Community policing

Material Information

Community policing how top executives view the concept and what they are doing to implement it
Morse, John P
Publication Date:
Physical Description:
247 leaves : ; 28 cm


Subjects / Keywords:
Community policing -- United States ( lcsh )
Police -- Attitudes -- United States ( lcsh )
Police chiefs -- Attitudes -- United States ( lcsh )
Community policing ( fast )
Police -- Attitudes ( fast )
Police chiefs -- Attitudes ( fast )
United States ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 239-247).
General Note:
School of Public Affairs
Statement of Responsibility:
by John P. Morse.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
47848511 ( OCLC )
LD1190.P86 2001d .M67 ( lcc )

Full Text
John P. Morse
B.S.B.A., University of Colorado, 1980
M.B.A., Regis University, 1984
M.P.A., University of Colorado, 1996
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Affairs

2001 by John P. Morse
All rights reserved.

This thesis for the Doctor of Philosophy
degree by
John P. Morse
has been approved

[Z Z00f
' Date
Fred W. Rainguet

Morse, John P. (Ph.D., Public Affairs)
Community Policing: How Top Executives View the Concept and What They are
Doing to Implement It
Thesis directed by Professor Richard J. Stillman, II
The purpose of this research is to explore the effects of community policing on
police departments policies, structure, and officer responsibilities, and to examine
the departments top executives attitudes concerning the underlying tenets of
community policing. The study also sought to construct and test a model to explain
the difference between police organizations that implement community policing
through specialized units (minimal implementers) and those that implemented it
through some restructuring (maximal implementers). The study was accomplished
in two parts; a written survey and follow-up telephone interviews. Survey data were
collected from a random sample of top police executives from local law enforcement
agencies employing at least fifty officers. Of the 557 executives surveyed, 388
(69.7%) responded. Thirteen of those that responded to the written survey
participated in thirty-minute follow-up telephone interviews.
The hypotheses were that (1) there would not be much variation between
departments in what was implemented as community policing; (2) there would not
be an impact on the policies, structure, or responsibilities of patrol officers due to
community policing; (3) there would be no difference in support levels for
community policing between citizens and police employees; (4) there would not be
organizational problems associated with implementing community policing (5)
police executives would not differ in their opinions concerning citizen participation,
citizen oversight, command devolution, and problem solving; and (6) there would
not be external factors helping to predict success with community policing.

The study found there was variation in what was being implemented as community
policing, there are some policy and responsibility changes, there are differences in
support, most of the organizational problems associated with implementation are
behind us, executives do vary in their views of community policings underlying
tenets, and the proposed model explains some of the variation between minimal and
maximal implementers. However, the follow-up telephone interviews revealed that
many departments are holding on to their professional model roots of random patrol,
rapid response, and retrospective investigations and may be just bolting
community policing onto their existing operations instead of implementing
fundamental changes to their operations that are consistent with the philosophy of
community policing.

I dedicate this dissertation to mother, Mary H. Morse, whose love, wisdom, and
guidance are always with me.

I know nothing except the fact of my ignorance.
As soon as you begin to acknowledge those that helped you risk leaving out
someone that was also very important. I know that will happen here, but it is a risk I
must take.
I am grateful to everyone who contributed to my education both inside the
classroom and out. I know the wisdom I have gained has been more attributable to
your ability to teach than to my ability to learn. This includes all the police chiefs
and sheriffs that took the time to complete the survey and especially those that
agreed to telephone interviews.
There is an impression as you contemplate beginning a dissertation that it is a
solitary process. As you begin the process you discover that although many hours
are spent in solitude, the assistance of others is the most critical element to
differentiating between a completed dissertation and one that ends up endlessly in
The professors in the Graduate School of Public Affairs deserve special
mention. Richard Box encouraged me to enter the program and once I was in
challenged me to make the most of it. Peter deLeon refused to take no for an answer
during my darkest hours. Richard Stillman stood with me shoulder to shoulder and
believed in me when I needed it most. Mary Dodge, Jae Moon, and Fred Rainguet
graciously agreed to serve on my committee and then offered much needed support
all along the way.
Finally, I would like to acknowledge my mother. Obviously, she has stuck with
me through thick and thin, but during this dissertation she gave very freely of her
time with no notice to attend to proofreading chores. Without that commitment, my
sanity would have come even closer to leaving me during this process.
Thank you all.

