organizations
work.” Commissioner Bratton insisted: “We reengineered the NYPD into
an organization capable of supporting our goals” (Bratton 1996: 1).
Reengineering acted like a booster cable to the NYPD’s battery, pro-
viding the cranking power needed to activate decentralization and com-
mand accountability. Relinquishing control of daily ground operations
was the most fundamental yet difficult challenge facing the new admin-
istration. Traditionally, the person at the apex of the NYPD pyramid
would retain control through standardized procedures and policies. But in
order to hold precinct commanders accountable for crime prevention, the
new leadership knew the organization must grant them more discretion.
Rather than allow headquarters to determine staffing and deployment on
a citywide basis, it was decided that reducing crime, fear of crime, and
disorder would flow from patrol borough and precinct coordination of
selected enforcement efforts.
In essence, the NYPD was able to achieve what Peters and Waterman
label “simultaneous loose-tight properties,” meaning the “co-existence
of firm central direction and maximum individual autonomy” (Peters
 Compstat’s innovation 273

and Waterman 1982: 318). So while the top NYPD levels were now
fortified with greater detail, this knowledge extends beyond activities such
as arrest particulars to more informative data such as the characteristics,
times, and locations of precinct-selected enforcement strategies and their
relationship to crime reduction.
These reengineering structural, operational, and strategic reconfig-
urations stemmed from a so-called “cultural diagnostic” which rested
on deliberations emanating from multirank and functional NYPD focus
groups. These groups examined various dimensions of NYPD missions
and strategies and the obstacles which hindered strategy effectiveness and
goal achievement.
Consequently, the reengineering reports questioned the NYPD’s oper-
ating procedures. What current policy yields, they claimed, was inade-
quate. The precinct organization report, for example, noted:
2 or 3 percent reduction in crime is not good enough. We need to change the orga-
nization to do more. The need to reengineer precincts is not immediately appar-
ent. City-wide crime continues to decline year after year. Every annual precinct
state of command report, without exception, includes evidence of neighborhood
improvements. Bureau and Special Unit Commanders to a man, or a woman,
will vigorously defend the effectiveness of the present system. Why fix what’s not
broken?
The answer is in the new mission of the department to dramatically reduce
crime, fear, and disorder. Slow, continuous improvement doesn’t cut it, and that
is all the present system can deliver . . . [R]eengineering in its simplest forms
means starting all over, starting from scratch. (New York City Police Department
1994: v)

Selling Compstat
What started out as a computer file and a book to satisfy crime infor-
mational needs has evolved and been reconstructed into a multifaceted
forum for coordinated, reenergized, and accountable organizational
crime-fighting strategies. Its strength lies in its adaptability and com-
pliance mechanisms. It is vitally important to recognize that Compstat’s
initial and prime raison d’être was and is to measure and hold managers
accountable for performance. In Moore’s words:
It becomes a powerful managerial system in part because the technical capacity
of the system allows it to produce accurate information on important dimensions
of performance at a level that coincides with a particular manager’s domain of
responsibility . . . [Compstat] is, in the end, primarily a performance measurement
system. (Moore 2003: 470, 472; see also Walsh 2001)

All these characteristics have contributed to its widespread replication

in various formats in numerous locations. The claims and counterclaims
274 Eli B. Silverman

are numerous and often difficult to evaluate since Compstat has fre-
quently been associated with and introduced at the same time as other
law enforcement and societal changes.
The boldest claims assert that Compstat plays a significant role in
crime reduction due to its accountability and decentralized components.
Kelling and Sousa include Compstat in their analysis of NYPD effective
crime reduction policing in six precincts. They conclude that

Both the problem-solving strategy and the notion of accountability came to

fruition in weekly NYPD headquarters meetings known as Compstat . . . But
the true effectiveness of Compstat lies in its ability to drive the development
of crime reduction tactics at the precinct level. By making precinct commanders
accountable, centralized Compstat allows the problem solving strategy to operate
in a decentralized manner. (Kelling and Sousa 2001: 17)

Compstat crime reduction efficacy is also frequently advocated by

police administrators, several of whom moved from the NYPD to head
other city police departments. Compstat’s introduction in New Orleans,
for example, corresponded with a decline in murders from 421 in
1994, diving 55 percent in 1999 to 162. Minneapolis’ Compstat’s ver-
sion, CODEFOR (Computer Optimized Deployment-Focus on Results)
has been credited for a double digit decrease in homicides, aggravated
assaults, robberies, burglaries, and auto thefts between 1998 and 1999
(Anderson 2001: 4). In 2000 Compstat was introduced in Baltimore by its
new chief, a former NYPD deputy police commissioner. By the end of the
year, the city experienced its first below 300 homicides in 20 years accom-
panied by an overall crime drop of 25 percent (Anderson 2001: 4; Clines
2001: 15; Weissenstein 2003: 27). Between 1999 and 2001, Baltimore’s
overall violent crime declined 24 percent, homicides dropped 15 percent,
shootings fell 34 percent, robberies dropped 28 percent, rapes 20 percent,
and assaults 21 percent (Henry and Bratton 2002: 307). Philadelphia’s
former police commissioner, another former NYPD deputy police com-
missioner, attributed a decline in the city’s crime to Compstat-driven
policing. “Social conditions in the city have not changed radically in the
two years and we have the same police department, the same number of
officers. Nothing has changed but how we deploy them and utilize them”
(Anderson 2001: 3).
Similar crime reduction assertions have been made for police agencies
around the world. Two Australian scholars, for example, recently pub-
lished an evaluation of the New South Wales Compstat-modeled Opera-
tion and Crime Review (OCM). The authors found that this process was
effective in reducing three of the four offence categories studied (Chilvers
and Weatherburn 2004).
 Compstat’s innovation 275

Although these claims are repeated elsewhere, some scholars are not
convinced. While Eck and Maguire acknowledge that “homicide rates
have declined significantly from 1994 to 1997, following the implemen-
tation of Compstat” in the United States, they raise serious reservations
(Eck and Maguire 2000: 231). First, they correctly assert that “Comp-
stat was implemented along with a number of changes in the NYPD . . .
Consequently it is difficult to attribute any reductions in crime to specific
police changes” (Eck and Maguire 2000: 230).
Beyond this valid point, the relationships are even more complicated.
A full Compstat program, as discussed earlier, rests upon and ulti-
mately encompasses many changes including police deployment driven
by more accurate Compstat crime mapping. The efficacy of these strate-
gies depends on the extent to which Compstat can permeate the agency’s
entire operations. So some Compstats may be viewed as “Compstat Lite”
while others may be more fully developed, embracing many dimensions of
Compstat’s crime strategy management. Compstat’s prime architect, the
late Jack Maple, referred to the pale Compstat imitations as the “knock off
versions” whereby administrators “just sit in a circle and chat about the
intelligence” (Maple 1999: 47). This vivid Compstat contrast explains
why, in my view, Eck and Maguire are compelled to maintain:

We do not know, however, whether this [homicide] reduction is produced through

general deterrence of all those who frequent hot spots, through specific deterrence
of hot-spot offenders who come under closer scrutiny of the police, or through
incapacitation of repeat offenders following their arrest at these hot spots. So,
there is an array of possible mechanisms through which directed patrol can reduce
crime at hot spots. (Eck and Maguire 2000: 231)

Finally, Eck and Maguire find that the rate of homicide reductions
in other cities does not lend credence to the independent impact of
Compstat on homicide rates. Yet they acknowledge that these findings
are tenuous:

Our analysis does not find that Compstat is ineffective. These explorations can
not be interpreted as a rigorous exploration of Compstat . . . It is possible, given
the complexity of homicide patterns, that the Compstat process had a subtle and
even meaningful but difficult to detect effect on crime in New York. (Eck and
Maguire 2000: 235)

Less controversial, however, is Compstat’s surging popularity and

adoption by numerous agencies and jurisdictions seeking to more
effectively administer their operations and hold managers accountable.
Omaha’s year-old Compstat is credited with improving cooperation
among all units. “Instead of one unit tackling a problem, everyone gets
276 Eli B. Silverman

involved . . . The entire culture has changed now” (Law Enforcement News,
March 2004: 11).
In 1996, New York City’s Corrections Department modeled its
TEAMS (Total Efficiency Accountability Management System) program
on Compstat with an examination of the department’s “most fundamen-
tal practices and procedures” (O’Connell 2001: 17). Again, accountabil-
ity is a major theme which, when fused with more accurate and timely
statistical reporting and analysis and interunit cooperation, has been
credited with a dramatic reduction in inmate violence. Between 1995
and 1999, stabbings and slashing declined from 1093 to 70 (Anderson
2001: 3).
TEAMS has evolved and now addresses more than just jail violence.
Its accountability system has been continually expanded to retrieve and
assess almost 600 performance indicators addressing such issues as
religious service attendance, maintenance work orders, health care, over-
time, compliance with food service regulations, completed searches con-
ducted, and the performance of personnel who have been the subject of
the department’s civility tests (O’Connell and Straub 1999a; 1999b;
1999c).
Since TEAMS is oriented to system-wide innovation, its impact
has contributed to a “shift in organizational culture and philosophy”
(O’Connell 2001: 19). A former commissioner observed:

The key is really the way that it assists us in using proactive and creative manage-
ment. It expands possibilities and get more people involved in the decisionmaking
process. Our people don’t just think of themselves as corrections officers anymore.
Now they see themselves as managers. (O’Connell 2001: 19)

Compstat accountability mechanisms, long a staple of the private sec-

tor, have also become increasingly attractive to non-law-enforcement
public agencies. There are numerous examples. The New York City
Department of Parks and Recreation, for instance, developed its own ver-
sion of Compstat, calling it Parkstat. When Parks officials visited NYPD
Compstat meetings in 1997, they realized that they could utilize this sys-
tem to develop and refine their Parks Inspection Program (PIP) which
oversees the maintenance and operation of over 28,000 acres of property
throughout New York City. Now parks department data analysis and
managerial accountability are combined with monthly meetings to assess
overall conditions, cleanliness of structural features such as benches,
fences, sidewalks, play equipment, and landscape features such as trees,
athletic fields, and water bodies (O’Connell 2001: 20). The percentage of
parks rated acceptably clean and safe increased from 47 percent in 1993
 Compstat’s innovation 277

to 86 percent in 2001 (Webber and Robinson 2003: 3). One observer’s

assessment of Parkstat ranks it comparable to Compstat’s high rating:

As with Compstat and TEAMS, an emphasis is placed on pattern or trend iden-

tification. All performance data are viewed through three lenses, which look for
district wide trends, borough wide comparison and trends, and citywide compar-
ison and trends. . .
Parkstat serves many functions, not the least of which is the fostering of organi-
zational learning. Senior administrators learn about what is occurring in the field
at the same time that districts are learning from one another. Effective practices
are openly shared and disseminated throughout the entire organization.
The Parkstat program is continually developing. Indeed, the department
recently renamed it Parkstat Plus and has expanded it to include a broader range
of performance measures . . . relating to personnel, vehicle maintenance, resource
allocation and enforcement activity to ensure superior service delivery. Parkstat
stands as an excellent example of how the Compstat model can be adopted and
successfully implemented outside the field of criminal justice. (O’Connell 2001:
21, 22)

Perhaps the most ambitious extension of Compstat’s managerial

accountability and informational exchange processes began in the City of
Baltimore in mid 2000 when its mayor was delighted with the results of
the Baltimore Police Department’s first year with Compstat. Baltimore’s
program, called Citistat (first developed by Compstat architect, the late
Jack Maple) is an attempt to evaluate and coordinate performance on a
citywide basis whereby supervisors report every two weeks (as opposed
to the previous quarterly basis) on their departments’ performance.
Citistat’s timely data permits the assessment and coordination of diverse
social services dealing with graffiti, abandoned vehicles, vacant housing,
lead paint abatement, urban blight, drugs, and drug treatment. Discus-
sions are based on up-to-date information. Citistat meetings are similar
to those of Compstat whereby data, graphs, and maps are projected to
track and display department performance.
So far, the city is pleased with Citistat’s development. There has been
a 40 percent reduction in payroll overtime with savings of over $15 mil-
lion over two years. Its director of operations maintains that “The charts,
maps and pictures tell a story of performance, and those managers are
held accountable” (Webber and Robinson 2003: 4). The fact that the
prestigious Innovations in American Government Award was awarded
to Baltimore’s Citistat ten years after NYPD’s Compstat received the
same award speaks to the enduring concepts embedded in this man-
agerial and organizational approach. Speaking for Harvard University’s
Kennedy School, Stephen Goldsmith stated: “Citistat is a management
tool for public officials that translates into real, tangible results for
278 Eli B. Silverman

citizens. Government leaders across the country and around the world
are taking notice of its success – and for good reason” (July 28, 2004).
Citistat, like most Compstat-type programs, seeks to lower the infor-
mational barriers that generally hinder intra and interagency collabora-
tion. Baltimore is constantly expanding the number of agencies included
in the data analyses and its Citistat meetings. It appears that Citistat
is the ultimate test of Compstat’s ability to serve as the informational
cement of reform, the central mechanism that provides communication
links to traditionally isolated specialized units. Fragmentation plagues
many organizations. Harvard management expert Rosabeth Kanter calls
it “segmentalism” and notes, “The failure of many organization-change
efforts has more to do with the lack . . . of an integrating, institution-
alizing mechanism than with inherent problems in an innovation
itself” (Kanter 1983: 301). Without Compstat, fragmentation would
continue to rule supreme. Compstat’s confrontation of informational
splintering can be indispensable to organizational well-being. Comp-
stat can serve as the organizational glue that bonds many changes
together.

Maintaining Compstat
Like any administrative-managerial-technological innovation, Compstat
is not an unalloyed asset when either of two conditions is not met. The
first pertains to the proper and full fundamental organizational and man-
agerial reforms necessary to establish Compstat as an innovation. The
absence of these reforms yields a Compstat that is more in name than in
substance. The second refers to the obligation to maintain these struc-
tural and managerial underpinnings.
Since the first condition was previously discussed (see pp. 271–273),
I now turn to the second condition of Compstat maintenance. Just as it
is mistaken to assume that all Compstats are the same, it is equally erro-
neous to suppose that all Compstats will receive steady attention to their
reform underpinnings. The vitality of a well-functioning Compstat rests
on its connections to, indeed its centrality to, a reenergized, restructured,
and adaptable problem-solving law enforcement organization.
Compstat’s proper introduction should not be confused with long-term
preservation. As a study of Compstat and organizational change in one
law enforcement agency observed: “police organizations are notoriously
resistant to change and any subsequent changes usually require many
years to take effect” (Willis et al. 2003: 58). Properly installing and main-
taining Compstat, therefore, requires vigilance.
 Compstat’s innovation 279

There are serious liabilities when Compstat is not properly imple-

mented or sustained. Compstat can be easily engulfed by a culture of
top-down directives dominated by numbers reflecting the predominant
control system of the particular agency.
When Compstat performance and accountability measurement
becomes excessively supervised, whether within a highly centralized orga-
nization or from external hierarchical organizations, the consequences
can be alarming. Subordinate units will naturally concentrate on those
items being measured. Or, as the saying goes, what gets measured gets
done. This can lead to crime statistics manipulation and/or downgrad-
ing which has been recently reported in numerous locales including
Philadelphia, Atlanta (Hart, 2004: 6), New Orleans (Ritea, 2003a:
1; Ritea, 2003b: 1), New York (Parascandola and Levitt, 2004: 5;
Gardiner and Levitt, 2003: 8) and Broward County, Florida (Hernan-
dez, O’Boye, and O’Neill, 2004: 9). Any performance measurement
instrument, including Compstat, has the capacity to aggravate this con-
dition, posing serious problems for law enforcement’s statistical accuracy
(Manning 2001: 331–333; Willis et al. 2003: 13).
Thus it is crucial to recognize that Compstat is not some distinct iden-
tity isolated from the law enforcement agency in which it functions. Mea-
surement of its success cannot be divorced from the structure in which it
operates since it often reflects the prevailing organizational and political
climate. To view Compstat in any other way is to reify it and endow it
with elements beyond its control.
Compstat’s evolving role in department priorities can be observed even
in the police departments where it gained the greatest notice. Over the
years, the NYPD’s Compstat has become increasingly centralized with
specialized units more likely directed from above:
In effect a centralization thrust has been superimposed on decentralized reforms.
But centralization now has a powerful weapon in its arsenal – Compstat.
Compstat, in many senses, has been turned on its head. Instead of a tool to
reevaluate objectives and tactics and scan the environment for future trends, the
information from computer-generated comparative statistics is becoming known
for only its most visible aspects – crime mapping and deployment activity. Greater
information now is used to bear down on the management of many street oper-
ations. Numbers, sometimes any numbers, rule the day.
The original technocratic components of Compstat have been reified and dupli-
cated, while its analytical underpinnings dwindle. In this mix, organizational
learning often gets lost in the shuffle. . .
And what happens when painstaking and constant lessons of implementation
are neglected for the quick fix? The short-haul dominates. (Silverman 2001: 206–
213)
280 Eli B. Silverman

Conclusion: Compstat as an innovation

Despite its liabilities in specific circumstances, Compstat constitutes a
significant law enforcement innovation. At its best, Compstat is a unique
blending of performance measurement, computer technology, crime
mapping, data analysis, information sharing, managerial accountability,
interactive strategy meetings, and organizational cohesiveness and learn-
ing. It is easy to forget the freshness of Commissioner Bratton’s approach
over ten years ago when entrance to his higher echelon was restricted to
commanders committed to double-digit crime reduction. Establishing a
specific objective – a 10 percent reduction in crime for 1994 – was the ini-
tial propellant for change. While target-setting is the norm for the private
sector, usually it is anathema for public organizations because it offers a
yardstick against which performance can be more accurately measured
and, if deficient, condemned.
The need to instill and recognize good performance in public policing
is a vital managerial imperative. Certainly, the call for improved police
performance grew out of felt necessities. For both real and/or imagined
reasons, there was wide discontent with the effectiveness and manage-
ment of police performance.
For Compstat, the question, then, is not should performance be mea-
sured. The questions are: what performance should be measured (Jones
and Silverman 1984); how should performance measurement be imple-
mented, assessed, and maintained; and by whom; and, more importantly,
in what type of organizational and political context.
In New York, as in many other jurisdictions, Compstat originally
focused on crime statistics and police arrest and quality of life enforce-
ment activities. But Compstat, per se, is not inherently restricted to these
measurements. Compstat’s beauty lies in its versatility and adaptability.
In New York, for example, Compstat, at various times, has included
citizen satisfaction survey statistics, precinct commander community
meeting activities, citizen complaints, and domestic violence issues and
data.
Furthermore, as previously discussed (see pp. 276–278) Compstat
lends itself to a variety of law enforcement and non-law-enforcement
contexts. Numerous additional agencies are currently adopting their own
versions of Compstat. These include New York’s Office of Health Insur-
ance with its Healthstat designed to assist uninsured New Yorkers enroll
in a publicly funded health insurance program. “In the first 18 months,
participating agencies enrolled about 340,000 eligible New Yorkers”
(Webber and Robinson, 2003: 4). The Department of Transporta-
tion instituted MOVE, an accountability and performance management
 Compstat’s innovation 281

system that meets twice a month to assess operational performance. The

cities of Miami and Pittsburgh are pursuing comprehensive Compstat-
like systems similar to Baltimore’s Citistat.
Societal needs for information sharing, data analysis, and effective
organizational-managerial performance will only continue to prolifer-
ate and even expand Compstat’s rapid diffusion. Two and a half years
before the 9/11 terrorist attacks on the United States, Compstat’s archi-
tect, the late Jack Maple, was asked about the future of Compstat. He
replied:

This should not be limited to the police department. It should involve every
city agency, the fire department, the building department, the transportation
department; everybody should be contributing and coordinating. And other law
enforcement agencies need to participate fully. The FBI, DEA and ATF offices
in a city should be running their own numbers and then bring those to Compstat
meetings at the police department. (Dussault 1999: 2)

Now in this post 9/11 era, there are numerous calls to overcome insti-
tutional barriers by federalizing the Compstat process in order to combat
terrorism.
The intelligence and accountability mechanism known as Compstat is
tailor made for combating terrorism. Applying this to America’s new war,
however, requires solving one of the most enduring problems in policing:
turf jealousy, especially between the FBI and local law enforcement agen-
cies.

The FBI’s anti-terrorism efforts should be Compstated in every city where the
bureau operates. Where a Joint Terrorism Task Force exists, the commanders
of the agencies should meet on a biweekly basis to interrogate task force mem-
bers about the progress of their investigations. Where JTTFs don’t exist, the FBI
should assemble comparable meetings with all relevant agency heads. The new
Fedstat meetings would have two purposes: to ensure that each ongoing investi-
gation is being relentlessly and competently pursued, and to share intelligence.
(MacDonald 2001: 27)

Compstat has provided significant advances in policing and organiza-

tional performance. It enables key decisionmakers to engage in face-to-
face discussions of issues and proactive practices that draw upon the
collective expertise of the entire organization. It has developed as an
interactive management device that enables the organization to learn,
teach, supervise, and evaluate personnel in one central forum. Compstat’s
multifaceted possibilities embrace an array of positive outcomes. It is
incumbent upon effective leadership to seize upon this potential.
282 Eli B. Silverman


Anderson, D. C. (2001). Crime control by the numbers: Compstat yields new
lessons for the police and the replication of a good idea. Ford Foundation
Report. New York.
Bowerman, M. and Ball, A. (2000). Great expectations: Benchmarking for best
value. Public Money and Management, 20 (2), 21–26.
Bratton, W. (1996, August 1). Management secrets of a crime-fighter extraordi-
naire. Bottom Line.
Chilvers, M. and Weatherburn, D. (2004). The New South Wales Compstat
process: Its impact on crime. Australian and New Zealand Journal of Crimi-
nology.
Clines, F. X. (2001, January 3). Baltimore gladly breaks 10 year homicide streak.
New York Times, A11.
Dussault, R. (1999). Jack Maple: Betting on intelligence. Internet, gov-
tech.net/publications, April.
Eck, J. E. and Maguire, E. R. (2000). Have changes in policing reduced violent
crime? An assessment of the evidence. In A. Blumstein and J. Wallman (eds.),
The crime drop in America (pp. 207–65). Cambridge: Cambridge University
Press.
Gardiner, S. and Levitt, L. (2003, June 21). NYPD: Some crime stats misclassi-
fied. NewYork Newsday, 11.
Giuliani, R. W. (2002). Leadership. New York: Hyperion.
Goldsmith, S. (2004). The innovations in American government awards. Cambridge,
MA: John F. Kennedy School of Government, Harvard University.
Gootman, E. (2000, October 24). A police department’s growing allure: Crime
fighters from around world visit for tips. New York Times, B1.
Hart, A. (2004, February 21). Report finds Atlanta police cut figures on crimes.
The New York Times, 26.
Henry, V. E. and Bratton, W. J. (2002). The Compstat paradigm: Management
accountability in policing, business and the public sector. New York: Looseleaf
Law Publications.
Hernandez, J., O’Boye, S., and O’Neill, A. W. (2004, February 25). Sheriff’s
office scrutinizes its crime data bonuses tied to performance under review.
Sun-Sentinel, A1, 6.
Jones, S. and Silverman, E. (1984). What price efficiency. Policing, 1 (1), 31–
48.
Kanter, R. M. (1983). The change masters. New York: Simon and Schuster.
Kelling, G. L. and Sousa, W. H. (2001). Do police matter? An analysis of the impact
of New York City’s police reforms. Civic Report No. 22. New York: Manhattan
Institute.
Law Enforcement News. (2004). Compstat is doing more than just driving down
Omaha’s rate. March.
MacDonald, H. (2001). Keeping New York safe from terrorism. City Journal.
Autumn.
Manning, P. K. (2001). Theorizing policing: The drama and myth of crime con-
trol in the NYPD. Theoretical Criminology, 5 (3), 315–344.
Maple, J. (1999). Crime fighter. New York: Broadway.
 Compstat’s innovation 283

Moore, M. (2003). Sizing up Compstat: An important administrative innovation

in policing. Criminology and Public Policy, 2 (3), 469–494.
New York City Police Department. (1994). Re-engineering team. Precinct Orga-
nization. New York.
O’Connell, P. E. (2001). Using performance data for accountability. Arlington, VA:
Price Waterhouse Coopers.
O’Connell, P. E. and Straub, F. (1999a). Why the jails didn’t explode. City Jour-
nal, 2, Spring, 28–37.
(1999b). Managing jails with T.E.A.M.S. American Jail, March/April, 48–54.
(1999c). For jail management, Compstat’s a keeper. Law Enforcement News, 9.
Parascandola, R. and Levitt, L. (2004, March 22). Police statistics: Numbers
scrutinized. New York Newsday, 14.
Peters, T. and Waterman, R. H. (1982). In search of excellence. New York: Warner.
Pooley, E. (1996). One good apple. Time, January, 55–56.
Ritea, S. (2003a, October 24). Five N.O. officers fired over altered crime stats.
The Times Picayune, 1.
(2003b, October 25). Crime, coercion and cover-up. The Times-Picayune, 1.
Silverman, E. B. (1999). NYPD battles crime: Innovative strategies in policing.
Boston: Northeastern University Press.
(2001). Epilogue. NYPD battles crime: Innovative strategies in policing. Boston:
Northeastern University Press.
Silverman, E. B. and O’Connell, P. (1999). Organizational change and decision-
making in the New York City Police Department. International Journal of
Public Administration, 22 (2).
Walsh, W. F. (2001). Compstat: An analysis of an emerging police managerial
paradigm. Policing: An International Journal of Police Strategies and Manage-
ment, 24(3), 347–362.
Webber, R. and Robinson, G. (2003). Compstamania. Gotham Gazette. New
York: Citizens Union, July 7.
Weisburd, D., Mastrofski, S. D., McNally, A. M., and Greenspan, R. (2001).
Compstat and organizational change: Findings from a national survey. Report
submitted to the National Institute of Justice by the Police Foundation.
Weisburd, D., Mastrofski, S. D., McNally, A. M., Greenspan, R., and Willis, J. J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2 (3), 421–456.
Weissenstein, M. (2003). Call on NY’s top cops: NYPD brass recruited by other
cities to lower crime rates. Newsday, January 2.
Willis, J. J., Mastrofski, S. D., Weisburd, D., and Greenspan, R. (2003). Comp-
stat and organizational change in the Lowell police department: Challenges and
opportunities. Washington, DC: Police Foundation.
15 Critic
Changing everything so that everything
can remain the same: Compstat and
American policing

David Weisburd, Stephen D. Mastrofski,

James J. Willis, and Rosann Greenspan

Compstat has come to be seen as a major innovation in American polic-

ing. It has received national awards from Harvard University and for-
mer Vice President Gore, and has been featured prominently along with
William Bratton (the police administrator who created the program)
in the national news media. Its originators and proponents have given
Compstat credit for impressive reductions in crime and improvements in
neighborhood quality of life in a number of cities that have adopted the
program (Silverman 1996; Remnick 1997; Gurwitt 1998; Bratton 1999).
And while introduced only in 1994 in New York City, police departments
around the country have begun to adopt Compstat or variations of it
(Law Enforcement News 1997; Maas 1998; McDonald 1998; Weisburd,
Mastrofski, McNally et al. 2003). Indeed, a Police Foundation survey sug-
gests that Compstat had literally burst onto the American police scene.
Only six years after Compstat emerged in New York City, more than a
third of American police agencies with 100 or more sworn officers claimed
to have implemented a Compstat-like program (Weisburd, Mastrofski,
McNally, and Greenspan 2001).
Drawing from a series of studies we conducted at the Police Founda-
tion (Weisburd et al. 2001; Greenspan, Mastrofski, and Weisburd 2003;
Weisburd et al. 2003; Willis, Mastrofski, Weisburd, and Greenspan 2004;
Willis, Mastrofski, and Weisburd 2004a, 2004b), we will argue in this
chapter that there is a wide gap between the promise of Compstat and
its implementation in American policing. While Compstat promises to
reinvigorate police organizations and to empower them to solve crime
problems in America’s cities, it appears to be more focused on main-
taining and reinforcing the “bureaucratic” or “paramilitary” model of
police organization (see Goldstein 1977; Bittner 1980; Punch 1983). In
turn, we argue that there is little substantive evidence of the much touted
crime control benefits of Compstat, and indeed that the Compstat model

284
 Compstat and US policing 285

in many ways inhibits the problem-solving capabilities of police agencies

that it is in theory meant to foster. This leads us to question whether
Compstat is more of a reaction than a reform in policing – a program
that enables police agencies to claim that much has changed, but in fact
allows the police to return to traditional models of police organization.
We begin our chapter by describing the promise of the Compstat model
as an innovation in American policing. We then illustrate how specific
elements of Compstat that reinforce traditional models of policing have
been dominant in its diffusion onto the American police scene, and argue
that the Compstat model includes a fundamental tension that undermines
Compstat’s promise of innovation in police strategies and tactics. Before
concluding, we speculate on the evidence regarding Compstat’s crime
control effectiveness.