Figures......................................................... xv
Tables........................................................ xvi
1. INTRODUCTION.....................................................1
Purpose of the Study.......................................3
Organization of the Dissertation...........................5
2. LITERATURE REVIEW............................................. 7
The Origins of Policing in the United States.............7
The Professional Model and Its Limitations...............11
What is Serious Crime?..............................17
Missed Opportunities to Prevent Crime...............20
Is the Problem with the Professional Model Police Management?... 22
Crime Prevention Under the Professional Model.............24
An Introduction to Community Policing.....................25
Defining Community Policing...............................28
Problems with the Broken Windows Thesis...........31

Problems with Other Aspects of Community Policing
Community Policing in Practice............................... 38
Community Policing in New York City...................41
Summary of the Practical Issues in Community Policing.47
An Aperture in the Literature................................53
Research Questions and Hypotheses...........................58
Research Question #1............................... 59
Null Hypotheses #1...................................59
Rationale for Hypotheses #1..........................59
Research Question #2.................................59
Null Hypotheses #2...................................60
Rationale for Hypotheses #2..........................60
Research Question #3.................................60
Null Hypotheses #3...................................60
Rationale for Hypotheses #3..........................60
Research Question #4.................................60
Null Hypotheses #4...................................61
Rationale for Hypotheses #4..........................61
Research Question #5.................................61

Null Hypotheses #5.....................................61
Rationale for Hypotheses #5............................61
Research Question #6............................... 62
Null Hypotheses #6.....................................62
Rationale for Hypotheses #6............................62
Quantitative Research Design.................................63
Sample Characteristics and Sampling Strategy.................70
Pilot Study and Instrument Development.......................72
Reliability Testing....................................73
Validity Testing.......................................75
Qualitative Research Design..................................77
Study Limitations............................................81
4. RESEARCH RESULTS WRITTEN SURVEY................................ 84
Sample Demographics and Characteristics......................86
Research Question #1: What is Community Policing?...........96
Research Question # 2: Community Policings Impact on
Policies, Structures and Officer Responsibilities............108
Research Question # 3: Support for Community Policing.......118

Research Question # 4: Operational Problems Associated with the
Implementation of Community Policing.......................122
Research Question # 5: Police Executive Opinions Concerning
Citizen Participation, Command Devolution, and Problem
Research Question # 6: Factors that Help Predict Successful
Implementation of Community Policing..................... 134
Some Exploratory Models....................................140
5. RESEARCH RESULTS TELEPHONE INTERVIEWS..........................147
Community Policing Priorities Facing Police Executives.....147
Philosophical Concerns..............................148
Operational Concerns................................150
Issues Facing the Concept of Community Policing............157
The Lack of a Definition............................157
Police Support for Community Policing...............161
Funding Community Policing Efforts..................165
Keeping up the Momentum.............................166
Reasons for Recent Changes to Departments Policies to
Incorporate Community Policing.............................167
Citizens Have Pushed for Change.....................167
Reaching Critical Mass............................169

Change Leaders and Officers and You Can Change
Policies and Culture..................................170
Complying with Grant Requirements.....................172
Future Changes to Community Policing.........................172
More Decentralization and Autonomy....................173
Handling Calls for Service More Effectively...........173
Carving Out Time from Patrol for Community Policing..174
Should Policies Be Changed to Implement Community Policing?. 175
Opposition to Policy Changes to Implement Community
Policies Must Be Changed to Incorporate
Community Policing................................... 177
Reasons To Be Optimistic about Community Policings Future..179
Community Support.................................... 179
Perception of Efficacy................................181
Available Funding.....................................183
Moving from the Professional Model to a Community Model.....184
Community Policing as a Supplement to the
Professional Model....................................185
The Professional Model Should be Replaced.............185
The Professional Model is the Future..................187
Implementing Community Policing..............................187

Specialized Unit Approach...........................187
Department-Wide Approach............................189
Contradictions Uncovered in the Real World.................190
6. FINDINGS AND CONCLUSIONS.........................................192
Quantitative Findings......................................192
What is Community Policing?.........................195
Community Policings Impact on Policies, Structures,
and Officer Responsibilities........................196
Support for Community Policing......................198
Operational Problems Associated with the Implementation
of Community Policing...............................198
Police Executive Opinions Concerning Citizen Participation,
Command Devolution, and Problem Solving.............199
Factors that Help Predict Successful Implementation of
Community Policing..................................200
Qualitative Results........................................201
Is Community Policing a Philosophy or a Program?....201
Is the Community Involved in Community Policing?...205
Are Departments Solving Problems or are they Just
Responding to Calls?................................206
Are Departments Adopting Fixed Beat Assignments?...207
Are Police Departments Empowering Their Officers to
Solve Problems?.....................................207