The promise of Compstat

In a ground-breaking article written in 1979, Herman Goldstein
described an underlying pathology in police agencies that he viewed as a
fundamental cause of the malaise that surrounded policing at the time.
Police agencies, Goldstein argued, were more concerned with the means
of policing than its ends. The police as an institution had in Goldstein’s
view lost its way, and was in need of fundamental change. He began
his article with a quote from an English newspaper suggesting just how
serious he thought the problem in policing had become.

Complaints from passengers wishing to use the Bagnall to Greenfields bus service
that “the drivers were speeding past queues of up to 30 people with a smile and a
wave of a hand” have been met by a statement pointing out that “it is impossible for
the drivers to keep their timetable if they have to stop for passengers.” (Goldstein
1979: 236)

Just as bus drivers in this English town had in the pursuit of meeting
schedules forgotten that the purpose of having a bus route was to pick
up passengers, so too Goldstein argued, American police in their efforts
to efficiently manage police organization had forgotten that the primary
task of policing was to solve community problems.
Though Goldstein’s article is often cited for its description of how
police strategies could become more “problem oriented,” he called more
generally for broad organizational and cultural change in policing that
would lead to police organizations that could be focused on solving prob-
lems. His article was the first step in a wider movement toward problem-
oriented policing practices in the United States. However, Goldstein gave
little guidance in his article on the fundamental question that he had
286 D. Weisburd, S. D. Mastrofski, J. J. Willis, R. Greenspan

raised about the reorganization of policing, and this question received

little attention in the vast problem-oriented policing literature that was to
develop in subsequent years. For scholars, including Goldstein, problem-
oriented policing was to become primarily a tactical innovation concerned
with how police solve problems and not how the police organization
should be managed.
Compstat is a systematic attempt to answer Goldstein’s original ques-
tion about how police organization could be redefined to focus on prob-
lem solving, and in this sense represents an important attempt to reform
American police agencies. Compstat is a policing model that seeks to
focus police organization on specific problems and to empower police
organizations to identify and solve those problems. For this reason we
have defined Compstat elsewhere as a type of “strategic problem solving”
(Weisburd et al. 2003). Compstat, as opposed to problem-oriented polic-
ing as it has been traditionally discussed and implemented, takes a “big
picture approach,” focusing not so much on the specific strategies that
police use as the ways in which police agencies can be organized as
problem-solving institutions.
But how does Compstat achieve this goal? The impetus behind Comp-
stat’s development in New York was Commissioner Bratton’s intention
to make America’s largest police agency, legendary for its resistance to
change, responsive to his leadership – a leadership that had clearly staked
out crime reduction and improving the quality of life in the neighborhoods
of New York City as its top priorities (Bratton 1999). Strictly speak-
ing, Compstat in New York City referred to a “strategic control system”
developed to gather and disseminate information on crime problems and
track efforts to deal with them. It has, however, become shorthand for
a full range of strategic problem-solving approaches. These elements are
most visibly displayed in Compstat meetings during which precinct com-
manders appear before the department’s top brass to report on crime
problems in their precincts and what they are doing about them.
Drawing from what those who developed Compstat have written (see
Bratton 1996; 1998; 1999; Maple 1999) as well as what those who have
studied Compstat have observed (see Kelling and Coles 1996; Silverman
1999; McDonald, Greenberg, and Bratton 2001), we identify six key
elements that have emerged as central to the development of Comp-
stat as a form of strategic problem solving: mission clarification; inter-
nal accountability; geographic organization of command; organizational
flexibility; data-driven problem identification and assessment; and inno-
vative problem solving. Together they form a comprehensive approach
for mobilizing police agencies to identify, analyze, and solve public safety
problems.
 Compstat and US policing 287

Mission clarification
Compstat assumes that police agencies, like military organizations, must
have a clearly defined organizational mission in order to function effec-
tively. Top management is responsible for clarifying and exalting the core
features of the department’s mission that serve as the overarching reasons
for the organization’s existence. Mission clarification includes a demon-
stration of management’s commitment to specific goals for which the
organization and its leaders can be held accountable – such as reducing
crime by 10 percent in a year (Bratton 1998).

Internal accountability
Internal accountability must be established so that people in the organi-
zation are held directly responsible for carrying out organizational goals.
Compstat meetings in which operational commanders are held account-
able for knowing their commands, being well acquainted with the prob-
lems in the command, and accomplishing measurable results in reducing
those problems – or at least demonstrating a diligent effort to learn from
the experience – form the most visible component of this accountabil-
ity system. However, while such meetings are the visual embodiment of
Compstat, they are part of a more general approach in which police man-
agers are held accountable and can expect consequences if they are not
knowledgeable about or have not responded to problems that fit within
the mission of the department. As Jack Maple, one of the program’s
founders in New York, remarked: “Nobody ever got in trouble because
crime numbers on their watch went up. I designed the process knowing
that an organization as large as the NYPD never gets to Nirvana. Trou-
ble arose only if the commanders didn’t know why the numbers were
up or didn’t have a plan to address the problems” (Maple 1999: 33).
Internal accountability in Compstat establishes middle managers as the
central actors in carrying out the organizational mission, and holds them
accountable for the actions of their subordinates.

Geographic organization of operational command

While Compstat holds police managers to a high level of accountability,
it also gives commanders the authority to carry out the agency’s mis-
sion. Organizational power is shifted to the commanders of geographic
units. Operational command is focused on the policing of territories,
so central decisionmaking authority in police operations is delegated to
commanders with territorial responsibility (e.g., precincts). Function-
ally differentiated units and specialists (e.g., patrol, community police
288 D. Weisburd, S. D. Mastrofski, J. J. Willis, R. Greenspan

officers, detectives, narcotics, vice, juvenile, traffic, etc.) are placed under
the command of the precinct commander, or arrangements are made to
facilitate their responsiveness to the commander’s needs. Silverman notes
that, in New York, “Rather than allow headquarters to determine staffing
and deployment on a citywide basis, it was decided that reducing crime,
fear of crime, and disorder would flow from patrol borough and precinct
coordination of selected enforcement efforts” (1999: 85).

Organizational flexibility
Middle managers are not only empowered with the authority to make
decisions in responding to problems, they are also provided with the
resources necessary to be successful in their efforts. Compstat requires
that the organization develop the capacity and the habit of changing estab-
lished routines to mobilize resources when and where they are needed
for strategic application. For example, in New York City, “Commanding
officers (COs) were authorized to allow their anticrime units to perform
decoy operations, a function that had previously been left to the City-
wide Street Crime Unit. Precinct personnel were permitted to execute
felony arrests warrants, and COs could use plainclothes officers for vice
enforcement activities. Patrol cops were encouraged to make drug arrests
and to enforce quality-of-life laws” (Silverman 1999: 85).

Data-driven problem identification and assessment

Compstat requires that data are made available to identify and analyze
problems and to track and assess the department’s response. Data are
expected to be available to all relevant personnel on a timely basis and
in a readily usable format. According to Maple, “We needed to gather
crime numbers for every precinct daily, not once every six months, to spot
problems early. We needed to map the crimes daily too, so we could iden-
tify hot spots, patterns, and trends and analyze their underlying causes”
(1999: 32).

Innovative problem-solving tactics

The five elements of Compstat described above would have little sub-
stance if Compstat did not encourage innovative problem-solving tactics.
Such tactics have been the core concern of problem-oriented policing and
they are central to the Compstat model (Bratton 1998). Middle managers
are expected to select responses because they offer the best prospects of
success, not because they are “what we have always done.” Innovation
 Compstat and US policing 289

and experimentation are encouraged; use of “best available knowledge”

about practices is expected. In this context, police are expected to look
beyond their own department by drawing upon knowledge gained in other
departments and from innovations in theory and research about crime
prevention.

Compstat and the bureaucratic paramilitary

model of control
Compstat promises a model of policing that is focused on solving prob-
lems. Indeed, in the model of Compstat we have just described policing is
not only focused on specific problems, it is organized around the problem-
solving process. But what does Compstat actually deliver in police agen-
cies that have implemented Compstat programs?
In a national survey we conducted at the Police Foundation we tried to
assess to what extent agencies that claimed to have implemented Comp-
stat or “Compstat-like” programs carried out the elements of Compstat
that we have described above (Weisburd et al. 2001). Surveying all US
police agencies with over 100 sworn police officers and with municipal
policing responsibilities, we examined the extent of implementation of
these core elements of Compstat in agencies that claimed to implement a
Compstat program, and compared this to that found in agencies that did
not claim to implement a Compstat program. The findings of the Police
Foundation study are instructive in understanding how police agencies
have used the Compstat model.
When the department goals of those who had recently developed a
Compstat program were compared with the goals of those who were not
intending to develop a Compstat program, the most significant differ-
ences were in the areas of crime control, increased control of managers
over field operations, improving rank-and-file policing skills, and improv-
ing police morale. Importantly, departments that had claimed to have
recently adopted Compstat were more concerned with reducing crime
and increasing internal accountability than departments that reported
neither having nor planning to develop a Compstat-like program. At the
same time, agencies that had adopted Compstat programs were much
less likely to focus on improving the skills and morale of street-level
officers. This suggests that Compstat represents a departure from the
priorities of “bubble-up” community and problem-oriented policing pro-
grams that had been predominant in police innovation until Compstat’s
arrival on the scene and had focused attention on the empowerment and
training of street-level police officers. Indeed, Compstat appears in this
sense to be modeled more closely on the traditional “bureaucratic” or
290 D. Weisburd, S. D. Mastrofski, J. J. Willis, R. Greenspan

“paramilitary” form of police organization (e.g., see Goldstein 1977;

Melnicoe and Menig 1978; Bittner 1980; Davis 1981; Punch 1983;
Weisburd, McElroy, and Hardyman 1988).
Police departments have traditionally relied on a highly articulated set
of rules defining what officers should and should not do in various sit-
uations to ensure internal control. This supervisory system is strongly
hierarchical and essentially negative, relying primarily on sanctions for
non-compliance with police rules and regulations. Importantly, this
bureaucratic, military model of organization has increasingly come under
attack (Greene and Mastrofski 1988; Goldstein 1990; Bayley 1994;
Mastrofski 1998). Innovations such as community-oriented policing
and problem-oriented policing (as it was developed outside the Comp-
stat model) included a strong current of dissatisfaction with traditional
bureaucratic, top-down command-and-control management (Weisburd
et al. 1988; Mastrofski 1998). They promoted the professionalization of
the rank-and-file, who – equipped with the necessary training, educa-
tion, and motivation to solve problems – are supposed to use their best
judgment to make important decisions about how to solve problems and
to serve the neighborhoods to which they are assigned. Some scholars
have called this a movement toward “decentralization of command” or
“debureaucratization” (Skolnick and Bayley 1987; Mastrofski 1998).
The Police Foundation study suggests that Compstat works to pre-
serve and reinforce the traditional hierarchical structure of the military
model of policing. While fairly strong on elements of mission clarification
and internal accountability, Compstat agencies were found to be largely
indistinguishable from non-Compstat agencies on measures that gauged
geographic organization of command, organizational flexibility, the timely
availability of data, and the selection and implementation of innovative
strategies and tactics. More generally, the Police Foundation data present
a picture of Compstat departments that have embraced control of middle
managers, tending to rely more heavily on punitive than positive conse-
quences. At the same time, Compstat departments are more reluctant
to relinquish power that would decentralize some key elements of deci-
sionmaking geographically (e.g., letting precinct commanders determine
beat boundaries and staffing levels), enhance flexibility, and risk going
outside of the standard tool kit of police tactics and strategies.
But does Compstat empower police organization more generally to
solve problems through its emphasis on the accountability of police
managers? In theory, the original developers of Compstat did not dispute
the community policing view that giving “cops more individual power
to make decisions is a good idea” (Bratton 1998: 198). However, they
believed that in the real world of police organization street-level police
 Compstat and US policing 291

officers “were never going to be empowered to follow through” (Bratton

1998: 199). Moreover, Compstat offered a more efficient approach for
large organizations to encourage problem solving. Middle managers were
expected to spearhead the Compstat approach, bringing the Compstat
message to the line officers who served under them.
Our intensive field observations in three police agencies that were seen
as model Compstat programs, suggest, however, that the rank and file
remain largely oblivious to Compstat and that it intrudes little, if at all,
into their daily work (Willis et al. 2004b). As one patrol officer put it
when asked about Compstat, “If you don’t go [to Compstat meetings],
you don’t know.” In that department, in contrast to almost the entire
command staff, only two or three patrol officers are present at any given
Compstat meeting. They may answer a question or two, and they may give
a brief presentation, but they play a peripheral role. A high-ranking officer
we interviewed remarked that “patrol officers can hide in the meeting and
get away without saying anything.”
Whereas members of the command staff in Compstat departments we
observed, in particular the sector captains, are expected to respond to the
chief’s questions, line officers are rarely called upon to explain a partic-
ular decision. It is true that a sector captain who has been “roasted” in
Compstat for an inadequate strategy may return to his sector and rebuke
his line officers, but the force of the message is considerably weakened for
three reasons. First, Compstat ultimately holds middle managers, not line
officers, accountable. Second, the message that is received by street-level
police officers (as opposed to middle managers) is not being delivered
by the highest-ranking official in the police department. And third, fail-
ure to innovate at the street level does not result in public censure on the
same scale as that experienced by middle managers in Compstat meetings
(Willis et al. 2004).
These field observations suggest that in Compstat agencies the
problem-solving processes are principally the work of precinct comman-
ders, their administrative assistants, and crime analysis staff (sometimes
available at the precinct level, as well as at headquarters). The pressures
on these people can be quite profound. One precinct commander noted
that Compstat was “very stressful.” Another told us: “We’re under con-
stant pressure. It’s the toughest job in this department. We’re held a little
closer to the fire . . . I’ll go home at night after ten hours at work and keep
working – 50–60 hours per week on average.” His lieutenant reinforced
this. “A precinct commander has no life [outside Compstat].” While this
level of intense accountability was commonly expressed by middle man-
agers in Compstat departments, nothing remotely resembling this was
found at lower levels in the organization, except on the rare occasion
292 D. Weisburd, S. D. Mastrofski, J. J. Willis, R. Greenspan

when a line officer was required to make a substantial presentation at a

department Compstat meeting.
In sum, specific components of Compstat that reinforce traditional
hierarchical structures of police organization have taken a predominant
role in Compstat programs nationally. The predominance of these com-
ponents, moreover, does not appear to empower police organization more
generally toward the goal of problem solving.

Internal contradictions in the Compstat model

The fact that police agencies that have adopted the Compstat model have
emphasized elements of command and control over other components
of Compstat does not in itself challenge Compstat as a programmatic
entity. The problem we have identified so far is one of implementation of
Compstat as it has diffused across the United States. The fact that many
departments emphasize specific elements of Compstat does not mean
that the program is not capable of being implemented more fully. But
our research at the Police Foundation suggests that the Compstat model
suffers from a more basic contradiction that hampers the ability of the
model to achieve its goal of solving crime problems.
We conducted in-depth field observations in three “model” Comp-
stat programs, chosen because they scored high on elements of Comp-
stat implementation in our survey and followed closely the New York
Police Department Compstat model (Willis et al. 2004; 2004a; 2004b).
Again, as in our survey, we found in these “model” departments that
Compstat’s implementation placed the greatest stress on mission clar-
ification and internal accountability. But perhaps more notably, when
we assessed the strength of problem-solving efforts in these agencies we
found that Compstat had done very little to change existing crime-fighting
strategies.
We did witness some innovation, such as the successful use of a com-
prehensive and coordinated problem-oriented approach to shutting down
a dilapidated, crime-ridden rooming house in one jurisdiction, but this
was the exception not the norm. The vast majority of problem-solving
approaches identified in these model Compstat agencies relied on tradi-
tional police strategies that had been used before – in particular, asking
patrol officers to identify suspects and keep an eye on things, area satura-
tion, stepping up traffic enforcement, “knock-and-talks,” and increasing
arrests. For example, in response to two unrelated incidents in a particu-
lar location (an increase in prostitution and the constant use of a specific
pay-phone by suspected drug dealers), one district commander put “extra
cruisers” in the area. During another Compstat meeting showing that the
 Compstat and US policing 293

street-side windows of several parked cars had been smashed, the chief
asked, “What kinds of things have we done in the past?” His deputy sug-
gested that in the past they clamped down on motor vehicle violations:
“You know, chief, sometimes you just get lucky. You catch a kid and they
just talk. We need to get people in to talk to them.” Similarly, when we
asked a district commander what he had done regarding a spate of vio-
lent crimes within an area of the city, he told us, “good old police work”
that included putting extra people in the area, increasing police visibility,
sending in the drug unit, etc.
One area where there was greater innovation was in the use of geo-
graphic models of policing in which police activities were concentrated
on specific locales identified by the Compstat process. While concentrat-
ing on “hot spots” represents an important innovation in policing that
has been supported by strong research evidence (Sherman and Weisburd
1995; Braga 2001; Weisburd and Braga 2003), there was little innovation
once hot spots were identified. In this sense, there was often innovation
in the focusing of police resources, but relatively little innovation in what
the police did once they had identified problems.
There did appear to be some very small, but observable differences
in each organization’s capacity to facilitate innovative problem-solving.
For example, one chief strongly encouraged his command staff to share
ideas on crime strategies, and in another jurisdiction task forces provided
a similar forum. However, we found that innovative thinking was gen-
erally not encouraged at Compstat meetings, in good part because of
time constraints. Questions and discussion were generally put off until
the end of the meeting. In fact, even when one of the top managers in one
agency advocated the creation of a poster delineating the most effective
problem-solving strategies at the department’s annual command meeting
to assess Compstat, his suggestion was opposed; the district commanders
responded that they already knew the answers to the problems in their
districts.
Our ethnographic field observations suggest that the failure to achieve
more in-depth problem solving in Compstat departments develops in
part from a fundamental tension in Compstat’s implementation. The
pressure of internal accountability strengthened the existing command
hierarchy, but appeared to work against two of Compstat’s other key
elements – innovative problem solving and geographic organization of
operational command. For example, officers were reluctant to brainstorm
problem-solving approaches during Compstat meetings for fear of under-
mining the authority of more senior officers. Thus the reinforcement of
the hierarchy in Compstat has the unintended consequence of stifling the
kind of creative and open discussion that is essential to creative problem
294 D. Weisburd, S. D. Mastrofski, J. J. Willis, R. Greenspan

solving. Since it is the rank hierarchy within the bureaucratic organization

that legitimates superior power by providing the foundation for authority
relations, any actions that would jeopardize the legitimacy of command
are likely to be avoided. Consequently, lower or same ranking officers are
reluctant to question their superiors during Compstat meetings.
In addition, the danger of “looking bad” in front of superior officers dis-
couraged district commanders from pursuing more creative crime strate-
gies with a higher risk of failure. Similarly, the decisionmaking authority
of middle managers was limited by top management’s willingness to ques-
tion their judgment and to intervene in deployment decisions. And in this
regard, actions that reinforce the legitimacy of the command hierarchy,
such as top leadership’s involvement in the decisions of their subordi-
nates, are difficult to change. Thus, paradoxically, the strengthening of
the command hierarchy under Compstat not only undermines innova-
tive problem solving, but it also interferes with the decentralization of
decisionmaking authority to middle managers.
There may be no silver bullet for resolving the tension between tra-
ditional bureaucratic features and Compstat’s core elements; that is, it
is zero sum. The arrangement of certain bureaucratic structures and
routines, such as hierarchical authority, is necessary for the successful
operation of a large administrative apparatus. It is the overdevelopment
or underdevelopment of these features that hinders the functioning of
the organization (Merton, Gray, Hockey, and Selvin 1952). For exam-
ple, too much control from top leadership stifles innovation, but too
little exposes the organization to excessive risk due to the reckless actions
of its employees (Simons 1995). With this in mind, the challenge for
any department that chooses to implement Compstat is picking the
compromise that most suits its needs and those of its constituencies.
Since there will always be friction between those Compstat elements
that conflict with existing bureaucratic features, implementing Comp-
stat resembles driving down the road with one foot on the accelerator
and the other on the brake: there is going to be a lot of friction at the
wheel.
The proponents of Compstat have seen the reinforcement of traditional
hierarchical models of policing as a key component of Compstat’s efforts
to focus police organization on problem solving. Our observations suggest
that the reality of Compstat in the nation as well as in model Compstat
programs departs markedly from this promise. Indeed, the emphasis on
the hierarchy has led to internal inconsistencies in the Compstat model,
which work against successful innovation. Compstat’s reinforcement of
the bureaucratic hierarchy of policing stifles rather than enhances creative
problem-solving approaches.
 Compstat and US policing 295

Does Compstat enhance crime prevention?

Whatever the internal contradictions of Compstat, from the outset it has
been seen by its originators and proponents as an effective approach to
reducing crime in America’s cities (e.g., see Bratton 1998; Henry 2002).
As early as 1995 an impressive set of statistics had been marshaled to
tout the program’s effectiveness in New York. Homicide rates were down
almost 31 percent from the same period the year before, and crime gener-
ally was down 18.4 percent (Bratton 1998: 289). Henry (2002: 1) reports
that by the year 2000, major crimes had declined 57.26 percent from their
level in 1993, and murder had declined 65.18 percent. Other cities also
attributed crime declines to Compstat (Anderson 2001), though to date
there is not a single systematic review of the impacts of the implementa-
tion of Compstat programs.
One problem in evaluating Compstat as a crime prevention approach is
that it includes many individual elements that may in themselves be suc-
cessful in reducing crime. For example, Compstat has become associated
with a focused geographic approach to crime problems. Crime mapping
has become an integral part of many Compstat programs and, accord-
ing to our Police Foundation survey, Compstat departments are much
more likely to utilize crime mapping and try to identify crime hot spots
than are agencies that do not have Compstat programs (Weisburd et al.
2001). This highly focused geographic approach, often termed hot spots
policing (Sherman and Weisburd 1995; Weisburd and Braga 2003), has
been found to have significant effects on crime and disorder at the places
that are targeted by the police (Braga 2001; Weisburd and Eck 2004).
But clearly police agencies can implement hot spots policing both in the
context of Compstat and without a Compstat approach.
Similarly, Compstat in New York has been strongly linked to what has
been called disorder or broken windows policing (Wilson and Kelling
1982; Kelling and Coles 1996; Bratton 1998; Kelling and Sousa 2001).
Disorder policing assumes that if the police focus aggressively on less seri-
ous crimes that affect the look and feel of neighborhoods they will in the
long run impact more serious offenses. This is because a developmental
sequence is assumed that begins with disorder on the street and ends
with serious criminal predators taking control of the urban environment
(Wilson and Kelling 1982). The effectiveness of disorder policing is much
debated and the research evidence supporting this approach is mixed
(Kelling and Sousa 2001; Taylor 2001; Weisburd and Eck 2004). But
again, Compstat has been implemented in some police agencies with-
out an emphasis on disorder policing, and disorder policing has been
implemented in police agencies without a Compstat program.
296 D. Weisburd, S. D. Mastrofski, J. J. Willis, R. Greenspan
Homicides/100,000 New Yorkers
35
30
25
20
15
10
5
0
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Figure 15.1 New York City homicide rate per 100,000 population,
1986–1998
Note: Vertical line shows year that Compstat was implemented.
Source: Eck and Maguire 2000

What this means is that any evaluation of the effectiveness of Compstat

will be confounded with crime control strategies that are not necessar-
ily part of Compstat. And, unfortunately, the police agencies that have
so strongly advocated Compstat as an approach have had little inter-
est in disentangling the complex network of effects that surround this
program’s impact. This is perhaps not surprising, when so much “good
news” emerged from the simple reporting of official crime rates.
But is there reason to be more sober about the impact of Compstat
even when relying upon such simple descriptive analyses? John Eck and
Edward Maguire (2000) suggest this in their analysis of the crime drop
in New York. One of the most important pieces of evidence brought
by Compstat advocates is the extremely large decline in homicides in
New York after Compstat was initiated. But Eck and Maguire show
that the trend toward lower homicide rates after 1994, when Comp-
stat was established in New York, follows a similar pattern to that found
between 1990 and 1994 (see Figure 15.1). While it is still noteworthy
that the homicide rate declined in New York for such a long period
of time, the decline began much before the Compstat program was
initiated.
Eck and Maguire also examine whether the decline in homicides in
New York could have been accelerated by Compstat, even if it did not
begin with the program. They do note that homicide rates fell more
quickly in New York after than before 1994, when Compstat began.
 Compstat and US policing 297

18000
16000 Newark
Index crimes/100,000

14000
12000
10000
Minneapolis
8000
6000
4000
2000 Lowell
0
1995 1996 1997 1998 1999 2000 2001 2002

Figure 15.2 Crime rates before and after Compstat implementation in

three model Compstat departments
Note: Vertical lines show year that Compstat was implemented.