Are Police Departments Providing Their Officers
Community Policing Training?...........208
Recommendations for Future Research.........209
A. SURVEY INSTRUMENT.............................211
RESPONDENTS................................. 224
E. SUPPLEMENTAL TABLES...........................226
REFERENCE LIST........................................239

4.1 Police Executive Age.......................................... 90
4.2 Years of Experience in Current Position....................... 93
4.3 Top ExecutivesLevel of Education............................. 93
4.4 Top Executive Ethnicity....................................... 94
4.5 Top Six Components of Community Policing...................... 98
4.6 Components with Significant Differences between Police and
Sheriffs Departments......................................... 99
4.7 Components with Significant Differences between Large and Small
Departments.................................................. 100
4.8 Components with Significant Differences between Minimal and
Maximal Implementers........................................... 101
4.9 Components with Significant Differences between Accredited and
Non-Accredited Departments..................................... 102
4.10 Six Largest Changes in Community Policing Components from
Gianakis and Davis to Current Study............................ 104
4.11 Differences in Officer Responsibilities between Police and Sheriffs
Departments.................................................... 115
4.12 Top Five Officer Duties for Maximal Implementers............... 117
4.13 Comparison of Factors of Support for Community Policing........ 121
4.14 Continuum of Executive Support for Tenets Underlying Community
Policing by Implementer Type................................... 133

2.1 Professional Model Philosophy and Practices Discussed in the
Literature...................................................... 15
2.2 Community Policing Philosophy and Practices for New York City.. 48
2.3 Community Policing Philosophy and Practices Discussed in the
Literature....................................................... 50
3.1 Research Questions and Variables................................. 65
3.2 Research Questions and Associated Survey Items................... 66
3.3 Research Questions and Proposed Statistical Analysis............. 67
3.4 Cronbachs Coefficient Alpha Results for Pilot Survey............ 75
3.5 Summary of Telephone Interview Participant Characteristics... 78
4.1 Summary of Population and Selected Sample Characteristics.... 86
4.2 Summary of Selected Sample and Sample Respondents
Characteristics............................................... 87
4.3 Average Age of Police Executives by Department Type.............. 89
4.4 Average Years of Overall Law Enforcement Experience of Top
Executive by Department Type..................................... 90
4.5 Average Years of Experience Prior to Becoming Top Executive by
Department Type...................................-.......... 91
4.6 Average Years of Top Executive Experience with Current
Department by Department Type.................................... 92

4.7 Average Total Years of Top Executive Experience by Department
Type....................................................... 92
4.8 Summary of Significant Demographic Differences between Police
Chiefs and Sheriffs......................................... 95
4.9 Components of Community Policing by Agency Type............. 97
4.10 Goals of Community Policing Implementation................... 106
4.11 Policy Changes to Implement Community Policing by Agency
Type........................................................ 109
4.12 Analysis of Change in Departmental Structure Due to
Implementation of Community Policing......................... 112
4.13 Analysis of Covariance for Effects of Department Size on Number
of Management Levels......................................... 113
4.14 Descriptive Statistics and Multiple Comparison Procedure for
Number of Management Levels by Community Policing
Implementation.............................................. 114
4.15 Patrol Officer Duties in Departments Implementing Community
Policing by Agency Type...................................... 116
4.16 Summary of Research Question #2.............................. 118
4.17 Analysis of Construction of Sub-Scales for Support for Community
Policing..................................................... 119
4.18 Levels of Support for Community Policing by Agency Type.... 120
4.19 Operational Problems Associated with Implementing Community
Policing by Agency Type...................................... 123
4.20 Comparison of Operational Problems Associated with
Implementing Community Policing between Gianakis and Davis
and Current Research......................................... 124