However, they also show that states surrounding New York also evidenced
more rapid declines in homicide in the latter period. Moreover, in com-
paring the homicide decline in New York to a series of large cities that did
not implement Compstat during the early to mid-1990s, they find that
the New York trend is “almost indistinguishable” from that in a number
of other cities that they examined.
While our Police Foundation studies did not focus on the crime out-
comes associated with Compstat’s implementation, we did examine crime
trends before and after the implementation of Compstat in the three
model Compstat program cities in which we conducted intensive field
observations. As can be seen in Figure 15.2, our observations follow
closely those of Eck and Maguire in New York. Despite the fact that in
each of these cities Compstat was given credit in one form or another
for the declining crime trends observed, we can see an overall decline in
index crimes even before Compstat began as a program.
While these analyses do not disprove Compstat’s alleged impact upon
crime rates in New York or other cities, they suggest the complexity of
making a claim of Compstat’s effectiveness without a systematic anal-
ysis of Compstat’s crime control effects. Compstat is yet to be proven
as an effective crime control strategy in New York or in other cities
that have adopted the Compstat approach. It is a sobering thought that
Compstat has spread so widely across American police agencies absent
strong empirical evidence that it actually does something about the crime
problem.
298 D. Weisburd, S. D. Mastrofski, J. J. Willis, R. Greenspan

Conclusion
In a classic Italian novel, The Leopard (Di Lampedusa 1991), about the
reunification of Italy in the nineteenth century, we are told about a meet-
ing between the Prince of Salina and his nephew in which the nephew has
just told the Prince of his decision to join the republican forces against the
King of Salina. The Prince responds: “You’re crazy, my son. To go and
put yourself with those people . . . a Falconeri must be with us, for the
King.” The nephew, Tancredi Falconeri, responds: “For the King, cer-
tainly, but which King? If we’re not there with them, that bunch is going
to make a republic on us. If we want everything to remain the same, then
everything is going to have to change” (Di Lampedusa 1991: 40, emphasis
added).
In some sense Compstat in American policing appears to follow
Tancredi Falconeri’s strategy for retaining the existing order in Italy in the
nineteenth century. Compstat promises to enhance police effectiveness by
capitalizing on recent police innovations, especially those that have been
associated with problem-oriented approaches to policing. Indeed, Comp-
stat, in theory, provides the first real attempt to implement problem-
oriented policing at the organizational level, an idea proposed by Herman
Goldstein more than a quarter of a century ago, but seemingly forgotten
by police scholars including Goldstein, who focused more on the practice
of problem-oriented policing than the nature of police organization.
We have argued that Compstat as it has actually been implemented by
American police agencies has been focused more on reinforcing and legit-
imating the traditional bureaucratic military model of police organization
than on innovation in the practices of policing. In contrast to reforms such
as community policing which emphasize the importance of challenging
the traditional policing models of command and control, Compstat works
to reinforce them. In this sense, Compstat can be seen more as a reaction
than as a reform in American police organization. It is a return to what
is comfortable for American police agencies, and provides a justification
for traditional models of police organization that have been under attack
by scholars for the last three decades. In this sense, Compstat is indeed a
method of changing everything so that everything can remain the same.
It is a reform in American policing that allows police agencies to return
to what is familiar and comfortable.
If in doing so Compstat can reinvigorate conventional police organiza-
tion for solving problems, then such a reform can of course still be seen
as innovative in the most positive sense. But we have also shown that
Compstat is plagued with a type of internal inconsistency which leads
to what might be termed bureaucratic dysfunction. By strengthening the
 Compstat and US policing 299

bureaucratic hierarchical model of policing, Compstat agencies impede

the ability of police agencies to develop innovative problem-solving
approaches. In this sense, the emphasis on command and control found
in Compstat departments has the ironic consequence of making it diffi-
cult for these agencies to achieve what Compstat is designed to do in the
first place.
In the end, the success of Compstat as an innovation will of course be
judged on whether its claims of crime control effectiveness are confirmed.
The problems we raise about Compstat do not mean that Compstat pro-
grams that are thoughtfully and fully implemented will not be successful
in preventing crime. However, as we have illustrated, the data to date
do not provide a solid basis for the claims of Compstat’s proponents.
This, combined with the trends in the implementation in Compstat we
have described, should dampen the present enthusiasm for Compstat
models.


Anderson, D. C. (2001). Crime control by the numbers: Compstat yields new
lessons for the police and the replication of a good idea. Ford Foundation
Report: 5. New York.
Bayley, D. (1994). Police for the future. New York: Oxford University Press.
Bittner, E. (1980). The functions of the police in modern society. Cambridge, MA:
Oelgeschlager, Gunn and Hain.
Braga, A. (2001). The effects of hot spots policing on crime. The Annals of Amer-
ican Political and Social Science, 578, 104–125.
Bratton, W. J. (1996). Cutting crime and restoring order: What America can
learn from New York’s finest. Heritage Foundation Lectures and Educational
Programs, Heritage Lecture #573. Retrieved from: www.nationalsecurity.org/
heritage/library/categories/crimelaw/lect573.html.
(1998). Turnaround: How America’s top cop reversed the crime epidemic. New York:
Random House.
(1999). Great expectations: How higher expectations for police departments
can lead to a decrease in crime. In R. H. Langworthy (ed.), Measuring what
matters: Proceedings from the Policing Research Institute meetings (pp. 11–26).
Washington, DC: National Institute of Justice.
Davis, E. M. (1981). Professional police principles. In H. W. More, Jr. (ed.),
Critical issues in policing. Cincinnati, OH: Anderson Publishing Company.
Di Lampedusa, G. (1991). The leopard. (Translated by Archibald Colquhoun).
New York: Pantheon Books [1960].
Eck, J. E. and Maguire, E. R. (2000). Have changes in policing reduced violent
crime? An assessment of the evidence. In A. Blumstein and J. Wallman (eds.),
The crime drop in America (pp. 207–65). Cambridge: Cambridge University
Press.
Goldstein, H. (1977). Toward a community-oriented policing: Potential, basic
requirements, and threshold questions. Crime and Delinquency, 33 (1), 6–30.
300 D. Weisburd, S. D. Mastrofski, J. J. Willis, R. Greenspan

(1979). Improving policing: A problem oriented approach. Crime and Delin-

quency, 24, 236–258.
(1990). Problem-oriented policing. New York: McGraw-Hill.
Greene, J. R. and Mastrofski, S. D. (1988). Community policing: Rhetoric or reality.
New York: Praeger.
Greenspan, R., Mastrofski, S. D., and Weisburd, D. (2003). Compstat and orga-
nizational change: Short site visit report. Washington, DC: Police Foundation.
Gurwitt, R. (1998). The comeback of the cops. Governing, January, 14–19.
Henry, V. E. (2002). The COMPSTAT paradigm: Management accountability in
policing, business, and the public sector. New York: Looseleaf Publications.
Kelling, G. L. and Coles, C. M. (1996). Fixing broken windows: Restoring order
and reducing crime in our communities. New York: Free Press.
Kelling, G. L. and Sousa, W. H., Jr. (2001). Do police matter? An anal-
ysis of the impact of New York City’s police reforms. Retrieved from:
http://www.manhattan-institute.org/cr 22.pdf
Law Enforcement News. (1997). NYC’s compstat continues to win admirers.
October 13.
Maas, P. (1998). What we’re learning from New York City. Parade, May 10, 4–6.
Maple, J. (1999). The crime fighter: Putting the bad guys out of business. New York:
Doubleday.
Mastrofski, S. D. (1998). Community policing and police organization structure.
In J. P., Brodeur (ed.), How to recognize good policing: Problems and issues
(pp. 161–189). Thousand Oaks, CA: Sage Publications.
McDonald, P. P. (1998). The New York City crime control model: A guide to
implementation. Unpublished manuscript. Washington, DC.
McDonald, P. P., Greenberg, S., and Bratton, W. J. (2001). Managing police
operations: Implementing the NYPD crime control model using COMPSTAT.
Belmont, CA: Wadsworth Publishing Co.
Melnicoe, W. B. and Menig, J. (1978). Elements of police supervision. Encino, CA:
Glencoe Publishing Company.
Merton, R. K., Gray, A. P., Hockey, B., and Selvin, H. G. (eds.). (1952). Reader
in bureaucracy. New York: Free Press.
Punch, M. (1983). Management, supervision and control. In M. Punch (ed.),
Control in the police organization. Cambridge, MA: MIT Press.
Remnick, D. (1997). The crime buster. The New Yorker, February 24 & March
3, 94–109.
Sherman, L. and Weisburd, D. (1995). General deterrent effects of police patrol
in crime ‘hot spots’: A randomized controlled trial. Justice Quarterly, 12 (4),
625–648.
Silverman, E. B. (1996). Mapping change: How the New York City police
department re-engineered itself to drive down crime. Law Enforcement News,
December.
(1999). NYPD battles crime: Innovative strategies in policing. Boston, MA: North-
eastern University Press.
Simons, R. (1995). Control in an age of empowerment. Harvard Business Review,
73, 1–7.
Skolnick, J. H. and Bayley, D. H. (1987). Theme and variation in community
policing. Crime and Justice. Washington, DC: National Institute of Justice.
 Compstat and US policing 301

Taylor, R. B. (2001). Breaking away from broken windows: Baltimore neighborhoods

and the nationwide fight against crime, grime, fear, and decline. Boulder, CO:
Westview Press.
Weisburd, D. and Braga, A. (2003). Hot-spots policing. In H. Kury and
J. Obergfell-Fuchs (eds.), Crime prevention: New approaches (pp. 337–355).
Mainz, Germany: Weisser-Ring.
Weisburd, D. and Eck, J. (2004). What can police do to reduce crime, disorder,
and fear. Annals of American Political and Social Science, 593, 42–65.
Weisburd, D., McElroy, J., and Hardyman, P. (1988). Challenges to supervision
in community policing: observations on a pilot project. American Journal of
Police 7 (2), 29–50.
Weisburd, D., Mastrofski, S. D., McNally, A. M., and Greenspan, R. (2001).
Compstat and organizational change: Findings from a national survey. Report
submitted to the National Institute of Justice by the Police Foundation.
Willis, J. J., Mastrofski, S. D., Weisburd, D., and Greenspan, R. (2004). Comp-
stat and organizational change in the Lowell police department: Challenges
and opportunities. Washington, DC: Police Foundation. Also available from:
http://www.policefoundation.org/pdf/compstat.pdf
Weisburd, D., Mastrofski, S. D., McNally, A. M., Greenspan, R., and Willis, J. J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2 (3), 421–456.
Willis, J. J., Mastrofski, S. D., and Weisburd, D. (2004a). COMPSTAT and
bureaucracy: A case study of challenges and opportunities for change. Justice
Quarterly, 21 (3), 463–496.
(2004b). Compstat in practice: An in-depth analysis of three cities. Washington
DC: The Police Foundation.
Wilson, J. Q. and Kelling, G. L. (1982). Broken windows: The police and neigh-
borhood safety. The Atlantic Monthly, March, 29–38.
Part VIII

Evidence-based policing
16 Advocate
Evidence-based policing for crime prevention

Brandon C. Welsh

Introduction
In recent weeks Big City has experienced an increase in young people sell-
ing drugs on the street. The police chief directs one of his managers to
organize a small group of officers to look into the matter and report back
with a plan of action. One of these officers is tasked with researching
what works to address the problem of street-level drug sales. The offi-
cer carries out literature searches, contacts other police departments and
police research organizations such as the Police Foundation, and follows
up with some of the researchers involved in evaluation studies. The most
rigorous evaluations are coded according to the outcomes of interest,
results are analyzed, and a report is prepared that shows what type of
intervention works best. Detailed observational and other information
on the problem gathered by the other officers is used to assess the appli-
cability of the “best practice” to the local context and conditions. On the
basis of this information, the chief authorizes the needed resources for a
program to be implemented to address street-level drug sales, and intro-
duces a monitoring and evaluation scheme to aid in-house policy and for
dissemination purposes.
This fictitious scenario, while brief on details, is an example of
evidence-based policing in action. Evidence-based policing involves the
police using the highest quality available research evidence on what works
best to reduce a specific crime problem and tailoring the intervention to
the local context and conditions. This approach likely holds wide appeal.
To local government and policymakers the police may be seen as being
efficient in their use of monetary and other resources. To police scholars
the police may be seen as acting like scientists: targeting crime problems
with accumulated scientific evidence of the highest quality on what works
best; here the police are seen as not reinventing the wheel. Evidence-
based policing could be another example of, in Bayley’s (1998: 174)
words, “smarter law enforcement.” To citizens, evidence-based policing
may also hold appeal because of its commitment to the use of what works

305
306 Brandon C. Welsh

best and its clear intent to bring about positive change. But for some
citizens, especially those who reside in high-crime neighborhoods, how
the police interact with residents and suspected offenders may ultimately
influence their feelings toward evidence-based policing. Arguably, these
residents may reserve judgment on all police practices.
This chapter argues that there is strong scientific merit in the police
adopting a formalized evidence-based approach to preventing crime and
that a great deal of public good can come about from this. This chapter
also argues that evidence-based policing holds much promise in mak-
ing a difference – indeed where other police innovations in recent years
have been unsuccessful – in those very neighborhoods where crime and
violence are most intractable and public confidence in the police is
lowest. The latter may come about not just because police are using
scientific evidence about what works best, but also because police are
targeting crime-risk factors and tailoring what works to local context and
conditions.
This chapter is not the first piece on evidence-based policing.
Sherman’s (1998) seminal piece, Evidence-Based Policing, and his earlier
works dating back to his Crime and Justice essay, “Attacking crime: polic-
ing and crime control” (Sherman 1992), spelled out the arguments in
support of the police using accumulated scientific evidence on what works
best to prevent crime, and set the stage for this approach being acknowl-
edged by police scholars (the editors of the present volume included)
as one of the important police innovations of the last two decades.
Importantly though, evidence-based policing is but one part of a larger
evidence-based movement, from which it draws institutional and intellec-
tual support. This may be one feature that distinguishes evidence-based
policing from other police innovations.
The chapter is organized as follows. The first part describes the
evidence-based model with special reference to evidence-based policing,
and overviews the movement that has given rise to its use in the social
sciences. It then reviews the state of evidence-based research on polic-
ing with a specific focus on the prevention of crime. One of the aims of
this part is to address the pressing question: Is there enough evidence
to develop evidence-based policing? This is followed by a discussion of
whether we can expect the police to use or perhaps institutionalize the
evidence-based model to prevent crime. The chapter concludes with a
discussion on why the police should adopt an evidence-based approach
to prevent crime.
The evidence-based model and policing
In characterizing the evidence-based model and policing it is important
to first define what is meant by the term “evidence.” Throughout this
 Evidence-based policing 307

chapter evidence is taken to mean scientific, not criminal evidence (see

Sherman 1998: 2, note 1). Evidence introduced in criminal court pro-
ceedings, while bound by laws and procedures, is altogether different
from scientific evidence. The latter “refers to its common usage in science
to distinguish data from theory, where evidence is defined as ‘facts . . .
in support of a conclusion, statement or belief’” (Shorter Oxford English
Dictionary 2002, as cited in Sherman 2003: 7).
At the heart of the evidence-based model is the notion that “we are
all entitled to our own opinions, but not to our own facts” (Sherman
1998: 4). Many may be of the opinion that hiring more police officers
will yield a reduction in crime rates. However, an examination of the
empirical research evidence on the subject reveals that this is not the
case (see below). Use of opinions instead of facts to guide crime policy
has a greater chance to result in harmful or iatrogenic effects (McCord
2003) or to not work at all, waste scarce public resources (Welsh and
Farrington 2000), and divert policy attention from the real priorities of
the day. Moreover, within the evidence-based paradigm, drawing conclu-
sions based on facts calls attention to a number of fundamental issues:
the validity of the evidence; the methods used to locate, appraise, and
synthesize the evidence; and implementation.
In an evidence-based model, the source of scientific evidence is empir-
ical research in the form of evaluations of programs, practices, and poli-
cies. But not all evaluations are made equal. Some are more scientifically
valid than others. The randomized controlled experiment is the most
convincing method of evaluating crime prevention programs (Farrington
1983; Weisburd, Lum, and Petrosino 2001). This type of evaluation
design involves the random allocation of subjects to treatment (the condi-
tion that receives the intervention) and control conditions. Through the
process of random assignment, treatment and control groups are equated
(prior to intervention) on all possible extraneous variables (e.g., age, gen-
der, social class, school performance) that may be related to the outcome
of interest, which for the present purposes is crime. Hence, any subse-
quent differences between them must be attributable to the intervention.
Technically, randomized experiments have the highest possible internal
validity in unambiguously attributing an effect to a cause (Shadish, Cook,
and Campbell 2002).
Other things being equal, a program evaluation in which experi-
mental and control units are matched or statistically equated prior to
intervention – what is called a non-randomized experiment – has less
internal validity than a randomized experiment. A program evaluation
with no control group has even less internal validity, since it fails to
address many threats to internal validity, such as history (an event occur-
ring at the same time as the intervention), maturation (a continuation
308 Brandon C. Welsh

of preexisting trends, for example, in normal human development), and

instrumentation (a change in the method of measuring the outcome)
effects (Cook and Campbell 1979).
Just as it is crucial to use the highest-quality evaluation designs to inves-
tigate the effects of police practices, it is also important that the most rig-
orous methods be used to locate, appraise, and synthesize the available
research evidence. The main types of review methods include the single
study, narrative, vote-count, systematic, and meta-analytic (Welsh and
Farrington 2001; see below for details on these review methods). Single
study and narrative reviews are less rigorous than the others, and are not
recommended in carrying out evidence-based research. Comprehensive-
ness, adherence to scientific rules and conventions, and transparency are
at the heart of the rigorous review methods.
The question, What does the evidence say?, is in direct reference to
existing evaluation studies. Rather than the self-selection of individual
studies to substantiate one’s position, evidence refers to the accumulated
body of studies (published and unpublished) on the subject of interest.
This degree of comprehensiveness is one way that the evidence-based
model attempts to address researcher or institutional bias. Adherence to
scientific rules and conventions is most germane to the use of appropriate
quantitative techniques in analyzing results, as well as the presentation
of these results. The transparency of search methods, criteria used to
include (and exclude) studies, quantitative techniques, and so on is also
important in allowing the reviews to be replicated by other researchers.
The importance of implementation to the evidence-based model is best
captured by the following: “Evidence-based policing assumes that exper-
iments alone are not enough. Putting research into practice requires just
as much attention to implementation as it does to controlled evaluations”
(Sherman 1998: 7).
Successful implementation calls for taking account of local context
and conditions. Some critics of the evidence-based paradigm (Lab 2003)
claim that it fails to adequately account for local context and conditions in
reaching conclusions about what works. The main thrust of this argument
is that unless local context and conditions are investigated undue weight
may be ascribed to any effects of the intervention on the outcome of inter-
est. Evidence-based policing has in place the capacity to take account of
these features. For example, those tasked with investigating the research
evidence on the effectiveness of police practices to deal with a particular
crime problem can question the original researchers or solicit unpub-
lished reports to learn about how local context and conditions may have
influenced the observed results. This information can then be integrated
into the existing profile of the program.
 Evidence-based policing 309

Evidence-based policing also has the capacity to appropriately tailor

proven strategies or practices to the local setting. While perhaps obvi-
ous and supported in research on diffusion of knowledge and replication
studies (see Ekblom 2002; Liddle, Rowe, Quille et al. 2002), not paying
attention to this (and using the “one-size-fits-all” approach) can severely
impact upon implementation as well as the overall effectiveness of the
intervention. Hough and Tilley (1998: 28) make this point clear:

Routinely used techniques often cannot be taken off the shelf and applied
mechanically with much real prospect of success. Standard, broad-brush, block-
buster approaches to problems tend to produce disappointing results. Where new
approaches are adopted it is likely that adjustments will be needed in the light of
early experience. All crime prevention measures work (or fail to do so) according
to their appropriateness to the particular problem and its setting.

Detailed observational and other information on the crime problem that

is the focus of attention, as well as the setting (e.g., urban density, unem-
ployment rates), can be matched with the proven practice and modifica-
tions can then be made as needed.
Another important feature of the evidence-based model is the outcome
of interest. While it is acknowledged that evidence-based policing can
serve other useful purposes (e.g., improving police training standards,
improving police–community relations), the main outcome of interest or
“bottom line” is crime prevention (see Welsh 2002). The parallel is with
evidence-based medicine’s primary focus on saving lives or improving the
quality of life of those suffering from terminal or chronic illnesses. For
evidence-based policing or evidence-based crime prevention in general,
the prevention of crime is a first tier or primary outcome. This is the focus
throughout this chapter.

Part of a larger movement

Evidence-based policing is a part of a larger and increasingly expanding
evidence-based movement. In general terms, this movement is dedicated
to the betterment of society through the utilization of the highest quality
scientific evidence on what works best. The evidence-based movement
first began in medicine (Millenson 1997) and has, more recently, been
embraced by the social sciences (Mosteller and Boruch 2002; Sherman,
Farrington, Welsh, and MacKenzie 2002; Sherman 2003).
In 1993, the Cochrane Collaboration was established to prepare, main-
tain, and make accessible systematic reviews (see below for details on this
review methodology) of research on the effects of health care and medical
interventions. The Cochrane Collaboration established collaborative
310 Brandon C. Welsh

review groups (CRGs) across the world to oversee the preparation and
maintenance of systematic reviews in specific areas, such as heart disease,
infectious diseases, and breast cancer. All reviews produced by Cochrane
CRGs follow a uniform structure. The same level of detail and consis-
tency of reporting is found in each, and each review is made accessible
through the Cochrane Library, a quarterly electronic publication.
The success of the Cochrane Collaboration in reviewing health care
interventions stimulated international interest in establishing a simi-
lar infrastructure for conducting systematic reviews of research on the
effects of interventions in the social sciences, including education, social
work and social welfare, and crime and justice. In 2000, the Campbell
Collaboration was established. It is named after the influential
experimental psychologist Donald T. Campbell (see Campbell 1969).
The Collaboration’s Crime and Justice Group aims to prepare and main-
tain systematic reviews of criminological interventions and to make them
accessible electronically to practitioners, policymakers, scholars, and the
general public.

The state of evidence-based research on policing for

crime prevention
It is not known to what extent police departments across the country
or even big city police departments for that matter undertake evidence-
based research to aid in the development and implementation of programs
and practices to prevent crime. In the absence of being able to report on
the state of evidence-based policing in real-world settings, this section
brings together leading reviews on the effectiveness of policing for crime
prevention. This serves two main purposes: (a) to take stock of what we
know (and do not know) about what works best in policing for crime
prevention; and (b) to assess if there is sufficient research evidence to
develop evidence-based policing.

Sherman and Eck (2002)

As part of a larger project to assess the scientific evidence on the effec-
tiveness of preventing crime in seven major institutional settings in which
crime prevention efforts take place (families, communities, schools, labor
markets, places, police agencies, and courts and corrections) in the
United States and internationally, Sherman and Eck (2002) undertook
a review of police practices. This piece updates Sherman’s (1997) chap-
ter in the widely cited report prepared for the US Congress, Preventing
crime: What works, what doesn’t, what’s promising (Sherman, Gottfredson,
MacKenzie et al. 1997).
 Evidence-based policing 311

Table 16.1 Sherman and Eck’s (2002) conclusions on the

effectiveness of police practices to prevent crime

What Works
Increased directed patrols in street-corner hot spots of crime
Proactive arrests of serious repeat offenders
Proactive drunk driving arrests
Arrests of employed suspects for domestic assault
Problem-oriented policing
What Does Not Work
Neighborhood block watch
Arrests of some juveniles for minor offenses
Arrests of unemployed suspects for domestic assault
Drug-market arrests
Community policing with no clear crime-risk factor focus
Adding extra police to cities, regardless of assignment or activity
What Is Promising
Police traffic enforcement patrols against illegally carried handguns
Community policing with community participation in priority setting
Community policing focused on improving police legitimacy
Warrants for arrest of suspect absent when police respond to domestic
violence

Source: Adapted from Sherman and Eck (2002: 321–322).

As shown in Table 16.1, the authors, on the basis of the highest-quality

available scientific evidence, identified five police practices that are effec-
tive in reducing crime. Most of these practices are highly specific, such as
increased directed patrols in street-corner hot spots of crime and arrests of
employed suspects for domestic assault. Also important to police man-
agers and other consumers of this evidence-based research is that this
review draws conclusions about what does not work, what is promising,
and what is unknown (the latter is not shown in Table 16.1). Some of the
police practices of unknown effectiveness include police recreation activ-
ities with juveniles (e.g., Police Athletic Leagues) and in-car computer
terminals (Sherman and Eck 2002: 322).
One of the first questions that may be asked of these conclusions is:
How much confidence can one have in them? This is a critical ques-
tion in any evidence-based research, and one that goes to the heart of
the underlying methodology employed (e.g., type of evaluation studies
included, method used to aggregate results). The saying, “garbage in,
garbage out,” needs to be kept in mind in judging the results of any
single study or review. As noted above, the evidence-based model is com-
mitted to the use of the highest-quality available scientific evidence and
the most rigorous methods.
312 Brandon C. Welsh

Sherman and Eck’s (2002) review used what is known as a vote-count

methodology. The vote-count method adds a quantitative element to the
narrative review,1 by considering statistical significance (the probability
of obtaining the observed effect if the null hypothesis of no relationship
were true). In essence, this method tallies-up the “number of studies with
statistically significant findings in favor of the hypothesis and the number
contrary to the hypothesis (null findings)” (Wilson 2001: 73). The main
problem with using statistical significance is that it depends partly on
sample size and partly on effect size. For example, a significant result
may reflect a small effect in a large sample or a large effect in a small
sample (Farrington, Gottfredson, Sherman, and Welsh 2002).
However, Sherman and Eck’s (2002) review adopted a more com-
prehensive vote-count methodology, first developed by Sherman et al.
(1997) and revised by Farrington et al. (2002). In addition to statisti-
cal significance, this vote-count method integrates a “scientific methods
scale” (SMS) that is largely based on the work of Cook and Campbell
(1979), and describes research designs that are most effective in elim-
inating threats to internal validity. The scale ranges from level 1 (cor-
relation between a prevention program and a measure of crime at one
point in time) to level 5 (random assignment of program and control
conditions to units), with level 5 being the highest. Only studies with
a minimum of a level 3 evaluation design (presence of before and after
measures of crime in experimental and comparable control conditions)
were included in the review. This was considered to be the minimum
interpretable design by Cook and Campbell (1979). (For more details
on the SMS, see Farrington et al. 2002; Welsh, Farrington, Sherman,
and MacKenzie 2002.) Importantly, external validity (the generalizability
of internally valid results) was addressed to some extent by establishing
rules for accumulating evidence.2

Eck and McGuire (2000) and Weisburd and Eck (2004)

As part of a larger project to characterize and explain the decline in
crime rates in the United States during the 1990s (Blumstein and Wall-
man 2000), Eck and Maguire (2000) examined whether the contribu-
tions of seven major policing strategies (increase in number of police
officers, community policing, zero tolerance policing, directed patrols in
hot spots, firearms enforcement, retail drug-market enforcement, and
problem-oriented policing) during this period of time reduced violent
crime. Their work is relevant here because in order to do this Eck
and Maguire reviewed the scientific evidence on the effectiveness of
these policing strategies. The authors found that some of the policing
 Evidence-based policing 313

strategies hold sufficient empirical support to reduce violent crime. These

included directed patrols in hot spots, firearms enforcement, retail drug
market enforcement (especially when combined with the threat of civil
action against property owners), and problem-oriented policing. For the
other three strategies the authors found that the evidence was mixed for
increasing the number of police officers, weak for community policing,
and non-existent for zero tolerance policing.
Weisburd and Eck’s (2004) review was carried out as part of a larger
National Academy of Sciences panel investigating police policy and prac-
tices (Committee to Review Research 2004). In reviewing the extant
research evidence about what works in policing, the authors set out to
address a number of critical questions facing American policing, includ-
ing “Does the research evidence support the view that standard models
of policing are ineffective in combating crime and disorder?” and “Do
recent police innovations hold greater promise of increasing community
safety, or does the research evidence suggest that they are popular but
actually ineffective?” (Weisburd and Eck 2004: 43).
As shown in Table 16.2, Weisburd and Eck (2004) found that focused
(or targeted) police strategies are more likely to yield crime reduction
benefits than unfocused (or uniformly applied) strategies. The authors
found moderate to strong evidence of effectiveness for recent innova-
tions of problem-oriented policing, problem solving in hot spots, focused
intensive enforcement, and hot spots patrols. Also revealing was the find-
ing that the standard model of policing (e.g., adding more police, general
patrol), which is both unfocused and relies almost exclusively on law
enforcement sanctions, is not supported by solid empirical study. On the
latter, the authors observe, “While this approach remains in many police
agencies the dominant model for combating crime and disorder, we find
little empirical evidence for the position that generally applied tactics
that are based primarily on the law enforcement powers of the police are
effective” (Weisburd and Eck 2004: 57).
While somewhat different in scope, these two reviews have been
grouped together because each reviewed only the most methodologically
rigorous studies. As in the Sherman and Eck (2002) review, these reviews
sought to separate out the wheat from the chaff, the former serving as the
empirical evidence upon which conclusions are drawn about effectiveness
or ineffectiveness.

Braga (2001)
Under the auspices of the Campbell Collaboration’s Crime and Justice
Group, Braga (2001) carried out a review of the effects of hot spots
314 Brandon C. Welsh

Table 16.2 Weisburd and Eck’s (2004) summary of findings on police

effectiveness research

Police Strategies that . . . Are Unfocused Are Focused

Apply a diverse array of Inconsistent or weak Moderate evidence of

approaches, including evidence of effectiveness: effectiveness:
law enforcement – impersonal community – problem-oriented
sanctions policing (e.g., newsletters) policing
Weak to moderate evidence Strong evidence of
of effectiveness: effectiveness:
– personal contacts in – problem solving in hot
community policing spots
– respectful police-citizen
contacts (improving
police legitimacy, foot
patrols (fear reduction))
Rely almost exclusively Inconsistent or weak Inconsistent or weak
on law enforcement evidence of effectiveness: evidence of effectiveness:
sanctions – adding more police – repeat offender
– general patrol investigations
– rapid response Moderate to strong
– follow-up investigations evidence of effectiveness:
– undifferentiated arrest for – focused intensive
domestic violence enforcement
– hot spots patrols

Source: Adapted from Weisburd and Eck (2004: 57; table 1).

policing on crime and disorder. This form of policing involves the

targeting of police enforcement measures on high-risk crime places. Also
examined were the effects on the displacement of crime (an unintended
increase in crime, for example, in a different location) and the diffusion
of crime control benefits (an unintended decrease in crime in an adjacent
location).
Nine studies were identified that met the inclusion criteria for the
review. The nine studies involved a range of enforcement strategies
(e.g., problem-oriented policing interventions involving both tradi-
tional enforcement tactics and situational crime prevention responses,
uniformed police patrol) targeted at a range of crime problems (e.g., ille-
gal gun carrying, street-level drug sales, assaults and robberies) in high-
risk places. Braga (2001) found evidence that targeted police actions can
prevent crime and disorder in hot spots, displacement of crime is rare, and
some enforcement actions produce unintended crime prevention benefits
in adjacent locations.3
 Evidence-based policing 315

Unlike the above reviews this one is focused on investigating the effec-
tiveness of one particular police practice or strategy. This narrow focus
is purposeful. Among other reasons, it is meant to enable the researcher
to carry out a comprehensive search for relevant studies. Through this
it is hoped that the researcher will uncover all available studies on the
subject. This narrow focus is also meant to allow the researcher to look
in depth at each study on multiple dimensions (and code them) to assess
any potential influence on the observed outcomes.
Braga’s (2001) review is known as a systematic review, which is the
most rigorous method for assessing the effectiveness of criminological
interventions (Chalmers 2003). Systematic reviews use rigorous methods
for locating, appraising, and synthesizing evidence from prior evaluation
studies, and they are reported with the same level of detail that char-
acterizes high-quality reports of original research. Systematic reviews,
according to Johnson and his colleagues (2000: 35), “essentially take an
epidemiological look at the methodology and results sections of a specific
population of studies to reach a research-based consensus on a given
topic.” They have explicit objectives, explicit criteria for including or
excluding studies, extensive searches for eligible evaluation studies from
across the world, careful extraction and coding of key features of studies,
and a structured and detailed report of the methods and conclusions of
the review.

Toward evidence-based policing?

Overall, these reviews, using different high-quality methods, demonstrate
(with a high degree of concordance) that there are a number of policing
practices with robust empirical evidence of effectiveness in preventing
crime. These reviews also show (also with a high degree of concordance)
that there are some promising practices that police currently employ, and
many more that police should consider abandoning or perhaps modifying.
(It is of course recognized that all crime prevention programs, policing
included, that become institutionalized and take on brand-name appeal,
think of DARE (Drug Abuse and Resistance Education) can be terribly
difficult to do away with or even change.)
These reviews of the empirical evidence on policing for crime preven-
tion also suggest that there is sufficient capacity to develop evidence-
based policing. Critics may charge that a handful of effective practices is
hardly sufficient grounds for the development of evidence-based polic-
ing. But underlying this aggregate number of effective policing practices
are a few important points that speak to the capacity for the development
of evidence-based policing. One is that the effective practices represent
316 Brandon C. Welsh

a diverse range of policing practices. This is important because different

communities have different policing needs. A second and related point is
that the effective practices are by no means obscure or relegated to one
region or another. Instead, they are used by police departments across
the country.
Another reason to believe there is sufficient capacity to develop
evidence-based policing is that there are even more promising practices,
or those with moderate evidence of effectiveness, that, with some modest
resources directed to replication studies, could conceivably become effec-
tive practices. This is by no means a long shot. Notwithstanding the con-
cern that “knowledge of many of the core practices of American policing
remains uncertain” (Weisburd and Eck 2004: 59), few if any areas of
research in criminal justice or crime prevention more generally, can boast
the level and high-quality nature of experimental evaluation in policing.
In the words of Sherman and Eck (2002: 321), “It is no small achieve-
ment that police crime prevention research has developed to the point of
having some conclusions to discard.”

Can we expect the police to use the evidence-based

model?
If history is any indication the answer to this question may be mixed. On
the one hand, we can look to the widespread adoption of some police
innovations in the absence of a link to research evidence, such as com-
munity policing. On the other hand, we can find a connection to the
evidence-based paradigm – although perhaps not necessarily conceived
of in this form – in the adoption of problem-oriented policing and other
targeted police practices that go beyond a reliance on police powers.
Sherman (1998) has critiqued community policing as not being linked
to research evidence on what works best in preventing crime. Commu-
nity policing “is much more about how to do police work – a set of out-
puts – than it is about desired results, or outcomes” (Sherman 1998: 5).
As noted above, Sherman and Eck (2002) found community policing
with no clear focus on crime-risk factors to be ineffective in preventing
crime. Critics may charge that community policing was never intended
just to prevent crime. This is true. But returning to a point made earlier,
within an evidence-based model, crime prevention is the chief concern of
policing.
In contrast to community policing, “[p]roblem-oriented policing is
clearly the major source of evidence-based policing” (Sherman 1998:
5–6). Sherman (1998) notes that this comes from problem-oriented
policing emphasizing systematic problem-solving responses. What may
 Evidence-based policing 317

be missing more often though is a commitment to the use of the scientific

evidence on what works best. That is, once the underlying problems of
the crime or disorder incidents have been identified, are the police using
practices with proven effectiveness? In the case of problem-oriented polic-
ing, with its focused or targeted nature and its use of a diverse array of law
enforcement sanctions and alternative crime control measures (Weisburd
and Eck 2004), proven practices need not always come from the policing
community. Indeed, the evaluation research on situational crime preven-
tion or place-based crime prevention (Clarke 1997; Eck 2002) and, to
a lesser extent, community crime prevention (Welsh 2003) offer police
many more efficacious options.
There will always exist barriers to getting the police to use research
evidence on what works best in preventing crime. Some of these have
already been addressed; others include administrative constraints (e.g.,
too few resources, need for training of personnel), philosophical differ-
ences, and institutional resistance to change. Overcoming the disconnect
between research evidence and practice may best be achieved through
the employment of what Sherman (1998) refers to as an “evidence cop.”
This individual, ideally working within a police department (for smaller
police departments it may be necessary to pool resources), would serve
as both a research scientist and manager. In the capacity of a research
scientist the evidence cop would be responsible for keeping up to date on
the latest police research findings and coming up with recommendations
based on the accumulated research evidence. In the capacity of a man-
ager the evidence cop’s role would be to monitor policing practices in the
field to ensure they are adhering to recommendations based on research
evidence. Importantly, his/her role would also be to “redirect practice
through compliance rather than punishment” (Sherman 1998: 3). The
police community could learn from similar initiatives that have shown
promise in the medical profession (Millenson 1997) and in agriculture
(MacKenzie 1998).

Why the police should adopt an evidence-based approach

Evidence-based policing brings scientific evidence to center stage in deci-
sions about which police practices should be used to deal with certain
crime problems. It also brings to policing a greater focus on the prevention
of crime. Improved effectiveness in the prevention of crime, greater effi-
ciency in the use of scarce public resources, and improved relations with
the public may be some of the outcomes of police adopting an evidence-
based approach. Of course, the police have achieved and continue to
achieve various degrees of success in these areas through a number of
318 Brandon C. Welsh

diverse practices. So what are some of the key reasons for why the police
should adopt an evidence-based approach?
Returning to the scenario that opened this chapter, one of the most
important reasons for the police to use an evidence-based approach is
that it brings yet another innovative tool to a field of criminal justice
that is a leader in innovation. More importantly perhaps, evidence-based
policing may serve as a catalyst in focusing limited resources on identifi-
able risks, further developing a knowledge base on what works best, and,
in continuing to develop new approaches to crime problems, encouraging
police departments to keep an eye on the bottom line of crime prevention
effectiveness.
In fact, at a time of police experimentation with new technologies like
Compstat and 3-1-1 non-emergency call systems to aid in the more effi-
cient use of police resources (see Mazerolle, Rogan, Frank, Famega et al.
2002; Weisburd, Mastrofski, McNally, Greenspan et al. 2003) and new
deterrence-based problem-solving strategies to reduce serious crimes (see
Kennedy 1997; Tita, Riley, and Greenwood 2003), the adoption of an
evidence-based approach in policing may not be such a great departure
from the current state of policing. While this is hardly a reason for the
police to adopt an evidence-based approach, it may go some way toward
selling the idea. And this cannot hurt.

   
I wish to thank David Weisburd and Anthony Braga for helpful comments and
suggestions on an earlier draft of this chapter.
1. Narrative reviews of the literature quite often include many studies and may be
very comprehensive. The main drawback of the narrative method is researcher
bias. This bias, whether intentional or not, typically starts right from the begin-
ning with a less than rigorous methodology for searching for studies. More
often than not, the researcher will limit his or her search to published sources
or even self-select studies to be included, based on the researcher’s familiarity
with them, quite possibly leaving many studies out of the review. This can
sometimes lead to an incorrect interpretation of the particular intervention’s
effect on crime; for example, what should have been presented as a desir-
able effect is instead reported as an uncertain effect (i.e., unclear evidence of
an effect). The one main advantage to the narrative review is that the reader
can usually glean a great deal more information about individual studies than
would otherwise be possible in the more rigorous methods of vote-count and
systematic reviews.
2. The aim was to classify all police practices into one of four categories: what
works, what does not work, what is promising, and what is unknown. What
works: These are programs that prevent crime. Programs coded as working
must have at least two level 3 to level 5 evaluations showing statistically sig-
nificant and desirable results and the preponderance of all available evidence
showing effectiveness. What does not work: These are programs that fail to
 Evidence-based policing 319

prevent crime. Programs coded as not working must have at least two level 3
to level 5 evaluations with statistical significance tests showing ineffectiveness
and the preponderance of all available evidence supporting the same conclu-
sion. What is promising: These are programs wherein the level of certainty from
available evidence is too low to support generalizable conclusions, but wherein
there is some empirical basis for predicting that further research could support
such conclusions. Programs are coded as promising if they were found to be
effective in significance tests in one level 3 to level 5 evaluation and in the
preponderance of the remaining evidence. What is unknown: Any program not
classified in one of the three above categories is defined as having unknown
effects.
3. Braga (2005) has since revised this piece to include only randomized controlled
studies. This had the effect of decreasing the sample size from nine to five
studies, but the findings did not change from the earlier systematic review.
Furthermore, Braga (2003) reported that a meta-analysis involving these five
studies supported the main finding that targeted police actions can prevent
crime and disorder in hot spots. A meta-analysis involves the statistical or
quantitative analysis of the results of prior research studies (Lipsey and Wilson
2001).


Bayley, D. H. (1998). Introduction: Effective law enforcement. In D. H. Bayley
(ed.), What works in policing (pp. 174–177). New York: Oxford University
Press.
Blumstein, A. and Wallman, J. (eds.). (2000). The crime drop in America. New
York: Cambridge University Press.
Braga, A. A. (2001). The effects of hot spots policing on crime. Annals of the
American Academy of Political and Social Science, 578, 104–125.
(2003). Personal communication, November 19, 2003.
(2005). Policing crime hot spots. In B. C. Welsh and D. P. Farrington (eds.),
Preventing crime: What works for children, offenders, victims, and places. New
York: Springer, in press.
Campbell, D. T. (1969). Reforms as experiments. American Psychologist, 24,
409–429.
Chalmers, I. (2003). Trying to do more good than harm in policy and practice:
The role of rigorous, transparent, up-to-date evaluations. Annals of the Amer-
ican Academy of Political and Social Science, 589, 22–44.
Clarke, R. V. (ed.). (1997). Situational crime prevention: Successful case studies, 2nd
ed. Guilderland, NY: Harrow and Heston.
Committee to Review Research on Police Policy and Practices. (2004). Fair-
ness and effectiveness in policing: The evidence. Washington, DC: National
Academies Press.
Cook, T. D. and Campbell, D. T. (1979). Quasi-experimentation: Design and anal-
ysis issues for field settings. Chicago: Rand McNally.
Eck, J. E. (2002). Preventing crime at places. In L. W. Sherman, D. P. Farrington,
B. C. Welsh, and D. L. MacKenzie (eds.), Evidence-based crime prevention
(pp. 241–294). New York: Routledge.
320 Brandon C. Welsh

Eck, J. E. and Maguire, E. R. (2000). Have changes in policing reduced violent

crime? An assessment of the evidence. In A. Blumstein and J. Wallman (eds.),
The crime drop in America (pp. 207–265). New York: Cambridge University
Press.
Ekblom, P. (2002). From the source to the mainstream is uphill: The challenge
of transferring knowledge of crime prevention through replication, innova-
tion and anticipation. In N. Tilley (ed.), Analysis for crime prevention: Crime
prevention studies, 13 (pp. 131–203). Monsey, NY: Criminal Justice Press.
Farrington, D. P. (1983). Randomized experiments on crime and justice. In M.
Tonry and N. Morris (eds.), Crime and justice: Annual review of research, 4
(pp. 257–308). Chicago: University of Chicago Press.
Farrington, D. P., Gottfredson, D. C., Sherman, L. W., and Welsh, B. C. (2002).
The Maryland scientific methods scale. In L. W. Sherman, D. P. Farrington,
B. C. Welsh, and D. L. MacKenzie (eds.), Evidence-based crime prevention
(pp. 13–21). New York: Routledge.
Hough, M. and Tilley, N. (1998). Getting the grease to the squeak: Research lessons for
crime prevention. Crime Detection and Prevention Series Paper 85. London:
Home Office Police Research Group.
Johnson, B. R., De Li, S., Larson, D. B., and McCullough, M. (2000). A sys-
tematic review of the religiosity and delinquency literature: A research note.
Journal of Contemporary Criminal Justice, 16, 32–52.
Kennedy, D. M. (1997). Pulling levers: Chronic offenders, high-crime settings,
and a theory of prevention. Valparaiso University Law Review, 31, 449–484.
Lab, S. P. (2003). Let’s put it in context. Criminology and Public Policy, 3, 39–44.
Liddle, H. A., Rowe, C. L., Quille, T. J., Dakof, G. A., Mills, D. S., Sakran,
E., and Biaggi, H. (2002). Transporting a research-based adolescent drug
treatment into practice. Journal of Substance Abuse Treatment, 22, 231–243.
Lipsey, M. W. and Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks,
CA: Sage Publications.
MacKenzie, D. L. (1998). Using the US land-grant university system as a model
to attack this nation’s crime problem. The Criminologist, 23 (2), 1, 3–4.
Mazerolle, L., Rogan, D., Frank, J., Famega, C., and Eck, J. E. (2002). Manag-
ing citizen calls to the police: The impact of Baltimore’s 3-1-1 call system.
Criminology and Public Policy, 2, 97–124.
McCord, J. (2003). Cures that harm: Unanticipated outcomes of crime preven-
tion programs. Annals of the American Academy of Political and Social Science,
587, 16–30.
Millenson, M. L. (1997). Demanding medical excellence: Doctors and accountability
in the information age. Chicago: University of Chicago Press.
Mosteller, F. and Boruch, R. (eds.). (2002). Evidence matters: Randomized trials
in education research. Washington, DC: Brookings Institution Press.
Shadish, W. R., Cook, T. D., and Campbell, D. T. (2002). Experimental and
quasi-experimental designs for generalized causal inference. Boston: Houghton
Mifflin.
Sherman, L.W. (1992). Attacking crime: Policing and crime control. In M. Tonry
and N. Morris (eds.), Modern policing. Crime and justice: A review of research,
15 (pp. 159–230). Chicago: University of Chicago Press.
 Evidence-based policing 321

(1997). Policing for crime prevention. In L. W. Sherman, D. C. Gottfredson,

D. L. MacKenzie, J. E. Eck, P. Reuter, and S. D. Bushway, Preventing crime:
What works, what doesn’t, what’s promising (ch. 8). Washington, DC: National
Institute of Justice, US Department of Justice.
(1998). Evidence-based policing. Ideas in American policing series. Washington,
DC: Police Foundation.
(2003). Misleading evidence and evidence-led policy: Making social science
more experimental. Annals of the American Academy of Political and Social
Science, 589, 6–19.
Sherman, L. W. and Eck, J. E. (2002). Policing for crime prevention. In
L. W. Sherman, D. P. Farrington, B. C. Welsh, and D. L. MacKen-
zie (eds.), Evidence-based crime prevention (pp. 295–329). New York:
Routledge.
Sherman, L. W., Farrington, D. P., Welsh, B. C., and MacKenzie, D. L. (eds.).
(2002). Evidence-based crime prevention. New York: Routledge.
Sherman, L. W., Gottfredson, D. C., MacKenzie, D. L., Eck, J. E., Reuter, P.,
and Bushway, S. D. (1997). Preventing crime: What works, what doesn’t, what’s
promising. Washington, DC: National Institute of Justice, US Department of
Justice.
Tita, G., Riley, K. J., and Greenwood, P. W. (2003). From Boston to Boyle
Heights: The process and prospects of a “pulling levers” strategy in a
Los Angeles bario. In S. H. Decker (ed.), Policing gangs and youth violence
(pp. 102–130). Belmont, CA: Wadsworth.
Weisburd, D. and Eck, J. E. (2004). What can police do to reduce crime, disorder,
and fear? Annals of the American Academy of Political and Social Science, 593,
42–65.
Weisburd, D., Lum, C. M., and Petrosino, A. (2001). Does research design affect
study outcomes in criminal justice? Annals of the American Academy of Political
and Social Science, 578, 50–70.
Weisburd, D., Mastrofski, S. D., McNally, A. M., Greenspan, R., and Willis, J. J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 3, 421–456.
Welsh, B. C. (2002). Technological innovations for policing: Crime prevention
as the bottom line. Criminology and Public Policy, 2, 129–132.
(2003). Community-based approaches to preventing delinquency and crime:
Promising results and future directions. Japanese Journal of Sociological Crim-
inology, 28, 7–24.
Welsh, B. C. and Farrington, D. P. (2000). Monetary costs and benefits of
crime prevention programs. In M. Tonry (ed.), Crime and justice: A review of
research, 27 (pp. 305–361). Chicago: University of Chicago Press.
(2001). Toward an evidence-based approach to preventing crime. Annals of the
American Academy of Political and Social Science, 578, 158–173.
Welsh, B. C., Farrington, D. P., Sherman, L. W., and MacKenzie, D. L.
(2002). What do we know about crime prevention? International Annals of
Criminology, 40 (1 and 2), 11–31.
Wilson, D. B. (2001). Meta-analytic methods for criminology. Annals of the Amer-
ican Academy of Political and Social Science, 578, 71–89.
17 Critic
Improving police through expertise,
experience, and experiments

Mark H. Moore

The words “expertise,” “experience,” and “experiment” all come from

the same Latin root: ex per, and it means from danger. The common root
suggests that at one time in human history, these ideas were closely con-
nected with one another. Interpreting unforgivably, we might conclude
that human beings have long been aware of the danger of acting with
uncertain knowledge of consequences; further, that an important way to
reduce this danger was to rely on the expertise that was rooted in experience
of which one important type was an experiment.
The commitment to “evidence-based policing” is today’s effort to
reduce the danger associated with relying on police practices and methods
whose consequences are, in an important sense, unknown, or at least
uncertain. The proposed solution lies in the careful accumulation of
actual concrete experience with particular police methods to discover
which methods seem to work to produce desired results. But the focus
on evidence-based, rather than experience-based knowledge suggests that it
is not just any old experience that can be used in developing a more solid
base for action. It is not, for example, the kind of experience we recog-
nize as commonsense. Nor is it the kind that accumulates as police lore. It
is not even the kind of experience captured in detailed case studies. It is
instead the kind of experience that is captured in observational studies
that reduce experience to numbers that can be systematically analyzed
to discover the generality and reliability with which a particular interven-
tion produces desired results. It may be even more particularly the kind
of experience generated by carefully designed, randomized experiments.
As a social scientist, and as a person who longs to put police work on a
more solid empirical basis, it is impossible to be against a movement that
supports “evidence-based policing.” I think it is important to find out
what works in policing. I think social science methods provide the most
powerful methods available to us to determine what works in policing as
well as in other fields. My only concern is that by focusing too much on
the experience that can be captured in quantitative observational studies
and controlled experiments – by assuming that these methods can stand

322
 Improving police 323

alone, and that they are the only ones that can provide a relatively firm
basis for action – we will end up, paradoxically, both reducing the amount
of experience that is available to us, and slowing the rate at which the field
as a whole can learn about what works in policing.
I set out this argument below in the form of a skeptical commentary
on evidence-based policing. For clarity and concreteness, I use Lawrence
Sherman’s (1998) paper on evidence-based policing as the point of depar-
ture. I understand that the idea of evidence-based policing is broader than
the approach suggested by Sherman, and that there are many others who
propose different ideas. But Sherman’s paper provides such a convenient
foil for my arguments, and is so clearly an exemplar of the commitments
that lie behind a larger movement for evidence-base medicine, evidence-
based education and evidence-based policing that it is irresistible to use
it as a target. I hope both Larry and others working on evidence-based
policing will forgive the distortion that this focus gives to my discussion.
I only hope that my arguments will have merit even when considered
against a wider and more nuanced idea of evidence-based policing.