4.21 Citizen Participation Sub-Scale Construction and Reliability
Analysis................................................... 125
4.22 Citizen Oversight Sub-Scale Construction and Reliability Analysis.. 126
4.23 Empowerment Sub-Scale Construction and Reliability Analysis. 127
4.24 Problem Solving Sub-Scale Construction and Reliability Analysis .. 128
4.25 Top Executives Opinions Concerning Underlying Tenets of
Community Policing All Departments.......................... 129
4.26 Top Executives Opinions Concerning Tenets Underlying
Community Policing Departments That Have Not Implemented
Community Policing............................................ 131
4.27 Top Executives Opinions Concerning Tenets Underlying
Community Policing Departments That Have Implemented
Community Policing......................................... 132
4.28 Top Executives Evaluation of Their Departments Success with
Community Policing by Department Type......................... 135
4.29 Summary of Factors That Do Not Help Predict Success in
Community Policing.......................................... 136
4.30 Summary of Factors That May Help Predict Success in Community
Policing...................................................... 137
4.31 Summary of the Explanatory Power of the Factors That May Help
Predict Success in Community Policing......................... 138
4.32 Correlation Matrix for Logistic Regression Model............... 142
4.33 Summary of Logistical Regression Statistics.................... 143
4.34 Summary of Multivariate Regression StatisticsSuccess of
Implementation................................................ 144

4.35 Summary of Multivariate Regression StatisticsCurrent Success
of Community Policing......................................... 145
4.36 Summary of Multivariate Regression StatisticsFuture Success of
Community Policing................................................. 146
6.1 Summary of Study Findings..................................... 193
6.2 Summary of Hypotheses Testing................................. 194
E-l Differences between Gianakis and Davis and Current Research in
Components of Community Policing by Agency Type............ 227
E-2 Differences between Gianakis and Davis and Current Research in
Components of Community Policing for Departments in Florida by
Agency Type.......................................................... 228
E-3 Differences between Gianakis and Davis and Current Research in
Goals of Community Policing Implementation by Agency Type ... 229
E-4 Differences between Gianakis and Davis and Current Research in
Goals of Community Policing Implementation for Departments in
Florida by Agency Type............................................... 230
E-5 Differences between Gianakis and Davis and Current Research in
Policy Changes to Implement Community Policing by Agency
Type................................................................. 231
E-6 Differences between Gianakis and Davis and Current Research in
Policy Changes to Implement Community Policing in Florida by
Agency Type.......................................................... 232
E-7 Differences between Gianakis and Davis and Current Research in
Patrol Officer Duties in Departments Implementing Community
Policing by Agency Type.............................................. 233
E-8 Differences between Gianakis and Davis and Current Research in
Patrol Officer Duties in Departments Implementing Community
Policing for Departments in Florida by Agency Type................... 234

E-9 Differences between Gianakis and Davis and Current Research in
Levels of Support for Community Policing for Departments by
Agency Type........................................................... 235
E-10 Differences between Gianakis and Davis and Current Research in
Levels of Support for Community Policing for Departments in
Florida by Agency Type................................................ 236
E-l 1 Summary of Multivariate Regression StatisticsSuccess of
Implementation................................................. 237
E-12 Summary of Multivariate Regression StatisticsCurrent Success
of Community Policing.......................................... 237
E-l3 Summary of Multivariate Regression StatisticsFuture Success of
Community Policing............................................ 238

Our society requires that the police deal with an incredibly broad range of troublesome
situations. Handling these situations within the limitations that we place on the police is the
essence of policing.
-Herman Goldstein
Community policing has taken the country by storm. Many police agencies around
the country, large and small, are embracing this concept as the wave of the future and
the solution to many of their problems. There are 13,533 police agencies in this
country (National Public Safety Information Bureau 1999, xix) and nearly all are run
by local governments. American policing is therefore community-oriented by its very
structure, but many police departments today formally commit to the concept of
community-oriented policing. Although the community policing concept has been
a common place idea for more than twenty years and many departments claim to have
adopted it, our understanding of what it is remains open to debate.
Many departments have implemented different strategies or tactics that they have
labeled as community policing, but there still is some agreement that all are doing
community policing without necessarily doing the same thingor even close to the
same thing. What they seem to agree about is that community policing is a good
philosophy for police agencies.