“Scientific evidence”
Let’s begin where Sherman does: with the apparently incontrovertible
claim that “police practices should be based on scientific evidence about
what works best.” Anyone who believes in science as an important driver
of human progress has to accept this idea. Since that includes virtually all
police researchers, and no small number of police practitioners, Sherman
starts off with a large and sympathetic audience. The difficulties arise
when we begin to think hard about what we mean by “scientific evidence,”
by “police practices,” and by “works best.”
Take, first, the issue of what constitutes “scientific evidence.” In
Sherman’s view, too much current police practice is based on noth-
ing more than “local custom, opinions, theories, and subjective impres-
sions” (1998: 6). Current practices of policing are viewed as “unsystem-
atic ‘experience’” that provides the raw material for, but not the true
essence of the kind of “scientific evidence” that would put policing on a
stronger intellectual footing (Sherman 1998: 4). What lies between this
“unsystematic experience” guided by “local custom, opinions, etc.” is the
rigorous examination of those practices through scientific methods –
ideally, randomized trials of particular methods used in dealing with
“repetitive circumstances.” Thus, Sherman sets up a simple dichotomy:
mere “custom, opinion and theory” versus “scientific facts.”
The problem is that this is a false dichotomy. The options that Sherman
describes anchor the ends of a continuum of scientific methods that can
324 Mark H. Moore

be used to test the efficacy of police methods; not a stark choice between
superstitious tradition on one hand, and scientific certainty on the other.
Science, as a human disposition and as a method, has never been strictly
limited to randomized experiments; it has always included many more
different types of investigations to acquire and use knowledge.
Science begins with curiosity about the size and character of some
phenomenon in the world (for example, different kinds of crime). It con-
tinues with the development of some (as yet untested) ideas about the
causal factors that are shaping the size and character of the phenomenon
of interest (such as levels of poverty, or the development of a culture
of violence). It proceeds with measurements made through particular
instruments specially designed to make the link between a theoretical
concept, and an empirical reality. It continues with the use of statistical
techniques designed to discover the relationship, if any, between changes
in the phenomenon being observed and changes in the variables that are
thought to be causing the change.
It is worth noting that this basic method of science (interest in a
phenomenon, speculation about causes, observation and measurement,
analysis of differences) can be used not only to discover the “natural
causes” of a phenomenon as it exists in nature, but also to investigate the
consequences of a specifically imagined human intervention designed
to alter the phenomenon that is observed. Thus, scientific methods
can, in principle, be used to discover not only the causes of crime, but
also the crime control effectiveness of particular interventions made by
police departments. The search for causes is often described as basic
research, while the search for remedies and cures is described as applied
research.
But the distinction between basic and applied research should not blind
one to the fact that both basic and applied sciences depend on exploring
relationships between so-called “dependent variables” on one hand, and
“independent variables” on the other.1 The major difference is that when
we are trying to explain the causes of crime, the independent variables
are social structural and population characteristics that we hypothesize
generate a certain amount of crime. When we are trying to explore the
impact of a crime control intervention, the independent variable of par-
ticular interest is the intervention to be evaluated.2
It is also worth noting that the general form of “observational studies”
described above (in which a phenomenon is observed in nature being
affected both by natural cases and certain kinds of policy interventions)
are generally only capable of showing a correlation between changes in the
independent variables on one hand, and the dependent variables on the
other (Winship and Morgan 1999). They cannot demonstrate causation
 Improving police 325

with any high degree of reliability. To show causation, nature has to be

manipulated a bit more carefully (Heckman 1997).
Specifically, to demonstrate that a particular hypothesized factor actu-
ally caused an effect to occur; or to show convincingly that a particular
effect was the result of a particular policy intervention, we have to be able
to compare two situations that are identical in all respects except for the
fact that one situation was affected by a cause, or received a particular
intervention, and the other did not so that any observed difference in
results can be attributed to that causal variable, or that particular inter-
vention. Repeated enough times, if we find that the treated situation pro-
duces a different result than the untreated situation, we can conclude with
a high degree of confidence that the intervention produced the result we
observed. It is this simple fact that makes randomized trials the “gold stan-
dard” for reliably inferring causation whether looking for natural causes
or estimating the impact of an intervention, since randomization of con-
ditions is the best method of ensuring that the two situations that are
being compared are similar (Campbell 1969).
While everyone agrees that randomized trials are the best way of cre-
ating confident knowledge about the existence and size of an effect of
police practices on levels of crime, an important question remains about
how much we should rely on randomized trials, and how much on other,
lesser means. This would not be an important question if we thought we
could produce randomized trials to test every police practice. But there
are many things that stand in the way realizing this particular dream –
including the cost of such experiments, the technical difficulty of exe-
cuting them, and various ethical cautions that arise when we experiment
with the fates of human beings (Cook and Campbell 1979).
If, realistically, we cannot expect to test every police practice through
randomized trials, then the issue of whether it makes sense to use some
of the lesser methods of science (simple inquiry, logical thought, direct
observation of actions and apparent results, program evaluations carried
out through statistical analyses rather than randomized trials) becomes
important. Put somewhat differently, both the research and the prac-
tice field in policing face the important question of how far down the
path of scientific sophistication they should go in their combined efforts
to establish a firm experiential and empirical basis for policing. More
provocatively put, they have to decide what to do with the knowledge that
lies between mere opinion on one hand, and results established through
randomized trials on the other.3
In my preferred vision of “evidence-based policing,” the standards
of what constituted “research,” or “evidence,” or “scientific knowl-
edge” would be more open and flexible than in Sherman’s conception.
326 Mark H. Moore

Obviously, one would want to get to the “gold standard” with as many
important police practices as we could. But the central aim of a movement
for “evidence-based policing” wouldn’t be simply to push some arbitrar-
ily selected police practices to the promised land of randomized trials, it
would be to subject as many police practices as possible to increasingly
stringent tests of their efficacy through methods that seemed appropriate
and feasible.
In this conception of evidence-based policing, it would be counted as
a gain if one simply laid out the common sense logic that lay behind a
particular police practice, and subjected it to what I would describe as
“the giggle test” that measured the common-sense plausibility of a claim
of efficacy.4 It would also count if one gathered facts about particular
kinds of crimes and problems that police face that could suggest inno-
vative ideas about how that sort of crime, or that kind of problem might
best be addressed. It would be counted as a bigger gain if one actually
went ahead and captured the experience a police department produced by
recording and measuring what the department did, and what happened
later, possibly but not necessarily as a consequence of the department’s
actions. These ideas are key elements in the problem-oriented policing
model proposed by Herman Goldstein (1990). It would be counted as an
even bigger gain if the experience produced in one domain was checked
with naturally occurring experiences in other domains that could be used
as “quasi-control groups” (i.e., rough equivalents to the kind of planned
variation that one would introduce in the design of randomized experi-
ments). And so on.
The point is to recognize that different pieces of evidence about efficacy
come with different degrees of weight and credibility. The task is not only
to produce evidence that has the particularly heavy weight associated with
randomized trials; it is also to produce and responsibly use evidence that
is less strong than that produced by randomized trials.
Widening the range of acceptable methods, and learning how to
calibrate their differential weight in policymaking is important for at
least three different reasons. First, sometimes the lesser methods produce
important knowledge. This occurs when the effect of an intervention is so
large, and so obvious that one doesn’t really need a fine-grained statistical
analysis to “tease out” the effect. Second, sometimes the lesser methods
are the only practicable methods to use. There is not enough time to wait
for the result, or not enough money to pay for the experiments, or not
enough capacity to structure the world into the form demanded by ran-
domized experiments. Third, many more people can get involved in the
effort to learn what works if we work hard on the cruder parts of science
such as developing “thick descriptions” of the phenomenon of interest,
 Improving police 327

reasoning about interventions that might make sense, and remembering

to look as closely as one can at what happened when one tried a partic-
ular intervention. Developing these basic scientific orientations among
a mass audience can provide not only a valuable base for learning, but
also help to create a strong base of support for the use of more sophis-
ticated methods. The alternative approach – allowing only a few elite
folks to carry out these investigations – raises doubts and generates resis-
tance among those for whom the research is ostensibly being conducted
(Hartley, Bennington, and Binns 1997; Hartley and Bennington 2000).

Police practices
Take next the question of what constitutes a “police practice” that could
be evaluated through scientific means. Again, the scientific method works
best (most reliably, most simply, most inexpensively) when the phe-
nomenon being examined is causally simple, easy to observe, and occurs
often. Such conditions often occur in manufacturing processes where
physical inputs are converted to physical outputs through a well-defined
technological process engineered to produce a specific result. That is the
reason that statistical control systems work as well as they do in manu-
facturing operations.
In principle, one can always characterize some specific police practice
as a kind of manufacturing technology that takes inputs (human labor,
professional skill, assistance from victims and witnesses, forensic support,
etc.) and produces a potentially valuable output (e.g., arrest of a murder
suspect with enough evidence to support an effective prosecution). All
one has to do is draw a “black box” around the inputs and the outputs,
and carry out a statistical analysis of the relationship between the inputs
and the outputs to identify the relevant “production function” (see e.g.,
Pindyck and Rubenfeld 2001). Look at enough homicide investigations
and burglaries in this way, and one can identify the factors that contribute
to the solution of the case (Greenwood, Chaiken, and Petersilia 1977).
Similarly, one can imagine that we have particular approaches to par-
ticular crimes – such as arresting offenders for domestic violence – and
experiment with the use of this technique to see whether it reduces the
recurrence of these events, and thereby results in a reduction in these
kinds of crimes.
In principal, at one higher level of abstraction, one can also think of
a police department as a whole as the cumulative sum of these partic-
ular practices, brought out by the personnel of the department at the
right time when triggered by an external demand for performance. Thus,
one can conceptualize a police department as a kind of manufacturing
328 Mark H. Moore

or service facility designed to prevent crime in a particular jurisdiction,

draw a box around the police department as a whole, and ask about the
degree to which it is effective in reducing crime throughout its particular
jurisdiction.
While this is fine in principle, when one begins to try to scientifically
design and evaluate police practices, some important difficulties emerge.
The first comes in defining a particular “police practice” that is worth
scientifically evaluating.
At the outset of modern police research, this was not a particularly
difficult problem. The “professional model” of policing relied on only a
few large “practices” that were thought to be just and effective in deal-
ing with all kinds of crime. The “logic model” of professional polic-
ing hypothesized (reasonably, I think) that all kinds of crime could be
prevented and controlled through a few general methods of policing:
namely, patrol, rapid response to crime calls, and retrospective investiga-
tions (Moore 2002). By threatening offenders with arrest, such methods
could be expected to prevent and control crime through general deter-
rence. By producing arrests that led to convictions, such tactics could
prevent and control crime through the additional mechanisms of specific
deterrence, incapacitation, and (sometimes) rehabilitation.5
These police practices – patrol rapid response, retrospective investiga-
tion – represented large, significant targets for evaluative research (includ-
ing both observational studies and randomized experiments) since they
encompassed a great deal of what police departments actually did to con-
trol crime. What that research showed, however, was that these methods
had much smaller effects on levels of crime and fear in the community
than we initially thought (Kelling, Pate, Dieckman, and Brown 1974;
Greenwood et al. 1977; Spelman and Brown 1984). Nor did they result
in the apprehension of offenders as often as we thought desirable (Green-
wood et al. 1977). And they were not able to cope very well with crimes
that were invisible to the police because they did not necessarily happen in
public, or produce a victim or witness who could alert the police (Moore
1983).
One of the most important effects of these studies was to encourage the
police to search for more specialized practices designed to deal with par-
ticular kinds of crime. And it is this approach to dealing with crime that
recently won the endorsement of the National Academy of Science’s Panel
on Policing (Committee to Review Research 2004). But an important
consequence of this development for those who would like police methods
to be carefully evaluated is that the more specialized focus on particular
kinds of crime implied a dramatic increase in the number of specific police
practices that had to be developed and evaluated. Instead of evaluating a
 Improving police 329

few police practices that used most police resources and were considered
robust in dealing with all manner of crime, the research task now became
the evaluation of a much larger number of more specialized practices
designed to deal with more particular kinds of crimes. We can, and have,
looked at how the police respond to domestic violence, to serial killers,
to robberies, to street-level drug dealing, to vandalism and vagrancy, and
to mentally ill individuals who are creating disturbances, or threatening to
commit suicide, and so on (Sherman 1992; Plotkin and Narr 1993;
Rossmo 1995).
Once we begin to think that it is wrong to look at a general method
for dealing with a homogenous problem called crime, and to rely instead
on a wide array of more particular methods for dealing with a highly
heterogeneous set of crimes, the challenge of evaluating police practices
begins to resemble the problem of investigating the huge variety of med-
ical practices used to cope with different kinds of illnesses. In medicine,
instead of imagining that there is a general cure for all disease, and that
hospitals are the organizations that do only that general thing to protect
us from disease, we have begun to distinguish many different kinds of
diseases, each with their own proper response, specially tailored to the
unique circumstances of an individual. We have also come to view the
practice of medicine in hospitals as bundles of these particular practices
that are hauled out by doctors and nurses when particular cases require
them. Perhaps police departments should be seen as collections of pro-
fessional police officers organized in a particular organization that brings
out particular treatments for particular cases as needed. Ideally, then, the
knowledge base of policing would consist of the knowledge of what partic-
ular interventions work with what particular diseases, and the adaptation
of those particular interventions to the particular condition of a given
circumstance – just what the idea of problem-oriented policing prescribed
for the field.
Unfortunately, just as in medicine, only a portion of what is now
thought to constitute the “best practices” in policing with respect to par-
ticular kinds of crime have actually been tested in randomized trials. That
means that the police are still operating in the dangerous situation of rely-
ing on untested best practices. On the other hand, it also suggests that
all we have to do is to increase the rate of experimentation so science
can catch up to our current practices. That is what Sherman seems to
believe. But I think there is a deeper problem here, and the difficulties that
medicine faces in evaluating its own practices should provide a cautionary
tale to policing as it seeks a firmer basis for its professional practice.
One part of the difficulty has to do with knowing how to divide
up the world of crime problems into meaningful categories for which
330 Mark H. Moore

particular best practices can be developed. So far, much of that work

seems to be guided by common sense – not some deductive logic. For
example, we started out distinguishing domestic violence from other
kinds of aggravated assault because we thought it was a “different” kind
of problem in at least the following ways: (1) we were less likely to hear
about it from victims; (2) the victims would be less cooperative in sup-
porting arrests and prosecutions; (3) there were often repeat offenses,
so that each offense could be used as an occasion to do something that
would make the next offense more or less likely to occur, and more or less
serious; (4) we had to be worried about the effect of the intervention we
made not only on the likelihood of future violence against the victim, but
also on the impact that the intervention had on children living in close
proximity to and dependent on the victims, and so on. All these features
of domestic violence seemed to make it a special category of crime that
deserved it own diagnosis and its own treatment (Chalk and King 1998).
Yet, as we have gone deeper into the exploration of methods for dealing
with domestic violence, two surprises have occurred that raise doubts
about categorizing domestic violence as a unique kind of crime. On one
hand, we have found that other kinds of violence share some important
characteristics with domestic violence. For example, we have discovered
that many homicides and aggravated assaults among young men occur in
the context of continuing relationships that are as criminogenic as those
between domestic partners (Kennedy, Piehl, and Braga 1996). They are
rivals for turf, for the leadership of a gang, or for the affections of a young
woman. Violence they commit on one another will create another round
of violence as families and friends of the victims seek revenge. In short,
gang murders are similar to domestic violence in that it is a continuing
relationship among individuals that seems to stimulate and focus the
violence.
On the other hand, we have learned that there are many different
kinds of domestic violence cases that seem to respond differently to
different kinds of interventions. Indeed, Sherman (1992) illustrates this
phenomenon in his own account of what happened with the experiments
in policing domestic violence. We began with the idea that there was a
relatively homogenous problem called “domestic violence,” and the chal-
lenge was to determine whether a particular method – mandatory arrest –
was successful in dealing with this general problem. What we learned
(after more than $10 million in experimental studies) was: (1) that arrest
reduces domestic violence in some cities but increases it in others; (2) that
arrest reduces domestic violence among employed people but increases
it among unemployed people; (3) that arrest reduces domestic violence
in the short run but can increase it in the long run; and (4) that the police
 Improving police 331

can predict which couples are most likely to suffer future violence, but
our society seems to value privacy too highly to allow effective preventive
action.
These observations suggest that the search for effective crime control
methods may not go on as systematically or as scientifically as we might
wish. While we could follow the approach of differentiating different kinds
of crime, and looking for the best method of dealing with each particular
kind of crime, it is not at all clear how such categories of crime should
be constructed, nor how many will have to be investigated. The more
particular our characterization of crimes, the more experiments we will
have to run. On the other hand, we could start with a generalized idea
of crime, and find a method that would be effective in dealing with all
kinds of crime. Unfortunately, we tried that, and produced only mixed
results. Perhaps our best chance is the method suggested by Goldstein
(1990): namely, to learn as much as we can about the likely causes and
potential points of intervention in dealing with particular crime prob-
lems through the development of a thick description of a certain class of
crimes, and then use common sense to imagine and test interventions to
see if they work. That clinical (as opposed to scientific) approach might
be both necessary and sufficient to help us develop the knowledge we
need to deal with a highly differentiated crime problem, just as it was
necessary and sufficient to give us most of the means we now rely on in
medicine.
If it is hard to evaluate any particular police operation focused on a
particular problem, imagine how much harder it is to evaluate the per-
formance of a police department as a whole. Of course, one could rule
out the evaluation of a whole department’s performance as not the kind of
“practice” Sherman had in mind. But Sherman seems inclined to think
that the overall performance of an entire department could be viewed
as a kind of “macro-practice” that could be evaluated scientifically as
well as the kind of “micro-practices” we associate with the “best prac-
tices” that exist for dealing with particular kinds of crime.6 He suggests
this inclination by explicitly introducing the analogy to manufacturing
organizations on one hand, and by claiming that departments as a whole
could be evaluated, motivated, and guided by scientific evidence of what
works in policing, on the other. In both moves, he changes the definition
of the police practice to be evaluated from a specific operational program
designed to deal with a specific kind of crime to the operations of the
police department as a whole.
When we focus on the “scientific evaluation” of the performance of a
police department as a whole rather than the efficacy of a particular pro-
cedure in dealing with a particular kind of crime, many things change.
332 Mark H. Moore

One key difference is that the kinds of police practices that need to be
evaluated change. Police practices observed at the organizational level of
the police would include all the specific operational practices used by the
department to deal with particular crimes as defined above. But police
practices observed at the organizational level would include the adminis-
trative practices that the police use to manage themselves; the methods
they use to recruit and train personnel, the methods they use to motivate
their line commanders and individual officers to achieve organizational
goals, the methods they use to minimize corruption and abuses of force,
and so on. These administrative practices might even include the arrange-
ments the department made to evaluate their own operational practices
(Moore, Sparrow, and Spelman 1997; Moore 2003).
A second key issue, however, is that while effective crime control is
certainly one thing that citizens want from police organizations, it is also
quite clear that citizens want other things from their police as well (Moore
2002). They want to have a subjective sense of security in their streets
that may be somewhat independent of the objective risks of victimiza-
tion. They want certain kinds of services from the police that make their
individual and collective life better.
But the problem continues. Often specific police activities and oper-
ations produce results that register on many different objectives of
policing – not just one. The DARE Program, for example, has long been
evaluated as though its most important justification is the impact such
a program would have on future levels of drug use by students exposed
to the program (Esbensen, Frend, Taylor, Peterson et al. 2002). But the
most important practical effect of the DARE Program might be to build
relationships between the police on one hand, and parents and youth on
the other, that will allow them to be more effective in dealing with all
kinds of problems both encountered and created by these individuals in
the future.
There may also be synergies (positive or negative) in the activities of
police viewed across the department as a whole. For example, the strong
work of an officer who was focused on building community relations in a
particular neighborhood could be undermined in an instant by a drug raid
carried out by a centralized unit that ended up in the wrong apartment
arresting the wrong person in a way that terrorized rather than reassured
the neighborhood about the competence and intentions of the police. In
short, policing may simply be too complex an endeavor to ever be sorted
out – just as it would be too complex to sort out the operations of a public
health system that included not only emergency rooms, but also long-
stay hospitals, neighborhood clinics, and public health immunization and
well-baby programs.
 Improving police 333

“What works?”
Take, finally, the question of what we mean when we say that a particular
police practice “works.” The obvious next question is: “Works to do
what?” A question that follows closely after that is: “Works at what cost,
and with what unintended and unexpected side effects?” In answering
these questions, one is gradually constructing an analytic framework that
one could use normatively to decide whether a given police practice was
worth continuing.
This is importantly an empirical question, of course. To determine
whether a practice is a good one or not, we have to be able to describe
the effects it produces in the world. And that is a profoundly empirical
issue.
But the construction of an analytic framework for evaluating a given
police practice also requires a normative stance. To identify an effect of
a program as something that would be worth noticing in any attempt to
evaluate its value to the wider society is inherently a normative enterprise.
The effects that we consider when evaluating a program are important
precisely because they have normative significance to the world.
For example, we could observe that police crackdowns on gang mem-
bers could reduce gun violence and fatalities in a city. That alone might
seem to make a strong case for engaging in more police crackdowns on
gang members. But suppose those police crackdowns produced other
effects as well such as increased resentment and fear of the police in the
neighborhoods in which the crackdowns occurred. There would be an
empirical question to be answered about whether this was a real effect
of the police crackdown. There would be a normative question about
whether such an effect should be counted when we were considering the
overall value of the crackdowns as a police response to a social problem.
If we thought this effect was a plausible one, and that it was normatively
significant, then the analytic framework for measuring the impact of the
crackdown would have to look at the effect it had on local residents’ atti-
tudes toward the police as well as on the level of gun violence. A new
dependent variable would have to be introduced into the analysis. This is
an example of an important conclusion drawn by the National Academy
of Sciences Panel on Policing. They concluded that police practices had
to be evaluated in terms of their impact on the legitimacy of the police
as well as their crime control cost effectiveness (Committe to Review
Research 2004).
Often, we do not feel obliged to say what we mean by a police practice
that works, because we assume that we know what the point of police
practices should be: namely, to control and reduce crime. We assume
334 Mark H. Moore

that all police practices should be evaluated only in terms of this single
dimension of performance. But I think it is clear that both micro police
practices and the overall operations of macro police institutions need
to be evaluated in more dimensions than their crime control efficacy.
Some of these added dimensions focus on the output or value side of
police operations. For example, we might be as interested in the capac-
ity of a given police practice to reduce fear, and reduce the burden of
self-defense as well as on their capacity to control a particular crime.
We might also be interested in the capacity of a given police practice to
produce the kind of justice we associate with both calling offenders to
account, and protecting the rights of citizens as well as efficacy in dealing
with crime. We might even be interested in the impact that the police
practice has on perceptions of police responsiveness and fairness as well
as effectiveness in dealing with particular crime problems (see e.g., Moore
2002).
Other dimensions of evaluation focus not on the value of the outputs,
but on the costs of mounting the operation. That includes monetary
costs for sure. But it also could include costs associated with abridging
individual rights, or exacerbating a sense of unfairness in the way that
the police do their business. These costs could show up as monetary
costs (when the police were successfully sued for wrongful conduct); or
they could show up in terms of reduced effectiveness as disillusioned
citizens stopped cooperating with the police in the production of justice
and security; or they could show up in nothing more than the continued
disgust and disappointment that citizens feel when they think their tax
dollars and liberty are being abused by an organization they would like
to be able to trust.
To say that a police practice, or a police organization, works, then, is
to make a set of claims about not only what effects a police department
produces, but also what a political community that authorizes the police
department does or should want. This means that we have to admit both
values and politics into the construction of the analytic framework we
use to evaluate police practices and police departments as a whole. What
effects of a police practice should be considered important for evaluation
purposes is not, strictly speaking, a value-free scientific enterprise. A fact
taken by itself may be value neutral. And in trying to develop a fact it
may be important to push one’s values to the side as best one can. But
to develop a fact that is useful in evaluating police practices, one has
to construct a value framework that makes the fact interpretable as an
evaluation of whether something is good or bad. And that is inevitably a
value question.
 Improving police 335

How does policing improve?

In the end, both police researchers and police practitioners have to
concern themselves with the way in which policing can be improved.
An important part of that effort depends on developing more reliable
knowledge about what works. And that, in turn, depends on deploying
the methods of science to help us separate false claims of efficacy from
true claims, or to gradually increase our confidence in the methods we are
using to accomplish the goals of policing. Thus, science, and particularly
the use of science in carrying out randomized trials of police practices,
could be the “rate determining step” in determining the rate of police
improvement.
If this is true, then the principal responsibility for improving policing
seems to lie primarily in the research community. It is there that police
practitioners must look to find out what is known to work. It is there
that police practitioners must turn to develop the guidelines they use to
instruct their officers to do their work. And it is from that source that any
new doubts about the efficacy of old methods, or any new ideas about
the proven efficacy of new methods must come.
That, at least, seems to be the core of the vision of how to improve polic-
ing that is contained in the idea of evidence-based policing: improvement
through science, and the work of scientists.
I hope I can be forgiven if I suggest that this puts a bit too much
of the load on science and scientists, and that it uses the commitment
and knowledge of the police profession a bit too little. Science cannot
become useful in the practical world of policing unless the practitioners
embrace some aspects of science as an important part of the way they do
their work. We have to create the equivalent in the profession not only of
academic researchers and clinicians, but also of clinical professors, and
reflective practitioners; and they have to learn to work together in a kind
of intellectual and practical partnership to solve concrete problems as
best they can – not compete with each other over whose knowledge is
more authoritative.
Science alone cannot answer important questions that are central to the
development of effective clinical practice. Science, by itself, cannot tell
us how to divide up the world of practical problems we see in front of us
in a way that is most amenable to their solution. Science, by itself, cannot
necessarily suggest the particular intervention that should be tried to deal
with a given problem. Science, by itself, cannot tell us in what terms we
ought to evaluate the interventions we make. For all this work, we need
the experience of practitioners as well as the experiments of science.
336 Mark H. Moore

If our vision of evidence-based policing going forward is one that

embraces these aspects of science, I am all for it. If our vision is one
that seeks to privilege a certain kind of science and a particular kind of
scientist, I am against it. The world of crime and policing is far too impor-
tant, far too complex, and far too urgent to leave entirely in the hands of
scientists. We need a great deal of practical wisdom as well as a rigorous
and responsive science to move the field forward.