What is a philosophy? In this context a philosophy is what you think and believe
(Trojanowicz and others 1998,4). Beliefs or ideas seem therefore to be what is
changing about policing, but is it changing actual police operations?
The prevailing wisdom has been for many years that the police control crime
they are the thin blue line between the law-abiding public and the violent criminal
element (Fogelson 1977, 187). That view continues to be popularized today by
television and movies. Yet scholars now generally recognize that the police are
largely involved, and always have been, not with crime fighting, but with order
maintenance (Wilson and Kelling 2000, 8). Within their order maintenance function,
police exercise wide discretion deciding whom to contact, whom to arrest, and whom
to ignore with little input from their supervisors or the public (Lipsky 1980, 201).
In recent years, advocates of community policing suggest that by partnering with
the community the police can enhance their order maintenance function and
therefore both their legitimacy and efficiency. Because their discretion would be
informed by community sentiment and, if the community would take some
responsibility for order maintenance, more if not better policing would get done with
the same commitment of police resources, so its proponents argue (Community
Policing Consortium 2000, 126).
Thinking about what the community wants and expects from its police department
and incorporating that into police operations is a radical departure from the way
police departments have functioned for many years which is epitomized by Joe

Fridays just the facts maam attitude. The police then viewed themselves, and
were viewed more generally by the public as the experts or professionals that would
handle the war on crime without much outside assistance or interference. This
professional model persists and certainly where it persists, has not eliminated police
problems. A case in point: the Rodney King beating in 1991. More recently, in 1997,
Abner Louima was assaulted with a toilet plunger while in the custody of the police in
New York City and in 1999 Amadou Diallo was shot forty-one times by New York
Police detectives as he reached for his wallet that the detectives mistook for a gun
(Lardner and Reppetto 2000, 330-1).
In spite of these well-publicized incidents or perhaps because of them, community
policing is popular with politicians and citizens alike. In 1993, President Clinton
made community policing the cornerstone of his crime fighting efforts and vowed to
put 100,000 new police officers on the street to do it (Roth and Ryan 2000, 9). At the
same time crime rates have fallen steadily and politicians have been quick to take
credit, although no one knows if this decline is attributable to community policing, or
has anything to do with that popular idea.
Purpose of the Study
How in fact does this new philosophy effect actual operations? Are police
departments indeed making a commitment to redefine themselves in a new image?
What does top leadership think about this idea, specifically how do police executives

define community policing? Little research addresses such questions. Because top
leadership in our police organizations is critical and has a strong impact on the
behavior of police organizations, community policing itself both as an idea and as a
practice ultimately rests on how local police executives, i.e., chiefs and sheriffs,
perceive this concept and then act upon their beliefs.
The purpose of this study is to learn what the top police executives around the
country think of community policing and try to evaluate its impact on their
community policing transitions. Community policing by now has permeated police
circles and so the time has come to see how the idea is practiced in its mature form
that is now popular throughout the United States. The views of citizens, police
employees, and politicians are also important, but they will not be included in this
study. Rather the focus is on top police executives, their perceptions, and how they
act upon those perceptions. For here is where the rubber meets the road, so to
speak or where ideas turn into reality. This view from the top can provide us with
rare insights into what is actually happening in police organizations, or what
community policing really means in practice. Hopefully, in addition this research will
give us some insight into the future of community policing.
In order to study police executives views this study collected data using a mail
survey followed by a limited number of telephone interviews. Data were collected
from the top executives on both what their departments were doing with community
policing and what they thought about its underlying tenets. The study investigates six

hypotheses concerning its definition, implementation, support, and underlying tenets.
The telephone interviews were used to put flesh on the bones of data collected by
the mail survey. It further revealed some contradictions between the day-to-day
practices of community policing and community policing in the literature.
Organization of the Dissertation
The police literature has three historic phases. The first will be called the
watchmen phase. The second phase is called the professional model and began in the
early 1930s. The last phase is the community policing phase that began in the 1980s
and continues today. Thus, the literature review in chapter two addresses the
development of these theoretical models and documents the problems with them.
Chapter three lays out the six research questions and the methodology used by this
study to examine where we are today with the community policing concept. The
methodology includes both quantitative and qualitative analyses.
Chapter four is the first of two chapters that presents the research results. Chapter
four sums up the results of the mail survey with 388 police executives responding
(police chiefs and sheriffs) from around the country. Chapter five provides the results
from the follow-up thirty minute telephone interviews with thirteen executives who
responded to the mail survey.

Finally, chapter six draws together the studys findings and conclusions. This
chapter concludes by offering suggestions for the next steps of future research
involving the community policing model.

To different minds, the same world is a hell, and a heaven.
Ralph Waldo Emerson
This chapter provides an overview of community policing by describing how the
concept emerged, how it is defined in the literature, and the challenges it currently
faces. This chapter begins with a brief discussion of the origins of policing in the
United States. Here the emergence and refinement of the professional model of
policing is outlined. Next, the review shows how the idea of community policing
began by describing its origins, evolution, and its current intellectual problems.
The Origins of Policing in the United States
As settlers came to America most lived in rural settings with little need for policing
in any formal sense. However, in the urban areas the citizens adopted the night watch
system to protect themselves, under the direction of the Constable of the Watch,
modeled after the system in urban England. In rural areas the citizens adopted the
county form of civil government and adopted the office of sheriff, common in rural
England, to keep the peace (Germann, Day, and Gallati 1973, 65).
Consistent with their mistrust of central authority, both the offices of constable and
sheriff in the United States were often elected positions with short terms, often less

than two years. As a result, and even to this day, sheriffs and constables are rarely
competitively selected, appointed, well-trained, or career-oriented people (Germann,
Day, and Gallati 1973,65).
Municipal police forces slowly emerged from these early forms of watchmen and
sheriffs. Initially, every able-bodied male over the age of sixteen was required to
serve on the night watch. Some members of the community who did not want to
serve paid others to take their places. In 1833, municipalities began to pay those who
were willing to serve. Slowly a process of using political influence to acquire these
positions emerged as some searched for a way to supplement their income by working
during the day in their chosen fields and sleeping, surreptitiously, during a paid tour
on the night watch (Germann, Day, and Gallati 1973,65-6).
By the mid-nineteenth century the larger cities like Boston, Philadelphia, and New
York eliminated their watch systems in favor of full-time paid police departments.
This movement was again, modeled after the British system that was created by the
passage of The Metropolitan Police Act of 1829. That act was sponsored by Sir
Robert Peel and was based on the following principles (Germann, Day, and Gallati
1973, 65-6):
1. The police must be stable, efficient, and organized along military lines.
2. The police must be under governmental control.
3. The absence of crime will best prove the efficiency of police.
4. The distribution of crime news is essential.

5. The deployment of police strength both by time and area is essential.
6. No quality is more indispensable to a policeman than a perfect command
of temper; a quiet, determined manner has more effect than violent
7. Good appearance commands respect.
8. The securing and training of proper persons is at the root of efficiency.
9. Public security demands that every police officer be given a number.
10. Police headquarters should be centrally located and easily accessible to
the people.
11. Policeman should be hired on a probationary basis.
12. Police records are necessary to the correct distribution of police strength.
The police organized under Metropolitan Police Act were focused on prevention
and not on the enforcement of criminal laws. In part that act states (Stead 1985,40-1;
emphasis in original):
It should be understood, at the outset, that the principal object to be attained
is the Prevention of Crime. To this great end every effort of the Police is
to be directed. The security of person and property, the preservation of the
public tranquility, and all the other objects of a Police Establishment, will
thus be better effected than by the detection and punishment of the offender,
after he has succeeded in committing the crime.
What were the origins of this formal approach to policing? In the late eighteenth
and early nineteenth centuries, England experienced an increase in crime that led to
wide-spread lawlessness (Germann, Day, and Gallati 1973, 59). At this time
Britain made use of the watch system, the same model later adopted by many cities
and towns in the United States, that used community residents without uniforms, paid
or unpaid, to maintain order. In theory these citizens became responsible to the
community for order maintenance (Germann, Day, and Gallati 1973, 59-60).

At the same time the French already had a strong, centralized, police force. In
France the policeman was a military man; his background was military, his uniform
was military and, like a soldier, he was deployed as an instrument of central
government. Some of the English admired the French police for their efficiency and
professionalism while others feared it because they viewed it as an infringement on
their liberty (Emsley 1983, 28-30).
In 1785, a bill was introduced in the English Parliament to provide policing to
London, modeled after the French. The leading British newspaper of the day wrote
our constitution can admit nothing like a French police; and many foreigners have
declared that they would rather lose their money to an English thief, than their liberty
to a Lieutenant de Police. This concern over surrendering liberty to a formal police
force took forty-four years to resolve before the Metropolitan Police Act of 1829
became law (Emsley 1983,28-30).
In the end, many of Peels ideas mirrored those of the French, but a significant
political battle ensued to achieve them, more then is usually acknowledged by those
referring to the British system today (Emsley 1983, 67). Americans were even more
suspicious of government than the British so the American police of the mid-
nineteenth century adopted only some of Peels ideas. For example, the British police
were centralized as part of the national government in 1829, but the Americans
continued to view the police as a local function and made policing subject to local
political control; a decision that bedevils the efficiency of the American police to