   
1. Karl Popper describes what is here called the dependent variable, the explanan-
dum – the thing to be explained, and the independent variables, the explanans –
or the thing that explains the explanandum (Popper, 1902).
2. James Q. Wilson (1985) has made some important observations about the
difference between criminology on one hand, and the effort to find important
means for controlling crime on the other.
3. Howard Raiffa (1924) has developed some systematic ways of thinking about
using imperfect information in practical decisions. He recommends a Bayesian
approach in which the decisionmaker begins with more or less strong prior
probability estimates of the likely effects of a given policy. He then updates
those prior probability estimates as new information comes in through obser-
vation and experimentation. Since each piece of information has a different
weight, each piece of information can be brought in to improve the estimate.
We don’t have to discard imperfect information; only discount its weight.
4. This is often accomplished as a necessary part of any program evaluation.
Wes Skogan (1990) has demonstrated the importance of developing the “logic
model” that connects a planned intervention to a desired effect as an important
step in carrying on a program evaluation. Often, once one goes through this
particular discipline, one can see errors in thought and planning even before
one gets into the field. That is what I mean by the “giggle test” – the use of
logic, common sense, and a sense of proportion to test the very plausibility of
an idea before one goes to the trouble of trying it out and carefully evaluating
it. In my experience, many police practices would be improved by taking this
very first step in science.
5. These basic ideas remain strong in policing because they align both with com-
mon sense, and with certain ideas of justice; namely that it would be just, as
well as practically useful, to call offenders to account for their crimes, and to
put them in prison. They show up these days in police tactics that encourage
aggressive preventive patrol focusing on disorder offenses.
6. Note that thinking about the idea of a police department as a macro prac-
tice could include two quite different ideas. One is that the department had a
general approach to dealing with crime that worked. That could be located in
a particular robust operational procedure such as aggressive preventive patrol.
Or it could be located in a particular administrative approach such as Comp-
stat. Or it could be located in a particular organizational strategy such as
community-oriented or problem-oriented policing. Or, it could be in some
combination of the police department having in its repertoire a wide variety
 Improving police 337

of responses that it uses, along with some principles that help it decide which
particular procedures to use.


Campbell, D. (1969). Reforms as experiments. American Psychologist 24,
409–429.
Chalk, R. and King, P. (eds.). 1998. Violence in families: assessing prevention and
treatment programs. Washington, DC: National Academy Press.
Committee to Review Research on Police Policy and Practices. (2004). Fair-
ness and effectiveness in policing: The evidence. Washington, DC: The National
Academies Press.
Cook, T. and Campbell, D. (1979). Quasi-experimentation: Design and analysis
issues for field settings. Boston: Houghton Mifflin Company.
Esbensen, F., Freng, A., Taylor, T., Peterson, D., and Osgood, D. (2002).
National evaluation of the gang resistance education and training
(G.R.E.A.T.) program. In W. Reed and S. Decker (eds.), Responding to gangs:
Evaluation and research. Washington, DC: National Institute of Justice, US
Department of Justice.
Goldstein, H. (1990). Problem-oriented policing. Philadelphia, PA: Temple
University Press.
Greenwood, P., Chaiken, J., and Petersilia, J. (1977). The investigative process.
Lexington, MA: Lexington Books.
Hartley, J. and Bennington, J. (2000). Co-research: A new methodology for new
times. European Journal of Work and Organizational Psychology, 7, 1–16.
Hartley, J., Bennington, J., and Binns, P. (1997). Researching the role of internal
change agents in the management of organizational change. British Journal
of Management, 8, 61–73.
Heckman, J. (1997). Instrumental variables: A study of implicit assumptions
used in making program evaluations. Journal of Human Resources, 33,
441–462.
Kelling, G., Pate, T., Dieckman, D., and Brown, C. (1974). The Kansas City
preventive patrol experiment: A technical report. Washington, DC: Police Foun-
dation.
Kennedy, D., Piehl, A., and Braga, A. (1996). Youth violence in Boston: Gun
markets, serious offenders, and a use-reduction strategy. Law and Contem-
porary Problems, 59, 147–196.
Moore, M. H. (1983). Invisible offenses: A challenge to minimally intrusive law
enforcement. In G. Kaplan (ed.), ABSCAM ethics: Moral issues and deception
in law enforcement. Washington, DC: Police Foundation.
(2002). Recognizing value in policing. Washington, DC: Police Executive
Research Forum.
(2003). Sizing up compstat: An important administrative innovation in polic-
ing. Criminology and Public Policy, 2, 469–494.
Moore, M. H., Sparrow, M., and Spelman, W. (1997). Innovation in policing:
From production lines to jobs shops. In A. Altshuler and R. Behn (eds.),
Innovation in American government. Washington, DC: Brookings Institution
Press.
338 Mark H. Moore

Pindyck, R. S. and Rubenfeld, D. (2001). Microeconomics (5th ed.). Upper Saddle

River, NJ: Prentice-Hall.
Plotkin, M. and Narr, T. (1993). The police response to the homeless. Washington,
DC: Police Executive Research Forum.
Popper, K. (1902). The poverty of historicism. Boston: Beacon Press [1957].
Raiffa, H. (1924). Decision analysis. New York: McGraw-Hill [1997].
Rossmo, D. K. (1995). Place, space, and police investigations: Hunting serial
violent criminals. In J. Eck and D. Weisburd (eds.), Crime and place. New
York: Criminal Justice Press.
Sherman, L. W. (1992). Policing domestic violence. New York: The Free Press.
(1998). Evidence-based policing: Ideas in American policing series. Washington,
DC: Police Foundation.
Skogan, W. (1990). Disorder and decline. New York: The Free Press.
Spelman, W. and Brown, D. (1984). Calling the police: Citizen reporting of serious
crime. Washington, DC: US Government Printing Office.
Wilson, J. Q. (1985). Thinking about crime. Revised edition. New York: Vintage
Books.
Winship, C. and Morgan, S. (1999). The estimation of causal effects from obser-
vational data. Annual Review of Sociology, 25, 659–706.
18 Conclusion: Police innovation and
the future of policing

Anthony A. Braga and David Weisburd

In this volume, a group of leading scholars presented contrasting perspec-

tives on eight major innovations in American policing developed over the
course of the 1980s and 1990s. Police departments needed to improve
their performance and innovation provided the opportunity to make these
improvements. These innovations represent fundamental changes to the
business of policing. However, as many of our authors point out, improv-
ing police performance through innovation is often not straightforward.
Police departments are highly resistant to change and police officers often
experience difficulty in implementing new programs (Sparrow, Moore,
and Kennedy 1990; Capowich and Roehl 1994; Sadd and Grinc 1994).
The available evidence on key dimensions of police performance associ-
ated with these eight innovations, such as crime control effectiveness and
community satisfaction with services provided, is also surprisingly lim-
ited. These observations are not unique to the policing field. For example,
as Elmore (1997) suggests, the field of education was awash in innova-
tion during the 1990s, but there is little evidence examining whether
those innovations advanced the performance of schools, students, or
graduates.
While our knowledge about the effects of these innovations on police
performance is still developing, we think there is much reason for opti-
mism about the future of policing. This period of innovation has demon-
strated that police can prevent crime and can improve their relationships
with the communities they serve. In the near future, we don’t anticipate
the dramatic strategic innovations that characterized the last two decades.
Rather, we expect further refinement of our knowledge of “what works”
in policing, under what circumstances particular strategies may work, and
why these strategies are effective in improving police performance. The
challenge for the future of policing is to continue making progress in fur-
ther developing and implementing promising strategies while addressing
the new problems of public safety that have been created by 9/11 and the
concerns that it has raised about the threat of terrorism and the need for
police commitment to homeland security.

339
340 Anthony A. Braga and David Weisburd

Categorizing recent police innovations

Moore, Sparrow, and Spelman (1997) suggest four distinct categories
of police innovation: programmatic, administrative, technological, and
strategic. These categories are not clearly separated from each other and,
as Moore and his colleagues admit (1997), assigning any one innovation
to one category over another is often a judgment call. Programmatic inno-
vations establish new operational methods of using the resources of an
organization to achieve particular results. These programs can include
arresting fences as a way to discourage burglary, using police officers
to provide drug education in the schools, and offering victim-resistance
training to women. Administrative innovations are changes in how police
organizations prepare themselves to conduct operations or account for
their achievements. These include new ways of measuring the perfor-
mance of an individual officer or the overall department as well as changes
in personnel policies and practices such as new recruiting techniques, new
training approaches, and new supervisory relations. Technological inno-
vations depend on the acquisition or use of some new piece of capital
equipment such as non-lethal weapons, DNA typing, or crime mapping
software.
Strategic innovations represent a fundamental change in the overall phi-
losophy and orientation of the organization (Moore et al. 1997). These
changes involve important redefinitions of the primary objectives of polic-
ing, the range of services and activities supplied by police departments,
the means through which police officers achieve their goals, and the key
internal and external relationships that are developed and maintained by
the police. Strategic innovations include shifting from “law enforcement”
to “problem solving” as a means of resolving incidents, forming work-
ing relationships with community groups as a tactic in dealing with drug
markets, and recognizing citizen satisfaction as an important performance
measure. These innovations are strategic because they involve changing
some of the basic understandings about the ends or means of policing
or the key structures of accountability that shaped overall police efforts
under the standard model of policing (Moore et al. 1997). We feel that the
eight innovations described in this volume represent related attempts to
change the ends and means of policing and, therefore, should be regarded
as strategic innovations.
Weisburd and Eck (2004) suggest that recent strategic innovations
expand policing beyond standard practices along two dimensions:
diversity of approaches and level of focus (see Figure 18.1). The “diver-
sity of approaches” dimension represents the content of the practices
employed or tools used by the police. As represented by the vertical axis,
 Conclusion 341

Wide array
Community Problem-oriented
Diversity of Approaches

Policing Policing
Mostly law enforcement

Hot Spots Policing

Standard
Model

Low High

Level of Focus

Figure 18.1 Dimensions of policing strategies

Source: Adapted from Weisburd and Eck 2004

tools can range from mostly traditional law enforcement to a wide array
of approaches. The horizontal axis represents the extent to which police
practices are focused or targeted. Weisburd and Eck (2004) contrast stan-
dard police practices with hot spots policing, problem-oriented policing,
and community policing. The standard model of policing, with its empha-
sis on enforcing the law and its generalized application of law enforcement
powers, scores low on both dimensions. Hot spots policing scores high on
focus, but low on the diversity of tools used to control hot spot locations.
Problem-oriented policing rates high on diversity of tools and focus as
the approach challenges police officers to implement strategies designed
to deal with the underlying conditions that give rise to discrete crime
problems. Community policing, where police draw on a wider array of
resources to prevent crime and engage the community in defining and
dealing with problems, scores high on diversity of approaches. However,
when implemented without problem-oriented policing, the approach is
not well focused on crime problems and provides a common set of ser-
vices throughout a jurisdiction.
Another dimension that could be added to Weisburd and Eck’s (2004)
classification of police practices is the degree to which the innovations
342 Anthony A. Braga and David Weisburd

change the goals of policing. Under the standard model, police depart-
ments were mostly focused on preventing serious crime by deterring
and apprehending criminal offenders, serving justice by holding offend-
ers accountable for their crimes, rendering immediate aid to people in
crisis, and providing non-emergency services such as controlling traffic
(Eck and Rosenbaum 1994). While the eight innovations described in
this book do not remove any of these goals from the tasks of policing,
the new strategies rearrange the priorities among the goals and add new
ones. Non-criminal and non-emergency quality of life problems receive
much more attention from the new police strategies. Community and
problem-oriented policing represent the most radical departures from
standard police work. Community policing, in its various manifestations,
challenges police officers to work with citizens to deal with a broader
range of concerns, most notably fear of crime and social and physical
disorder (Skogan, in this volume). Problem-oriented policing similarly
adds new goals to policing, but it also reorganizes police actions from
focusing on incidents as units of work to focusing on classes of prob-
lems to be addressed by responses that can be quite different from rou-
tine police activities (Eck in this volume). Other innovations represent
less dramatic changes to standard police goals. For example, disorder
policing, if engaged without community and problem-oriented polic-
ing, expands the police mandate to include social and physical disorder
but does not radically change the tactics engaged by the police to deal
with these problems (Sousa and Kelling in this volume; Taylor in this
volume).

Crime and disorder control effectiveness

When police departments focus their efforts on identifiable risks, such
as crime hot spots, repeat victims, and serious offenders, they are able
to prevent crime and disorder (Braga 2002; Eck 2003). The strongest
evidence comes from evaluations of hot spots policing initiatives
(Weisburd and Braga 2003; Weisburd and Braga, in this volume). Braga
(2001) presents evidence from five randomized controlled experiments
and four quasi-experimental designs to show that hot spots policing pro-
grams generate crime control gains without significantly displacing crime
to other locations. Instead, in the five evaluations that examined immedi-
ate spatial displacement, hot spots policing initiatives were more likely to
generate a “diffusion of crime control benefits” to areas immediately sur-
rounding the targeted hot spots (Clarke and Weisburd 1994). While the
rigor of evaluation designs varies from simple before-after comparisons
without control groups to randomized experiments, problem-oriented
 Conclusion 343

policing, when appropriately focused on specific crime problems, has

been found to be effective in preventing crime (Sherman and Eck 2002;
Weisburd and Eck 2004; Eck in this volume). The available scientific
evidence on third-party policing is derived from a similar mix of stud-
ies with varying degrees of rigor. Nonetheless, Mazerolle and Ransley
(in this volume) report that third-party policing is effective in dealing with
drug problems, violent crime problems, and problems involving young
people. Pulling levers strategies also seem to be promising in controlling
the violent behavior of groups of chronic offenders (Braga et al. 2001;
Kennedy in this volume). However, the research base for this approach
is very limited (Braga and Winship in this volume; Wellford, Pepper, and
Petrie 2005).
As a general strategy, community policing has not been found to be
effective in preventing crime (Mastrofski in this volume). The avail-
able research shows that unfocused community-oriented tactics such as
foot patrol, storefront offices, newsletters, and community meetings do
not reduce crime and disorder (Committee to Review Research 2004;
Weisburd and Eck 2004). However, as will be discussed below, there is
strong evidence to suggest that community policing tactics reduce fear of
crime (Committee to Review Research 2004; Weisburd and Eck 2004).
The available empirical evidence on the crime control effectiveness of
broken windows policing is mixed (Sousa and Kelling in this volume;
Taylor in this volume). It remains unclear whether police departments
that engage a broad-based broken windows policing strategy actually
reduce crime. Simple analyses of crime trend data suggest that cities expe-
rience decreases in crime after their police departments adopt Compstat
(Silverman in this volume). However, since Compstat programs are
often implemented in conjunction with other crime prevention initia-
tives such as broken windows and hot spots policing, it is very difficult to
untangle the influence of Compstat on any observed crime control gains
(Weisburd, Mastrofski, Willis, and Greenspan in this volume). More-
over, in New York City and three other cities, further analysis revealed
the observed decreases in crime began before the implementation of
Compstat (Eck and Maguire 2000; Weisburd, Mastrofski, McNally et al.
2003; Weisburd et al. in this volume). Compstat has yet to be proven
as an effective crime control strategy in cities that have adopted the
approach.
Evidence-based policing has not been empirically tested as an overall
model of policing (Welsh in this volume). However, evidence-based police
departments would draw policies and practices from a solid research
base of strategies that have proven to be effective in controlling crime
(Sherman 1998). While an evidence-based approach to policing may
344 Anthony A. Braga and David Weisburd

have the unintended effect of limiting the ability of police to innovate

by privileging evidence over experience (Moore in this volume), we do
not believe that engaging an evidence-based approach would undermine
the crime and disorder control effectiveness of police departments.

Community reaction to innovative police strategies

In addition to concerns over the crime control effectiveness of the stan-
dard model of policing, police innovation in the 1980s and 1990s was
also driven by high levels of community dissatisfaction with police ser-
vices and a growing recognition that citizens had other concerns that
required police action, such as fear of crime. Since citizen involvement
in policing is a core element of community policing programs (Skogan in
this volume), it is not surprising that we know most about citizen reaction
to these types of programs. In general, broad-based community policing
initiatives have been found to reduce fear of crime and improve the rela-
tionships between the police and the communities they serve (Committee
to Review Research 2004; Weisburd and Eck 2004). Community polic-
ing strategies that entail direct involvement of citizens and police, such
as police community stations, citizen contract patrol, and coordinated
community policing, have been found to reduce fear of crime among
individuals and decrease individual concern about crime in neighbor-
hoods (Pate and Skogan 1985; Wycoff and Skogan 1986; Brown and
Wycoff 1987).
Community policing also enhances police legitimacy. Citizen support
and cooperation are closely linked to judgments about the legitimacy of
the police (Tyler 2004). When citizens view the police as legitimate legal
authorities, they are more likely to cooperate and obey the law (Tyler
1990). Public judgments about the legitimacy of the police are influenced
by their assessments of the manner in which the police exercise their
authority (Tyler 1990; 2004). The available evidence suggests that the
police generally obey the laws that limit their power (Skogan and Meares
2004). However, minorities consistently express significantly lower confi-
dence in the police when compared to whites (Tyler 2004). Community
policing improves citizens’ judgments of police actions (Skogan in this
volume). For example, over an eight-year period of community policing,
Chicago residents’ views of their police improved on measures of their
effectiveness, responsiveness, and demeanor (Skogan and Steiner 2004).
Importantly, these improvements were shared among Latinos, African-
Americans, and whites (Skogan and Steiner 2004). Clearly, community
policing has been a strategic innovation that has helped bridge the police
confidence gap in minority communities.
 Conclusion 345

While there is a growing body of systematic research on the effects

of community policing on citizen satisfaction with the police, there is a
noteworthy lack of research assessing the effects of other police inno-
vations on police–community relations. This gap in knowledge is note-
worthy as many of the contributions to this volume suggest a tension
between the crime prevention effectiveness of focused police efforts and
their potential harmful effects on police–community relations (Braga and
Winship; Mazerolle and Ransley; Meares; Rosenbaum; Taylor; Weisburd
and Braga, all in this volume). Certainly, legitimacy is linked to the ability
of the police to prevent crime and keep neighborhoods safe. However, the
police also need public support and cooperation to be effective in prevent-
ing crime. While residents in neighborhoods suffering from high levels
of crime often demand higher levels of enforcement, they still want the
police to be respectful and lawful in their crime control efforts (Skogan
and Meares 2004; Tyler 2004). Residents don’t want family members,
friends, and neighbors to be targeted unfairly by enforcement efforts or
treated poorly by overaggressive police officers. If the public’s trust and
confidence in the police is undermined, the ability of the police to prevent
crime will be weakened by lawsuits, declining willingness to obey the law,
and withdrawal from existing partnerships (Tyler 1990; 2001). The polit-
ical fallout from illegitimate police actions can seriously impede the ability
of police departments to engage innovative crime control tactics.
This dilemma has been described elsewhere as “the trust dilemma”
(Altshuler and Behn 1997). Innovation may be necessary for establish-
ing public faith in the ability of government agencies to perform. But
before the public grants government agencies a license to be truly inno-
vative, it needs to be convinced that these same agencies have the ability
to perform (Altshuler and Behn 1997). Police departments should be
encouraged to pursue effective strategies that aggressively focus on iden-
tifiable risks such as hot spots, repeat victims, and high-rate offenders.
However, police departments must be careful in their application of these
approaches to crime prevention. For example, anecdotal evidence sug-
gests that broken windows policing strategies enjoy broad community
support as a legitimate way to reduce crime and disorder (Sousa and
Kelling in this volume). However, when the broken windows approach
is distorted into so-called zero-tolerance policing, indiscriminate and
aggressive law enforcement can negatively affect police–community rela-
tions (Taylor in this volume). To avoid engaging tactics that will gen-
erate strong negative community reaction, police departments should
encourage and embrace community involvement in their crime preven-
tion efforts. In Boston, the involvement of black ministers in the police-
led pulling levers violence prevention strategy allowed law enforcement
346 Anthony A. Braga and David Weisburd

agencies to pursue more intrusive and aggressive tactics that would not
have been possible without community involvement (Braga and Winship
in this volume).

Police reaction to innovative strategies

The eight innovations differed in their degree of departure from the stan-
dard model of policing. The police most easily adopt innovations that
require the least radical departures from their hierarchical paramilitary
organizational structures, continue incident-driven and reactive strate-
gies, and maintain police sovereignty over crime issues. In its most
basic form, hot spots policing simply concentrates traditional enforce-
ment activity at high crime places. The familiarity of the approach
to police is straightforward as they have a long history of temporarily
heightening enforcement levels in problem areas. While law enforcement
tools are deployed in a new way, the pulling levers deterrence strategy
focuses existing criminal justice activities on groups of chronic offenders.
Broken windows policing involves making arrests of minor offenders to
control disorder and, as an end product, reduce more serious crime. As
Kennedy (in this volume) observes, “law enforcement likes enforcing the
law.” Strategies such as hot spots, broken windows, pulling levers polic-
ing appeal to law enforcement practitioners primarily because they allow
mostly traditional tactics to be deployed in new ways with the promise
of considerably greater results. Compstat, as implemented by most
American police agencies, has been focused more on reinforcing and
legitimating the traditional bureaucratic military model of police orga-
nization than on innovation in the practices of policing (Weisburd et al.
2003; Weisburd et al. in this volume).
While all major American police agencies report some form of com-
munity policing as an important component of their operations (Bureau
of Justice Statistics 2003), the police have been generally resistant to
its adoption. This is not surprising since community policing involves
the most radical change to existing police organizations. Skogan (in this
volume) and Mastrofski (in this volume) report many shortcomings in the
practical application of its three core elements: citizen involvement, prob-
lem solving, and decentralization. Citizens are generally used as informa-
tion sources rather than engaged as partners in producing public safety.
Officers prefer law enforcement strategies to developing and implement-
ing alternative problem-oriented responses. Most “community-oriented”
police agencies haven’t made the organizational changes necessary to
decentralize decisionmaking authority to the neighborhood level. Simi-
larly, the available research on problem-oriented policing suggests that
 Conclusion 347

police officers experience difficulty during all stages of the problem-

oriented process (Braga and Weisburd in this volume). Problem anal-
ysis is generally weak and implemented responses largely consist of tra-
ditional enforcement activities. Problem-oriented policing as practiced
in the field is but a shallow version of the process recommended by
Herman Goldstein (1990). Given its close relationship to community
and problem-oriented policing, it seems likely that police departments
engaging third-party policing will encounter similar practical problems.
It is not remarkable that the strategies that require the most radical
changes to existing police practices and structures report the greatest dif-
ficulties in implementation. Nonetheless, the available evidence indicates
a gradual transformation in police attitudes toward adopting these new
strategies. In addition to the widespread reporting of innovative police
practices across the United States, police officers’ views toward the com-
munity and problem-oriented policing philosophy are becoming more
positive. As summarized by Skogan (in this volume), studies point to
positive changes in officers’ views once they are involved in community
policing, positive findings with respect to job satisfaction and views of
the community, and growing support for community policing in districts
that engage the strategy compared to districts that maintain traditional
activities. Police history shows that it takes a long time for new models
of policing to fully develop. The standard model of policing was itself
a reform in reaction to corrupt and brutal police practices during the
so-called “political era” of policing (Walker 1992). Initially, the reform
movement progressed very slowly; in 1920, only a few departments could
be labeled “professional” or engaging in the basic tenets of the standard
model. It wasn’t until the 1950s that virtually all American police depart-
ments were organized around the principles set forth by O. W. Wilson,
August Vollmer, and other reformers (Walker 1992).

The future of policing

What will the future bring? Will the police continue to innovate at a rapid
pace? We don’t anticipate a new wave of strategic police innovation in the
near future. The current context of policing suggests that future innova-
tion will be incremental in nature. The conditions in the 1980s and 1990s
that created the pressure for innovation simply no longer exist. Indeed,
the atmosphere is precisely the opposite of earlier decades. Crime is down
and federal funds available for demonstration projects to spur innovation
are very limited. While the available research evidence is not as strong as
some police executives believe, there is a general sense that these police
innovations work in preventing crime and satisfying community concerns.
348 Anthony A. Braga and David Weisburd

This perspective on the crime control effectiveness of new police prac-

tices is reinforced by the modest research evidence described in this
volume and by a cursory examination of crime trends over the past decade
(see Figure 18.2). The Federal Bureau of Investigation’s Uniform Crime
Reports reveals a 29 per cent decrease in the Index crime rate from 5820
per 100,000 residents in 1990 to 4118 per 100,000 residents in 2002
(www.bjsdata.ojp.usdoj.gov). While no single factor, including innovative
policing, can be invoked as the cause of the crime decline of the 1990s
(Blumstein and Wallman 2000), the “nothing works” view of policing in
the 1970s and 1980s described in our introduction is no longer a topic
of discussion in most policing circles.
Many of these innovative changes to policing appear positive and have
shown crime control and community benefits. However, this volume
reveals the need for greater research and knowledge about the effects of
these innovations on police departments and the communities they serve.
Relative to other criminal justice institutions, the police are very open to
research and evaluation activities with universities and other research
institutions (Committee to Review Research 2004). We believe that the
police will continue to work with researchers to better understand crime
problems, community concerns, police behavior, and structural issues in
their organizations. These collaborations will support the police in further
refining their practices to become even more effective, fair, efficient, and
accountable agencies in the twenty-first century (Committee to Review
Research 2004).
Over the next couple of decades, we anticipate that individual police
departments will continue to institutionalize innovative practices by mak-
ing administrative adjustments to their organizations and by developing
a set of supporting strategies that fit the nature of their crime problems in
the neighborhoods they police. Administrative arrangements and portfo-
lios of crime prevention interventions will necessarily vary across depart-
ments as the police become more specialized in dealing with local crime
problems. In essence, we believe police departments will continue their
evolution from “production lines” that engage a static set of processes
that are used over and over again to produce the same result to “job
shops” where each police assignment is treated as a new challenge that
might require a new solution (Moore et al. 1997).
We also anticipate some modest innovation in the development of
systems to measure the performance of police departments. As police
departments engage a broader set of tactics to deal with a wide range
of community problems and concerns, there will be a need for sensi-
ble performance measurement that captures the value created by police
along a number of dimensions such as reducing criminal victimization;
 8000

7000

6000

5000

4000

Rate per 100,000

3000

2000

1000

0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Year

Figure 18.2 Index crime rates per 100,000 residents in the United States, 1990–2002
Source: Federal Bureau of Investigation, Uniform Crime Reports
350 Anthony A. Braga and David Weisburd

calling offenders to account; reducing fear and enhancing personal

security; guaranteeing safety in public spaces; using financial resources
fairly, efficiently, and effectively; using force and authority fairly, effi-
ciently, and effectively; and satisfying customer demands/ achieving legit-
imacy with those policed (Moore 2002). Appropriate measurement plays
a vital role in transforming police departments into the learning labora-
tories they are now positioned to become in the future (Maguire 2004).
American police departments, however, will be challenged to maintain
their current trajectory by the new set of homeland security demands
created in the wake of the 9/11 tragedy. In many ways, this is a new crisis
for police departments, as their goals will be further expanded by a new
focus on preventing future terrorist attacks and dealing with potentially
catastrophic events. On the one hand, this new set of demands, with its
emphasis on collecting intelligence on terrorist networks, apprehending
terror operatives, and protecting likely targets, may push policing back to
a more professional model that is distant from the community. Indeed,
there is real potential for a backward shift as federal financial support and
attention has been directed toward enhancing local law enforcement’s
role in maintaining homeland security while, at the same time, funding
for community crime prevention efforts has been drastically reduced. On
the other hand, this crisis may create a new source for innovation as police
departments will strive to continue their recent success in dealing with
crime and community concerns. The US Department of Justice Com-
munity Oriented Policing Services office has already sponsored working
group sessions and conferences on using community policing strategies
to respond to the challenge of homeland security (US Department of
Justice 2004).
Over the last two decades, the police industry has undergone radical
changes in the ends and means of policing. This period of innovation has
yielded a set of very promising strategies that can improve the ability of
the police to prevent crime and enhance their relationships with the com-
munities they serve. Police departments will be challenged to continue
developing these new approaches while meeting the homeland security
demands of the post-9/11 world. Nonetheless, we believe that the future
is promising for police agencies, as they will continue to evolve into more
effective and legitimate governmental institutions.


Altshuler, A. and Behn, R. (1997). The dilemmas of innovation in American
government. In A. Altshuler and R. Behn (eds.), Innovations in American
government: Challenges, opportunities, and dilemmas. Washington, DC:
Brookings Institution Press.
 Conclusion 351

Blumstein, A. and Wallman, J. (2000). The recent rise and fall of American
violence. In A. Blumstein and J. Wallman (eds.), The crime drop in America.
New York: Cambridge University Press.
Braga, A. (2001). The effects of hot spots policing on crime. Annals of the American
Academy of Political and Social Science, 578, 104–125.
(2002). Problem-oriented policing and crime prevention. Monsey, NY: Criminal
Justice Press.
Braga, A., Kennedy, D., Waring, E., and Piehl, A. (2001). Problem-oriented
policing, deterrence, and youth violence: An evaluation of Boston’s operation
ceasefire. Journal of Research in Crime and Delinquency, 38, 195–225.
Brown, L. and Wycoff, M. (1987). Policing Houston: Reducing fear and improv-
ing service. Crime and Delinquency, 33, 71–89.
Bureau of Justice Statistics. (2003). Local police departments 2000. Washington,
DC: Bureau of Justice Statistics, US Department of Justice.
Capowich, G. and Roehl, J. (1994). Problem-oriented policing: Actions and effec-
tiveness in San Diego. In D. Rosenbaum (ed.), The Challenge of Community
Policing: Testing the Promises. Thousand Oaks, CA: Sage Publications.
Clarke, R.V. and Weisburd, D. (1994). Diffusion of crime control benefits: Obser-
vations on the reverse of displacement. Crime prevention studies, 2, 165–183.
Committee to Review Research on Police Policy and Practices. (2004).
Fairness and effectiveness in policing: The evidence. Washington, DC: National
Academies Press.
Eck, J. (2003). Police problems: The complexity of problem theory, research and
evaluation. In J. Knutsson (ed.), Problem-oriented policing: From innovation to
mainstream. Monsey, NY: Criminal Justice Press.
Eck, J. and Maguire, E. (2000). Have changes in policing reduced violent crime?
An assessment of the evidence. In A. Blumstein and J. Wallman (eds.), The
crime drop in America. New York: Cambridge University Press.
Eck, J. and Rosenbaum, D. (1994). The new police order: Effectiveness, equity,
and efficiency in community policing. In D. Rosenbaum (ed.), The chal-
lenge of community policing: Testing the promises. Thousand Oaks, CA: Sage
Publications.
Elmore, R. (1997). The paradox of innovation in education: Cycles of reform and
the resilience of teaching. In A. Altshuler and R. Behn (eds.), Innovations in
American government: Challenges, opportunities, and dilemmas. Washington,
DC: Brookings Institution Press.
Goldstein, H. (1990). Problem-oriented policing. Philadelphia, PA: Temple
University Press.
Maguire, E. (2004). Police departments as learning laboratories: Ideas in American
policing series. Washington, DC: Police Foundation.
Moore, M. (2002). Recognizing value in policing. Washington, DC: Police Execu-
tive Research Forum.
Moore, M., Sparrow, M., and Spelman, W. (1997). Innovations in policing: From
production lines to job shops. In A. Altshuler and R. Behn (eds.), Innovations
in American government: Challenges, opportunities, and dilemmas. Washington,
DC: Brookings Institution Press.
Pate, T. and Skogan, W. (1985). Coordinated community policing: The Newark
experience. Technical Report. Washington, DC: Police Foundation.
352 Anthony A. Braga and David Weisburd

Sadd, S. and Grinc, R. (1994). Innovative neighborhood oriented policing: An

evaluation of community policing programs in eight cities. In D. Rosenbaum
(ed.), The challenge of community policing: Testing the promises. Thousand Oaks,
CA: Sage Publications.
Sherman, L.W. (1998). Evidence-based policing: Ideas in American policing series.
Washington, DC: Police Foundation.
Sherman, L.W. and Eck, J. (2002). Policing for prevention. In L. Sherman,
D. Farrington, B. Welsh, and D. MacKenzie (eds.), Evidence-based crime
prevention. New York: Routledge.
Skogan, W. G. and Meares, T. (2004). Lawful policing. Annals of the American
Academy of Political and Social Science, 593, 66–83.
Skogan, W.G. and Steiner, L. (2004). Community policing in Chicago, year ten.
Chicago: Illinois Criminal Justice Information Authority.
Sparrow, M., Moore, M., and Kennedy, D. (1990). Beyond 911: A new era for
policing. New York: Basic Books.
Tyler, T. (1990). Why people obey the law. New Haven, CT: Yale University Press.
(2001). Public trust and confidence in legal authorities: What do majority and
minority group members want from legal authorities? Behavioral Sciences and
the Law, 19, 215–235.
(2004). Enhancing police legitimacy. Annals of the American Academy of Political
and Social Science, 593, 84–99.
US Department of Justice. (2004). Applying community policing principles post 9/11:
Homeland security working session. Washington, DC: US Department of Jus-
tice, Office of Community Oriented Policing Services.
Walker, S. (1992). The police in America: An introduction. (2nd ed.). New York:
McGraw-Hill.
Weisburd, D. and Braga, A. (2003). Hot spots policing. In H. Kury and O. Fuchs
(eds.), Crime Prevention: New Approaches. Mainz, Germany: Weisser Ring.
Weisburd, D. and Eck, J. (2004). What can police do to reduce crime, disorder,
and fear? Annals of the American Academy of Political and Social Science, 593,
42–65.
Weisburd, D., Mastrofski, S., McNally, A.M., Greenspan, R., and Willis, J.
(2003). Reforming to preserve: Compstat and strategic problem solving in
American policing. Criminology and Public Policy, 2, 421–456.
Wellford, C., Pepper, J., and Petrie, C. (eds.). (2005). Firearms and violence: A
critical review. Committee to Improve Research Information and Data on
Firearms. Committee on Law and Justice, Division of Behavioral and Social
Sciences and Education. Washington, DC: The National Academies Press.
Wycoff, M. and Skogan, W. (1986). Storefront police offices: The Houston
field test. In D. Rosenbaum (ed.), Community crime prevention: Does it work?
Thousand Oaks, CA: Sage Publications.
 Index

abusive policing 254 Baltimore Police Department 277

Academy of Criminal Justice Sciences 238 Bardach, E. 184
accountability 171–187, 287 Baumer, E. 226
ACLU (American Civil Liberties Union) Bayley, D. H. 147, 227, 290, 305
217 police strategies and 1, 9, 10, 14, 16
adult homicide in Boston 177 Beat Health program 106, 137
African-Americans 4, 30, 31, 344 Beavers, Alderman William (Chicago) 217
Agnew, R. 226 Beckman, K. 160
AGVU (Anti-Gang Violence Unit) 181 Beenleigh (Australia) Calls for Service
Ajzen, I. 50 Project 232
Akers, R. 226 Behn, R. D. 3, 345
Alcohol Management Accord 197 Benitez, C. 53
Alcohol, Tobacco and Firearms(ATF), Bennington, J. 327
Bureau of 179 Bennis, J. 53
Altschuler, A. A. 3, 345 Bennis v. Michigan 210–211
American Bar Foundation 91, 92, 93 Bergman, W. T. 49
American Civil Liberties Union (ACLU) Berrien, J.
217 pulling levers and 183, 185, 342
American Society of Criminology 171 Boston 180, 181, 182, 183
analysis 7, 124–125, 136–140, 248–249 BIDs (business improvement districts) 77,
“problem-analysis triangle” 120 91
Anand, G. 181 Biebel, E. 35
Anderson, D. C. 125, 274, 276, 295 Binns, P. 327
Andrews, C. 134 Bittner, E. 63, 123, 284, 290
Annan, S. 253 Black, D. 107
Anti-Gang Violence Unit (AGVU) 181 Blacks 213
Antunes, G. E. 9 Block, C. 229
assessment 141–144 Blumstein, A. 94, 102, 312, 348
ATF (Bureau of Alcohol, Tobacco and Boba, R. 120, 136, 147
Firearms) 179 Boritch, H. 107
Atlanta 83, 279 Boruch, R. 309
Atlantic 77, 79, 85, 89 Boston 92, 168, 345
Attorney’s Office (US) 179 homicide in 175, 177
Australia 274 pulling levers and 16, 155, 156, 157
Ayres, I. 194 accountability and 171–183, 187
impact of 158
Baltimore 119 Boston City Hospital 180
broken windows and 85, 100, 107, 109 Boston Community Centers’ streetworker
Compstat and 274, 277, 278, 281 program 179
pulling levers and 158, 168, 171, 178, Boston Gun Project 155, 178, 179, 247
184 Operation Ceasefire and 173–174

353
354 Index

Boston Police Department 92, 149, 179, Broward County 279

185 Brown, C. E. 6, 227, 328
accountability and 180, 181, 182, 183 Brown, Dale K. 7, 8, 227
Operation Ceasefire and 172, 173, 174 Brown, E. O. 227
Boston Strategy II 174 Brown, L. 344
Bowling, B. 80 Bryant, K. A. 50
Braga, A. A. 12, 16, 36, 86 Buerger, M. E. 52, 92, 105, 134, 135
Boston and 171, 172, 173, 184 hot spots and 17, 246
Operation Ceasefire 173, 174, 178 third-party policing and 191, 198
Compstat and 293, 295 Bundy, McGeorge 5
evidence and 319, 330 Bureau of Alcohol, Tobacco and Firearms
future and 345, 346 (ATF) 179
crime and disorder 342, 343 Bureau of Justice Statistics 12, 346
hot spots and 231, 232, 234 community policing and 27, 29, 30, 33,
limits 246, 247, 252, 253, 258 35, 37, 40
principles, practice and 133, 145, 146, Bursik Jr, R. J. 226, 250
147, 149 Bushway, S. 101, 229, 256
problem definition 134, 137, 141 business improvement districts (BIDs) 77,
problem-oriented policing and 121, 124, 91
125, 347 Butterfield, F. 40
evolution 119, 120, 121 Bynum, T. 121, 135, 136, 253
pulling levers and 118, 155, 158, 161,
162, 164, 165, 168 Caeti, T. 232
groups and networks 162, 164 Cahn, M. F. 7
Braithwaite, J. 192, 193, 194 California 86, 158
Braman, D. 207 Cameron, J. 121
Brandl, S. G. 67 Campbell Collaboration 310
Brantingham, P. J. 227, 228, 250 Crime and Justice Group 310, 313
Brantingham, P. L. 227, 228, 250 Campbell, D. T. 248, 325
Bratton, Police Commissioner William 11, evidence-based policing and 307, 308,
17, 18, 52, 148 310, 312
broken windows and 77, 92, 102 Canadian policing 107
Compstat and 284, 290, 291, 295 Capowich, G. E. 38, 134, 137, 140, 339
innovation 270, 272, 274, 280 CAPS (Chicago Alternative Policing
promise of 286, 287, 288 Strategy) 258
Briggs, L. R. 17, 92, 229 Casey, J. D. 9
Brisbane 197 Caulkins, J. P. 215
Brisbane City Council 197 causal reasoning 87–88
Britain 200, 204, 228 Ceasefire working group 174, 183
broken windows policing 14–15, 92, see also Operation Ceasefire
103–106 Centre for Problem-Oriented Policing 121
criminology and criminal justice 77–95 Chacon, R. 182
causal reasoning and 87–88 Chaiken, J. 9, 227, 327, 328
crime 91–93 Chainey, S. 121
discourses 77–80, 81–87 Chalmers, I. 315
moral complexity and 88–90 Chandek, M. S. 46
policy options and 90–91 Chandler (AZ) 87
incivilities reduction, zero tolerance and change and crisis 11–18
coproduction 98–111 character, understanding form and 13–18
coproduction 106–107, 109–110 Charlotte-Mecklenburg Police Department
incivilities thesis 98–103, 111 149
zero tolerance 107–108 Chatterton, M. 123
model 255 CHEERS (Community, Harm,
Bromley, M. L. 51 Expectations, Recurring and
Brooklyn 271 Similarity) 120
 Index 355

Cheh, M. M. 194, 200, 211 community policing and 45, 56, 63, 64
Chenery, S. 125 organization and practice 51, 54, 55
Chermak, S. 46, 125 evidence-based policing 313, 328, 333
pulling levers and 155, 158, 161, 162, future and 343, 344, 348
178 hot spots and 231, 234
Chicago 83, 84, 100, 148, 344 problem-oriented policing and 122, 127,
community policing and 52, 53, 54, 133, 145
59–60 communication and relationship, common
promise of 28, 30, 31, 32, 33, 34, ground 166–167
36–40 Communities, Department of (Brisbane)
hot spots and 246, 258 197
third party policing 211, 212, 217 community
Chicago Alternative Policing Strategy -police relations 253–254
(CAPS) 258 problem-oriented theories and 250–251
Chicago City Council 212, 217 reaction to innovation 344–346
Committee on Police and Fire 107 Community, Harm, Expectations,
(1992) 217 Recurring and Similarity
Chicago Police Department (CPD) 257, (CHEERS) 120
258 Community Oriented Policing Services
Chilvers, M. 274 (COPS), Office of 1, 12, 350
Citistat (Baltimore) 277, 278, 281 community policing and 48, 62, 63
Citizen Law Enforcement Analysis and problem-oriented policing and 120, 121,
Reporting (CLEAR) program 246 128, 136
Citizens on Patrol 104 community policing 14, 67, 91, 225
Citywide Street Crime Unit (New York) promise of 27–43, 67
288 community involvement 28–34
civil libertarianism 209–212 decentralization 36–40
Clark, Police Commissioner Kevin P. problem solving 34–36
(Baltimore City) 107 prospects 41
Clarke, L. 65 organization and practice 47–55
class and race bias 255–256 outcomes 55–63
CLEAR (Citizen Law Enforcement public expectations 45–46
Analysis and Reporting) program terrorist-oriented 63–65
246 Compstat 2, 11–12, 17–18, 176, 195
Clear, T. 215 American policing and 284–301
Cleveland 40 internal contradictions in model
Clines, F. X. 274 292–297
Cloward, Richard 93, 226 para-military model of control
CMRC (Crime Mapping Research 289–292
Center) study 237 promise of 285–289
Cochran, J. K. 51 broken windows and 90, 92, 94, 102
Cochrane Collaboration 309, 310 community policing and 41
collaborative review groups (CRGs) crime rates 297
309, 310 evidence-based policing and 318, 336
CODEFOR (Computer Optimized future and 343, 346
Deployment-Focus on Results) hot spots and 225, 246
274 innovation of 267–283
Cohen, J. 124, 125 concluding thoughts 280–281
Cohen, L. E. 228, 250 full Compstat 269–271, 273–278
Coles, Catherine M. 15, 77, 79, 80, 88, maintaining 278–279
89, 103, 179, 286, 295 understanding 268, 269
Comey, J. 29 principles, practice and 147, 148
Committee to Review Research on Police problem-oriented policing and 122, 127
Policy and Practices 1, 5, 9, 11, 12, Computer Optimized Deployment-focus
214 on Results (CODEFOR) 274
356 Index

confidence, crisis of 4–11 DARE (Drugs Abuse Resistance

Congress 30, 64, 310 Education) Program 36, 122, 315,
Constitution (US) 210 332
Cook, T. D. 248, 307, 308, 312, 325 community policing and 49, 57, 58
coproduction 106–107 Davies, H. T. O. 18
see also incivilities reduction, zero Davies, S. L. 64
tolerance and coproduction Davis, E. M. 290
coproduction model 109–110 Davis, M. 51
COPS see Community Oriented Policing Davis, R. C. 107, 251
Services (COPS), Office of decentralization 36–40
Corbett, R. 179 DeJong, C. 67
Cordner, G. W. 13, 35, 105 Department of Justice see Justice
problem-oriented policing and 119, 133, Department
140, 146 deterrence 161–162, 249–250
Corman, H. 86 Dieckman, D. 6–13, 227, 328
Cornish, D. B. 228, 250 DiMaggio, P. J. 49
Costello, S. K. 254 disorder 55–58, 83–85, 102
costs displacement 252
hot spots policing 256 distribution of policing’s benefits 59–60
third party policing 214–218 district advisory committees (Chicago) 59
Crank, J. P. 46, 50 diversifying responses 125–126
crime Dixon, D. 253
criminology and criminal justice 91–93 DNA 2, 340
disorder; control and ; fear and 55–58 Dorchester Youth Collaborative 179
effects of hot spots policing 256–257 Douglas, W. O. 213
experts 251–253 Doyle, D. 208
mapping adoption 237 Drahos, P. 194
prevention 311, 314 Drugs Abuse Resistance Education see
Compstat and 295 DARE (Drugs Abuse Resistance
evidence-based policing and 305–319 Education)
situational 92 DuBois, J. 29, 32, 61, 246
rates Due Process 210, 211
Compstat and 297 Duffala, D. C. 228
index 349 Dussault, R. 281
see also violent crime Dvorak Keyboard 3
Crime and Disorder Act (1998) (UK) 200,
204 Earls, F. 31, 58, 216, 257
Crime and Justice (Tonry and Morris) Eck, J. E. 5, 9, 12, 156, 164
306 broken windows and 84, 105, 106
Crime Mapping Research Center community policing and 34, 36, 47, 50,
(CMRC) study 237 56
Criminal Justice Commission (Australia) Compstat and 295, 296, 297
232 innovation 275
Criminology and Public Policy 267 evidence-based policing and 316, 317
crisis research 311, 310–313, 314
change and 11–18 future and 340, 341, 342, 344
of confidence 4–11 crime, disorder and 343
Cullen, F. 18 homicide rates and 296
culture in community policing 50–52 hot spots and 225, 228, 229, 231, 234,
Curtis, R. 253 235
limits of 245, 246, 252, 258
Dabdoub, M. 229 policing strategies and 341
Dakof, G. A. 309 principles, practice and 44, 133, 134,
Dalton, E. 155, 157, 158, 171 145, 149
groups, networks and 162, 163 problem definition 134, 136, 137, 141
 Index 357

problem-oriented policing and 122, 123, Ferman, B. 110

124, 127 Ferracuti, F. 226
diversifying responses 125, 126 Ferrara, A. 79
evolution of 119, 120, 121, 122; Ferraro, K. F. 83
guidance improvement 121; Finn, P. 197
theoretical underpinning 120 Fitzgerald, B. 179
third party policing and 77, 191, 193, Florida 103, 279
197, 201 Fogelson, R. M. 44, 104
effective policing 124–126 Ford Foundation 5, 6
Ekblom, P. 309 Forrester, D. 123
Elmore, R. 339 Forst, B. 253
empirical evidence 231–234 FoxPro 271
enforcement, traditional 164–166 Frank, J. 67, 253, 318
England 136 Fregly, S. 229
Ericson, R. V. 64, 193, 195 Frend, A. 332
Erzin, T. 107
Esbenson, F. 332 Gaffigan, S. J. 48
evaluation Gaines, L. 253
evidence 202, 201–203 Gallagher, C. 62
hot spots policing 233 Gardiner, S. 279
evidence Garland, D. 107, 108, 193
-based models 225–229 Garofalo, J. 99
empirical 231–234 Gartin, P. P. 92, 135
evaluation 202, 201–203 Gartin, P. R. 17, 229, 246
scientific 323–327 Gendreau, P. 18
evidence-based policing 18 GIS (digital) mapping 248
for crime prevention 305–319 Giuliani, Mayor Rudolph W. 77, 80, 82,
model for 306–310; larger movement 95, 267
309–310 Glazer, N. 79
police use of model 316–317; reasons Goldsmith, S. 277
for 317–318 Goldstein, Herman 11, 52, 168, 347
research 310–315, 316; Sherman and broken windows and 92, 105
Eck 311, 310–312; towards Compstat and 17, 284, 290, 298
315–317 promise of 285, 286
improvement through expertise, evidence-based policing and 326, 331
experience and experiments hot spots and 227, 249, 251
322–337 principles, practice and 133, 134, 148,
police practices 327–332 149
policing improvement 335–336 problem definition 134, 135, 140,
practices working 333–334 145, 146
scientific evidence 323–327 problem-oriented policing and 16, 118,
119, 120
Fagan, J. 172, 176 adopting approach 123
Famega, C. 318 third party policing and 16, 191, 192,
Farrell, G. 125 201, 204
Farrington, D. P. 307, 308, 309, 312 Goodwin, B. 126
hot spots and 227, 251 Gootman, E. 268, 269
fear 55–58, 83–85 Gore, Vice President A. 284
Federal Bureau of Investigation (FBI) 6, Gorr, W. 124
348, 349 Gottfredson, D. C. 49, 59, 245, 310, 312
Compstat and 268, 281 Gottfredson, M. R. 9, 16, 226
Fedstat 281 Gottfredson, S. D. 99
Feeley, M. 193 Gould, J. B. 55
Felson, Marcus 191 governance and regulation 192–195
hot spots and 226, 228, 250 Grammich, C. 155
358 Index