this day. The British police focused on crime prevention while the American police
focused on responsiveness to local control by elected politicians (Fogelson 1977,17).
The net effect of this development was that politicians came to rely on police
departments to provide a wider variety of services to their constituents, such as
checking on runaway children, running ambulances, managing employment bureaus,
operating soup kitchens, monitoring polling places, and bringing out the vote. In turn,
police officers came to rely upon the politicians for their positions in the department.
At issue during the nineteenth century was not whether the police departments would
be operated in someones interests, a point firmly settled, but in whose interests they
would be operated (Fogelson 1977, 21). This vision dominated American policing
until the 1930s. Even though police were largely beholden to local political leaders,
they still'had large discretion over how they functioned. [T]he big-city police have
always done more than just enforce the law, keep the peace, and serve the public.
They have also decided, or at least helped to decide, which laws to enforce, whose
peace to keep, and which public to serve (Fogelson 1977, 12).
The Professional Model and Its Limitations
According to a prominent police historian, Samuel Walker, because of political
influence, nineteenth century American police did not effectively prevent crime
(although this was not their focus as we have seen), nor did they effectively provide
other services (Walker 1980, 61). The police faced many challenges at the turn of the

century including incompetence, corruption, and lawlessness. Much of local
government was undergoing some fundamental professional reform due to the
These progressive reformers stressed the need for administration that was rational,
impartial, and free of political influence (Stillman 1974,8-9). Within the police ranks
this reform movement was led by August Vollmer, the police chief in Berkeley,
California, O.W. Wilson, one of Vollmers proteges who served as police chief in
Wichita, Kansas and later in Chicago, Illinois, and J. Edgar Hoover who headed the
FBI from 1924 to 1972 (Unger 1976,48).
These police reformers, along with their Progressive counterparts in local
government, believed that politics was the source of the problems plaguing the police
and that professionalization was the solution (Fogelson 1977,49; Kelling and Moore
2000,96). The reformers came up with three distinct recommendations to
professionalize the police. First, centralize control of the department with the chief in
charge and insulate him from political pressure. Second, upgrade the quality of the
rank-and-file officers by improved training, and finally, focus on efficient crime
fighting instead of trying to be a service agency that provides a variety of functions to
the community (Fogelson 1977,159-61).
This professionalization movement had a strong scientific management emphasis
that focused on efficiency and economy as the dominant value via strong, centralized,
bureaucratic control of the police (Fogelson 1977, 53). Although some reformers,

like Vollmer (1971), emphasized improving the quality of the individual police
officers through better training, others, like O.W. Wilson (1950), concentrated on
improving the police organization. Both seemed convinced that there was one right
way to police; remove the police from politics and professionalize the department and
individual officers. The overall effect was to isolate the public from policing and to
leave policing to the professionals.
This focus on professionalized policing has been dubbed the professional model
and its emphasis is on professional crime fighting to reduce crime as its goal. So,
the professional model emphasized a shift from being responsive for serving those
who controlled them to fighting crime by direction from within the police
organization. The English police focused on crime prevention since their inception in
1829, but police in the United States largely did not accept this mandate until a
century later in the 1930s. The professional model used three primary strategies to
deal with crime; motorized random patrol, rapid response to citizens calls for service,
and retrospective investigation of crimes reported to the police (Moore, Trojanowicz,
and Kelling 2000,41).
The professional model thus represented a substantial shift in the underlying
philosophy of policing from local responsiveness and community service to a
professional autonomy focused on crime fighting. The reformers especially viewed
this as the way to prevent crime. This view of crime prevention was very narrow.
Chief Parker of the Los Angeles Police Department once highlighted its narrow