Grasmick, H. G. 226, 250 Hartnett, S. M. 105, 109, 184, 246

Gray, A. P. 294 community policing and 47, 52, 61
Great Society 82, 92, 93 promise of 29, 32, 33, 34, 37
Green, L. Harvard University 278, 284
broken windows and 86, 102 Hassell, K. D. 48
hot spots and 229, 232 Hauritz, M. 125, 133
limits 252, 258 Hawkins, D. F. 254
principles, practice and 137, 141, 145 Hawkins, G. 166, 218
problem-oriented policing and 124, Hawkins, R. 218
125 Hawthorne effect 56
third party policing and 197, 198, 201 He, N. 51
see also Green-Mazerolle, L.; Mazerolle, Health Insurance (New York), Office of
L.G. 280
Green-Mazerolle, L. 234, 252, 258 Health and Safety codes 199
Greenberg, S. 286 Healthstat (New York) 280
Greene, J. A. 235 Heckman, J. 325
Greene, J. R. 13, 105, 251, 290 Heggarty, K. D. 195
community policing and 38, 47, 48, Helmick, D. O. 120
49 Henry, V. E. 274, 295
Greenspan, R. 11, 122, 135, 195, 318 Herman Goldstein Award for Excellence in
community policing and 47, 52, 61 Problem-Oriented Policing 134
Compstat and 284 Hernandez, J. 279
innovation 267, 268, 269 Hester, J. 120
future and 343 High Point (NC) 158
hot spots and 225, 236 high-crime places 146
Greenwood, P. J. 9, 227 Hill, R. 226
evidence-based policing and 327, 328 Hirschi, T. 9, 16, 94, 226
Greenwood, P. W. 318 Hoare, M. A. 119
Grinc, R. M. 32, 33, 49, 51, 339 Hockey, B. 294
Groff, E. 236 Hofstadter, D. R. 128
Grogger, J. 214 Holbrook, B. 253
Grossman, H. 217 Holsinger, A. M. 54, 86
groups and networks 162–164 Homel, R. 125, 133
Groves, W. B. 58, 226 Homeland Security, Department of 64
Grunwald, M. 181 homicide 175, 177, 296
Guerry, A. M. 228 Hope, T. 125, 232, 253
guidance improvement 121–122 hot spots policing 17, 92
Gurwitt, R. 284 limits of 245–259
abusive policing 254
Haar, R. N. 52 benefits 245
Hagan, J. 107 collective efficacy 259; rethinking 257
Haggerty, K. D. 64, 193 community and problem-oriented
Hallman, H. W. 110 theories 250–251
Hamilton, E. E. 50 costs 256
Hanson, G. S. 49 deterrence model 249–250
Harcourt, B. E. 122, 127 effects on crime 256–257
broken windows and 83, 84, 87 labeling stigma 254–255
incivilities 100, 102, 103 police as experts 251–253; adverse
Hardyman, P. 290 effects 252; displacement 252
harm principle, collapse of 103 police-community relations 253–254
Harries, K. 121 problem analysis 248–249
Harris Poll (1999) 46 problem definition 247–248
Hart, A. 279 race and class bias 255–256
Hartley, J. 327 response options 249
Hartman, J. L. 54, 86 short-term impact 246–247
 Index 359

as a model 231, 225–239 Kansas City 7, 227, 246

emergence of 229–231 Kansas City Crack House Police Raids
empirical evidence 231–234 Program 232
evaluations 233 Kansas City Gun Project 182, 232
evidence based 225–229 Kansas City Police Department 6–7
research 234–238 Kansas City Preventive Patrol Experiment
Hough, M. 228, 309 17, 19
Houston 33, 83 crisis of confidence and 6, 8
Houston Police Department 158 hot spots and 229, 230
Houston Targeted Beat Program 232 Kanter, R. M. 278
Hsia, H. M. 255 Karmen, A. 94
Hunter, A. 109, 110 Katz, C. M. 48, 84, 87, 88
Hunter, R. D. 133, 228 Compstat and 267, 274, 286, 295
Huo, Y. J. 57 crisis of confidence and 6, 7, 13
Hutchins v. District of Columbia 211 future and 342, 343, 345
Hylton, M. O. 197 innovation and 14, 15
Kelling, George L. 14, 127, 179, 227,
incivilities reduction, zero tolerance and 328
coproduction 98–111 broken windows and 88, 91, 92, 94,
increasing inequality 103 107, 108
Index crime rate 348 discourses 77, 84, 86, 155; ideas of
Indianapolis 67, 157, 158, 168 79, 80; moral issues 89
Informix 271 incivilities 99, 102, 103
innovation 2–11, 339–352 origins of 103, 105
Islamic citizens 64 Compstat and 30, 274, 286, 295
Ismaili, K. 107 crisis of confidence and 6, 215
IT capabilities of law enforcement 246 future and 342, 343, 345
innovation and 14, 15
Jacobs, J. 79 Kennedy, David M. 16, 50, 92, 318, 330
Jang, S. J. 83, 84, 85, 93 future and 79, 343, 346
Jeffery, R. 106 hot spots and 247, 253
Jeffrey, C. R. 133 problem-oriented policing 119, 120,
Jersey City Drug Markets Analysis 123, 125, 126
Program 232 pulling levers and 77, 155, 165, 166,
Jersey City (NJ) 86, 213 168
problem-oriented policing and 141, 146 Boston 171, 172, 174, 178, 181, 184;
problem definition 137, 139 network of capacity 179; Operation
Jersey City Problem-Oriented at Violent Ceasefire 173, 174
Places Study 232 groups and networks 162, 163, 164
Jill Dando Institute for Crime Sciences impact 158
121 police knowledge 159, 160, 161
John F. Kennedy School of Government Kelly, Clarence M. (Kansas City Police
267, 277 Chief) 6, 7
Johnson, B. R. 84, 85, 94, 315 Kenney, D. 126
Johnson, C. C. 47 Kerik, B. B. 52
Johnson, President Lyndon 81, 82 Kerlikowske, R. G. 64
Joint Terrorism Task Force 281 Kerner Commission on Civil Disorders
Jones, S. 280 (1968) 4, 104
Jordan, J. 179 Kerner, O. 104
Justice Department 5, 120, 350 Kessler, S. L. 219
pulling levers and 157, 171, 174, 176 King, W. R. 48
Kirchner Jr, R. E. 9
Kadleck, C. 102, 105, 106, 124, 197 Klarman, M. J. 214
Kahneman, D. 250 Kleiman, M. 253
Kane, R. J. 255 Klein, M. W. 250, 256
360 Index

Klinger, D. A. 11 Maas, P. 284

Klockars, C. B. 63 McArdle, A. 107
Knobler, P. 18 McCluskey, J. D. 55, 56
Koper, C. S. 48, 145, 246 McCord, J 307
McDevitt, J. 164
La Grange, R. L. 83 MacDonald, H. 281
Lab, S. P. 227, 308 McDonald, P. P. 284, 286
labeling stigma 254–255 McElroy, J. E. 13, 290
LaFree, G. 4, 44 McEwen, J. T. 135
Larson, R. C. 7 McEwen, T. 236
Latin Kings (gang) 163 McGarrell, E. F. 125
Latinos 31, 33, 60, 255, 344 pulling levers and 155, 158, 161, 162,
Laub, J. H. 99, 226 178
Lauritsen, J. 226 McIlwain, G. 125, 133
LaVigne, N. 121, 236, 246 McKay, H. D. 215, 226
Lavrakas, P. J. 248 MacKenzie, D. L. 18, 245
Law Enforcement Assistance evidence-based policing and 309, 310,
Administration 92 312
Law Enforcement News 276, 284 McLaughlin, E. J. 49
Lawton, B. 101 McLaughlin, Paul 181
LeBeau, J. 228 McNally, A. M. 11, 39, 122, 195, 318,
legal issues 199–200 343
legislation Compstat and 267, 268, 284
Crime and Disorder Act (1998) (UK) hot spots and 225, 236
200, 204 Madison 119
Omnibus Crime Control and Safe Maguire, E. R. 84, 343, 350
Streets Act (1968) 5 community policing 62, 63
Violent Crime Control and Law partnerships and 48
Enforcement Act (1994) 1, 40 reform 48
Levine, J. P. 8, 9, 227 Compstat and 296, 297
Levitt, L. 279 innovation 275
Levitt, S. 176 crime prevention and 312–313
Liddle, H. A. 309 homicide rates and 296
Lind, E. A. 218 Maher, L. 17, 92, 229, 253
Link Valley operation (Houston PD) Mallia, J. 182
158 Mamalian, C. 236, 246
Lipsey, M. W. 227, 319 Mann, K. 17
Liquor Licensing Authority (Australia) 197 Manning, P. K. 46, 51, 279
Loader, I. 195 Maple, Deputy Commissioner (New York)
Loftin, C. 226 Jack
Logan, J. R. 109 Compstat and 286, 287, 288
London 119 innovation 269, 270, 275, 277,
Los Angeles 158 281
Lovell, R. 255 mapping (GIS) 248
Lovrich, N. P. 51 Maquoketo v. Russell, City of 211
Lowell (Massachusetts) 164 Martinson, R. 227
Lowell Police Department 269 Maryland 119
Luckenbill, D. 208 Massachusetts 164, 176
Lum, C. M. 101, 307 Mastrofski, S. D. 11, 13, 105, 195, 318
crime mapping and 237 community policing and 44, 46, 48, 67
hot spots and 229, 236, 238 organization and practice 50, 53, 54,
Lurigio, A. J. 34, 251 55; philosophy and culture 50, 51,
Lynch, J. P. 216 52
Lyons, W. 46, 54, 64 outcomes 61, 62, 63
promise of 39, 40
 Index 361

Compstat and 284 hot spots policing 225–239

innovation 267–273 para-military of control 289–292
para-military model of control 290 SARA 140
future and 343, 346 Moffitt, T. 226
hotspots and 225, 236 Molotch, H. 109
police-community partnerships and 48 Monkkonen, E. H. 107
problem-oriented policing 122, 135, Moon, Y. 50
147 Moore, M. H. 14, 178, 273
Matthews, R. 125, 133 community policing and 50, 51, 52
Maxfield, M. G. 83 evidence-based policing 328, 332, 334
Maxwell, C. D. 51 future and 339, 340, 344, 348, 350
Mayhew, P. 228 moral issues 88–90
Mazerolle, L. G. 17, 19, 36, 232, 318 Morgan, S. 324
broken windows and 86, 102, 105, Morningstar Baptist Church 181
106 Mosteller, F. 309
future and 343, 345 MOVE (Department of Transportation)
problem-oriented policing 124, 125, 280
126, 135, 137 Moy, P. 30
third party policing 191, 203 Muir Jr, W. K. 51
critical view 207, 208, 213, 214 Mulvihill, M. 182
dimensions 196, 197, 198, 200, 201 Murphy, P. V. 6, 7
‘means-over-ends-syndrome’ 118
Meares, T. L. 204, 216, 344, 345 Nagin, D. 215
Melnicoe, W. B. 290 Najaka, S. S. 50
Menig, J. 290 Narr, T. 329
Merton, R. K. 226, 294 National Academies 55, 56
Meyer, J. W. 49, 225 Panel on Improving Information and
Miami 281 Data on Firearms 176, 178
Michigan 210–211 National Academy of Sciences 232, 234,
Michigan State University 158 313
Microsoft 271 Panel on Policing 328
Mid-Town Manhattan Community Court National Crime Prevention Council 197
77 National Crime Victimization Surveys 98
Millenson, M. L. 18, 309, 317 National Institute of Justice (NIJ) 5, 8,
Minneapolis 137, 229, 246, 274 171, 238, 268
pulling levers and 158, 168, 171, 178, Crime Mapping Laboratory 236
184 National Research Council (NRC) 45
Minneapolis Hot Spots Experiment 17, see also Committee to Review Research
230, 231, 231, 236, 238 on Police Policy and Practice
Minneapolis Hot Spots Patrol Program Neighborhood Watch 47, 58, 64, 104
232 neighborhoods 157, 171, 179
Minneapolis Medical Research Foundation Nettler, G. 226
7 network of capacity development
Minneapolis Police Department 229 178–180
Minneapolis Repeat Call Address Policing New Haven 163
(RECAP) 232 New Orleans 279
mission clarification 287 New South Wales 274
Mission Hill housing projects 180 New York 158, 176, 246, 255
Missouri 227 broken windows and 89, 90
Mocan, N. 86 order maintenance 94, 95
models research 85, 86
Compstat 269, 292–297 Compstat and 286, 287, 288
deterrence 249–250 crime prevention 295, 296, 297
evidence-based policing 225–229, innovation 271, 279, 280; full 269,
306–310, 316–317 270
362 Index

New York City 33, 41, 148, 168, 276, 343 O’Shea, T. C. 58
broken windows and 79, 80, 82, 88, 89, Ostrom, E. 11, 106, 110
94, 102 outcomes of policing 55–63
Compstat and 11, 17, 284, 286, 288 building social capital 58–59
Corrections Department 276 crime, disorder and fear 55–58
Department of Parks and Recreation distribution of benefits 59–60
276 enhanced legitimacy 62–63
homicide rates 296 success 61–62
hot spots and 235, 255
New York City Police Department Paoline III, E. A. 51
(NYPD) Papachristos, A. V. 247
broken windows and 85, 86, 89, 90, 92, Papachristou v. City of Jacksonville 213
94 paramilitary model of control 289–292
Compstat and 269, 279, 287, 292 Parascandola, R. 279
organization 272, 273 Parks Inspection Program (PIP) (New
origins 270, 271 York) 276
selling 274, 275, 276, 277 Parks, R. B. 11, 106
understanding 268 police practice and 50, 51, 55, 58, 67
Newark Foot Patrol Experiment 14 Parkstat (New York) 276, 277
Newark (NJ) 83, 84, 93, 99 partnership, accountability and innovation
Newport News 119, 134 171–187
Newport News Police Department 137, Pate, A. M. 6, 33, 79, 227, 328, 344
149 Paternoster, R. 250, 251
Nicholls, K. 58 Pease, K. 123, 125
Nisbett, R. E. 250 Pelfrey, W. V. 105
North Carolina 158 Pepper, J. 171, 343
Novak, K. J. 54, 86, 88 Percy, S. L. 106
Numez v. City of San Diego 211 Perkins, D. D. 101
Nutley, S. 18 Peters, T. 272
Petersilia, J. 9, 227, 251, 327, 328
Oakland 86, 106, 137 Petersilia, M. 9
O’Boye, S. 279 Peterson, D. 332
O’Connell, P. E. 272, 276, 277 Petrie, C. 171, 343
O’Malley, P. 193 Petrosino, A. 307
Omnibus Crime Control and Safe Streets Pfau, M. 30
Act (1968) 5 Philadelphia 83, 274, 279
One Police Plaza 271 philosophy, and culture in community
O’Neill, A. W. 279 policing 50–52
Operation Ceasefire (Boston) 149, 168, Piehl, A. M. 16, 119, 120, 247, 330
172–174, 179, 180 pulling levers and 155, 161, 165, 173,
accountability and 180, 182, 183 174
implementation 182, 183 Pierce, Glenn L. 17, 92, 164, 229
other jurisdictions 183, 184, 185 Pindyck, R. S. 327
working group 174, 183 Pittsburgh 281
Operation and Crime Review (New South Plotkin, M. 329
Wales) 274 PNAP (Pharmacy Narcotics Abuse
Operation ID 104 Prevention) 51
Operation Nightlight 179 police athletic leagues 57
Operation Whistlestop 104 police as crime experts 251–253
Oregon 158 Police Development Fund 5
organization and practice in community Police Executive Research Forum 134, 136
policing 246 Police Foundation 5–7, 27, 91, 135, 227,
Compstat 271 236, 305
philosophy and culture 252 Compstat 284, 289, 290, 292, 295, 297
programs and structures 246 innovation 268, 269, 270
street level 52–55 studies 8, 12, 14, 15
 Index 363

police innovation, and the future 339–352 public expectations 45–46

categorizing innovations 340–342 pulling levers policing 16, 92
community reaction 344–346 lessons of 155–168
crime and disorder control 342–344 communication, relationship and
police reaction 346–347 common ground 166–167
police, practices 327–332 concluding issues 167–168
police-community relations 253–254 definition 156–157
policing strategies 341 deterrence 161–162
Pooley, E. 269 evidence of impact 157–159
Pope, C. 255 groups and networks 162–164
Portland 158 knowledge 159–161
Potter, G. 253 traditional enforcement 164–166
Powell, W. W. 49 partnership, accountability and
practices, police 327–332 innovation 171–187
Presidency 30 Boston Gun Project and Operation
Presidential Commission on Law Ceasefire 173–174
Enforcement and the implications 183–185
Administration of Justice 81, 82, network of capacity 178–180
83, 92 police-community relations 180–183
Press, S. J. 9 serious violence 174–178
principles, and practice in Punch, M. 284, 290
problem-oriented policing 133–152 Purcell, C. M. 119
problem
analysis triangle 120 Queensland 197
definition 120–121, 247–249 Quetelet, A. J. 228
identification and assessment 288 Quille, T. J. 309
solving 34–36, 91, 145–148
tactics 288–289 race
Problem Solving Partnerships 136 class bias and 255–256
problem-oriented policing 15–16, 152 equity 212–214
principles, practice and 133–152 Raine, A. 226
definition 134–144; analysis Ransley, Janet 343, 345
136–140; assessment 141–144; third party policing and 191, 201, 203,
responses 140–141; scanning 207, 208
134–136 Raudenbush, S. W. 31, 58, 216, 257
in-depth problem solving 147–148 broken windows and 87, 93, 100
shallow problem solving 145–147 research 84, 85
science, values and 117–128 Read, T. 120, 133, 134, 136, 140
adopting approach 122–126; science Ready, J. 105, 213, 235
and 123–124; values and 122–123 Reagan, Ronald 193
alternatives 126–128 RECAP (Minneapolis Repeat Call Address
definition 117–119 Policing) 232
evolution 119–122; guidance Reinier, G. H. 236
improvement 121–122; location, Reisig, M. D. 46, 58, 62
scale and quality 119–120; problem Reiss Jr., A. J. 10, 46, 66, 67
definition 120–121; theoretical Remnick, D. 284
underpinnings 120 Rengert, G. F. 228
theories 250–251 research 45
process-oriented policing 54–55 broken windows 83–87
Professional Security 127 disorder-fear and disorder-crime 83–85
programs and structures in community evidence-based policing 310–316
policing 47–50 hot spots policing 234–238
Project on Policing Neighborhoods, The policy evaluation 85–87
67 see also Committee to Review Research
Project Safe Neighborhoods 157, 171 on Police Policy and Practice
Prusoff, L. 9 responses 125–126, 140–141, 249
364 Index

Revere, E. 213 scanning 134–136

Rice, K. 120 Schaefer, D. R. 84
Rich, T. F. 49 Schafer, J. A. 51, 52
RICO, prosecution 163 Schnelle, J. F. 9
Ridgeway, G. 155 Schuck, A. M. 252
Riley, K. J. 155, 318 science, values and problem-oriented
Ritea, S. 279 policing 117–128
Ritti, R. R. 44, 46, 48, 50, 61, 225 scientific evidence 323–327
Roberts, D. 212, 213, 214 scientific methods scale (SMS) 312
Robinson, G. 277, 280 Scott, M. S. 121, 133, 136, 144, 147, 192
Robinson, J. 101 Scott, W. R. 49, 63
Rochester 158 Seattle 54
Roehl, J. 47, 102, 105, 106, 339 Sechrest, L. B. 227
hot spots and 234, 252, 258 Second World War 1, 5, 6, 13
problem-oriented policing 124, 134, security surveys 104
137, 140 Selvin, H. C. 94
third party policing and 196, 197, 198, Selvin, H. G. 294
200, 214 Shadish, W. R. 248, 307
Rogan, D. P. 182, 318 Shaw, C. R. 215, 226
hot spots and 232, 234, 246, 252, 253 Shaw, J. 145, 235
problem-oriented policing and 124, 125, Shearing, C. 195
145 Sherman, D. 329, 330, 331, 343
Rogers, Everett 3 Sherman, Lawrence W. 7, 92, 182, 201,
Roman, C. 160 203, 293, 295
Rose, D. 215 community policing and 55, 56, 59, 63
Rosenfeld, R. 226 evidence-based policing and 18, 311,
Ross, C. E. 83 306–312, 313, 316
Ross, L. 250 improvement 323
Rossmo, D. K. 329 model 307, 308, 309, 316, 317
Roth, J. A. 10, 47, 49, 50, 63 hot spots and 17, 225, 231, 239
partnerships and 48 emergence 229, 230, 231
Rouse, A. 51 limits 245, 246, 252, 253
Rowan, B. 49, 225 model 226, 229
Rowe, C. L. 309 research 232, 234
Rubenfeld, D. 327 problem-oriented policing and 124, 125,
Ryan, J. F. 48 135, 145
Shin, Y. 48
Sabol, W. J. 216 Shtull, P. 33
Sadd, S. 33, 49, 51, 253, 339 Shumaker, S. A. 99
Safe Neighborhoods Initiative 179 Silverman, E. B. 102, 343
St. Clair Commission 180 Compstat and 272, 279, 280, 284, 286,
St. Louis Problem-Oriented Policing at 288
Drug Market Locations Study 232 Simon, J. 193
Sampson, Rana J. 121, 215, 216 Simons, R. 92, 294
broken windows and 84, 85, 87, 93 Singh, P. 124
incivilities 100 situational crime prevention 92
community policing and 31, 58 future and 342, 344, 345, 346, 347
hot spots and 226, 250, 257–259 hot spots and 246, 248, 251, 253, 254
San Diego 137, 211 Skogan, Wesley G. 9, 12, 148, 184, 336
San Diego Police Department 158 broken windows and 83, 84, 87, 100,
San Francisco 83, 158, 161, 184 105, 109, 284
sanctions and penalties 200 community policing and 27, 37, 47, 61,
Sandburg, C. 45 66
SARA (Scanning, Analysis, Response and involvement 29, 30, 31, 32, 33, 34
Assessment) 105, 121, 134, 140 outcomes 52, 53, 54, 58, 59, 60, 61
 Index 365

future and 342, 344, 345, 346, 347 Suffolk County District Attorney’s Office
hot spots and 246, 248, 251, 253, 254 179
Skolnick, J. H. 147, 290 ‘Summer of Opportunity’ program 179
Slovic, P. 250 Supancic, M. 83
SMART (Specialized Multi-Agency Supina, A. 53
Response Team) 86 Supreme Court 210, 213
Smart Ware (software) 271 confidence in 30
Smith, M. 196, 253 Susmilch, C. E. 119, 123, 133
Smith, P. 18 Sutherland, E. 226
SMS (scientific methods scale) 312 Sviridoff, M. 253
Snipes, J. B. 44, 50, 51, 53, 54
social capital, building 58–59 targets 198–199
Solè, R. 126 Taylor, R. B. 141, 295, 342, 343, 345
Solomon, A. L. 256 broken windows and 83, 85, 88, 105,
Sousa, William H. 127, 267, 274, 295 109
broken windows and 80, 86, 89, 91, incivilities 99, 100, 101, 102
102, 108 hot spots and 228, 229, 246, 248
New York 94 Taylor, T. 332
future and 342, 343, 345 TEAMS (Total Efficiency Accountability
Spaar, S. A. 17, 92, 229 Management System) 276
Sparrow, M. K. 50, 52, 194, 332, 339, Ten Point Coalition 180, 181–183, 185
340 Tench, M. 182
Specialized Multi-Agency Response Team Terrill, W. 50, 51, 53, 105
(SMART) 86 terrorist-oriented policing 63–65
Spelman, W. 191, 201, 227, 249, 328, 332, Texas 33
340 Thacher, D. 88, 89, 90, 101
broken windows and 105, 106 Thatcher, Margaret 193
principles, practice and 133, 134, 137, theoretical underpinnings of
149 problem-oriented policing 120
problem-oriented policing 119, 120, third party policing 16–17, 191–206, 219
121, 123 critical view 207–219
rapid response and 7, 8 civil libertarianism 209–212
Spiegelman (author) 19 costs and benefits 214–218
statutes, Chicago (1992) 212 racial equity 212–214
Steiner, L. 246, 344 dimensions of 196–201
community policing and 30, 31, 32, 53 focal point 196–197
Stenning, P. 195 implementation 201
Stephens, D. 52 initiators 196
Stewart, G. 80, 119 legal basis 199–200
Stewart, James K. 238 problems 197
Stockton (Ca) 158 purpose of action 196
Strategic Alternatives to Community sanctions and penalties 200
Safety Initiative 171 targets 198
Strategic Approaches to Community tools and techniques 200–201
Safety Initiative 157 evaluation evidence 202, 201–203
Straub, F. 276 governance, regulation and 192–195
street level community policing 52–55 Tien, J. M. 49
engaging the community 53–54 Tilley, N. 120, 133, 134, 136, 140, 309
process-oriented 54–55 Tita, G. 155, 158, 162, 168, 171, 318
providing services 54 Tittle, C. R. 250, 251
structures in community policing 47–50 Tolbert, P. S. 49, 61
Stuart, Carol 180, 181, 183 Tonry, M. H. 103, 251
Stuart, Charles 180 Total Efficiency Accountability
Sturman, A. 228 Management System (TEAMS)
subject creation 103 276
366 Index

Transportation, Department of 280 computerized crime mapping and 237

Trasler, G. 226 crime, disorder and 342, 343
Travis, J. 160, 171–187, 256 crisis, change and 5, 7, 11, 12
Trojanowicz, R. 13, 105 evidence-based policing and 307,
Tuch, S. A. 107, 253 312–313, 314, 316, 317, 318
Turner, M. G. 54, 86 future and 344, 345, 346, 347
Tversky, A. 250 hot spots and 17, 225, 229, 230, 231,
Tyler, T. R. 218, 235, 253 231, 232
citizen support and 344, 345 limits 245, 246, 252, 258
community policing and 55, 56, 57 model 227, 228, 229
research 234, 235, 236, 237, 238
Uchida, C. 253 police innovation and 13, 17, 18, 340,
Unified Crime Reports (FBI) 268, 348 341
Uniform Building Standards 199 policing strategies and 341
Uselton Jr., P. H. 9 principles, practice and 133, 135,
Utne, M. 79 137
problem solving 145, 147, 148
Valley Alcohol Management Partnership problem-oriented policing 122, 124,
197 127
Vietnam War 4 effectiveness 125, 126
violent crime 10, 174–178 Weisel, D. L. 47, 121
places 138, 142 Weiss, A. 11, 46, 125, 236
Violent Crime Control and Law Weissenstein, M. 274
Enforcement Act (1994) 1, 40 Weitzer, R. 107, 253
Vira, B. 106 Wellford, C. 171, 176, 178, 343
Virginia 119, 134 Welsh, Brandon C. 307, 308, 309, 312,
Visher, C. 18, 145, 227 317, 343
Vollmer, August 347 Whitaker, G. P. 11, 106
White House 40
Wakeling, S. 155, 158, 161, 162 White, R. 198
Wales 136 White, S. O. 227
Walker, S. 347 Whitehead, J. T. 227
Wallman, J. 94, 102, 312, 348 Wilkinson, D. L. 49, 51, 52
War on Poverty 93 Williams, R. 218
Waring, E. J. 36, 86, 105, 120, 125, 137 Willis, J. J. 147, 148, 225, 343
hot spots and 232, 247 Compstat and 284, 291, 292
pulling levers and 158, 171 innovation 269, 278, 279
Wartell, J. 36, 197, 234, 258 Wilson, D. B. 49, 312, 319
problem-oriented policing and 124, 125, Wilson, James Q. 14, 15, 45, 125, 227,
141 251, 295, 336
Washington, DC 30, 160, 165, 211 broken windows and 77, 79, 80, 82, 84,
Waterman, R. H. 272, 273 89, 92, 105
Watson, T. S. 126 incivilities 99, 102
Waul, M. 256 Wilson, O. W. 347
Weatherburn, D. 274 Wilson, W. J. 215
Webb, V. J. 84 Winship, C. 324, 343, 345, 346
Webber, R. 277, 280 pulling levers and 157, 174, 185
Weisburd, David L. 156, 162, 164, 193, accountability 180, 181, 182, 183
195 Winston-Salem (NC) 158
broken windows and 86, 92, 94, 101 Wisconsin 119
community policing and 36, 39, 50, 51, Wolfgang, M. E. 226
56 Wood, R. L. 51
Compstat and 284, 286, 289, 290 Worden, R. E. 44, 50, 51, 67, 253
innovation 267, 268–269, 270 Worrall, J. L. 63, 86
model 293, 295 Wortley, R. 125, 133
 Index 367

Wright, R. 226 zero tolerance policing 1, 89, 94, 312, 345

Wyckoff, L. A. 235 see also broken windows policing;
Wycoff, M. A. 62, 344 incivilities reduction, zero tolerance
and coproduction
Yang, S. M. 101, 229 Zhao, J. 48, 51, 63
youth homicide, in Boston 175 Zimring, F. E. 166, 218
Youth Violence Strike Force (YVSF) 173, Zucker, L. G. 49, 61
179, 181, 182 Zuger, A. 18