application by suggesting that the police could only deal with the effects of crime and
not its causes. As the so-called thin blue line that separated the lawless from the
law-abiding, the police could deter criminal activity by increasing the likelihood of
intervention, apprehension, and punishment (Fogelson 1977,187).
By embracing this professional theory police departments made three substantial
changes in their operations. First, they stopped trying to address the causes of
crime and increasingly moved resources away from such community services as boys
clubs, employment bureaus, athletic leagues, and other welfare programs. Second,
they taught citizens to lock their cars and homes, call the police to report suspicious
activities, and generally, be aware that criminals would try to victimize them. Finally,
the police themselves became more proactive by aggressively attempting to find
criminals instead of waiting for citizen complaints. As a result, the police stopped
people on the street to search out violations, to ask possible suspects about their
activities, and search for contraband. This strategic shift represented a dramatic new
way to use police resources (Fogelson 1977,187). Using Trevor Bennetts analytic
framework, the professional models philosophy, principles, goals, organizational
strategies, general operational strategies, and specific operational strategies are
summarized in Table 2.1 (1994, 229).
The professional model certainly can be said to have been an appropriate approach
when Vollmer and O.W. Wilson were alive and faced departments that were

disorganized, poorly equipped, poorly trained, inefficient, and corrupt. However,
over the last thirty years that model has been severely criticized as having a poor
Table 2.1 Professional Model Philosophy and Practices Discussed in the Literature
Philosophy Principles Goals Organizational Strategy General Operational Strategy Specific Operational Strategies
Professional Professionalism Political Neutrality Civil Service Job Requirements Examinations
Model of Policing Crime fighting Impartial law enforcement Intercept criminals in the act Incarcerate criminals Random Patrol Rapid Response Retrospective Investigations In-progress arrests Response time Follow-up arrests
track record of preventing crime, and for insulating the police from the public they
served to the point the police became unresponsive (Goldstein 2000a, 19; Moore,
Trojanowicz, and Kelling 2000,41; Community Policing Consortium 2000,120).
The underlying assumption of the professional model is that it reduces serious
crime. Police departments embraced the professional model to fight serious crime by
actively seeking out and arresting law breakers. The first line of defense is random
patrolsperhaps the police will be in the right place at the right time. Failing this,
they rapidly respond to the reported crime and arrive quickly enough to arrest the
perpetrator. If that tactic does not work, the retrospective investigation can identify
suspects leading to their arrest, conviction, and incarceration. This military-minded
logic makes sense on the surface and over the years broadly appealed to both the
police and those to whom they are accountable. However, serious questions about its

efficacy arouse by the early 1970s (Moore, Trojanowicz, and Kelling 2000,49;
Skolnick and Bayley 1988,212).
In 1974, the results of the Kansas City Preventative Patrol Study (Kelling and
others) were published. The study documented that even when the number of officers
patrolling the streets was doubled, neither the crime rate nor the fear levels decreased.
These results stunned the policing world, but many police departments have not
incorporated these findings into their operations to this day, even though it appears
that random patrol is not an effective way of reducing crime (Kelling and others 1974,
v; Sparrow, Moore, and Kennedy 1990,15; Gianakis and Davis 1998,491; Moore,
Trojanowicz, and Kelling 2000,49).
A second study in Kansas City addressed the effectiveness of rapid response. That
study found that the speed of the citizens call to the police, not the police response
time was the critical variable. The study also documented that citizens often delayed
their call to the police to either ensure the suspects flight or to seek moral support
from loved ones before dealing with the police investigation. The delay was usually
between twenty and forty minutes, long enough to ensure the police were responding
to a cold situation and not an emergency (Kansas City Police Department 1977;
Kelling and Coles 1996, 92-3; Moore, Trojanowicz, and Kelling 2000, 49).
Finally, a study of retrospective investigations appeared in the late 1970s. The
authors concluded that the most critical factor in follow-up investigations for
identifying the suspects was whether or not the victim or witnesses knew them. If

they could not help, no amount of forensic wizardry would make up the difference,
and the police rarely solved the case in follow-up investigation (Greenwood, Chaiken,
and Petersilia 1977,225; Sparrow, Moore, and Kennedy 1990,47; Moore,
Trojanowicz, and Kelling 2000,49).
Interestingly, this three-pronged attack on the assumptions underlying the
professional model tactics focused on the tactics thought to be most effective for
dealing with crimes like murder, rape, robbery, assault, and burglary. These crimes
have some likelihood of being seen by a passing patrol car, are more likely to be
reported to the police, and are usually assigned high priority within police
departments for follow-up investigation. However, these tactics, basic to the
professional model operation, proved unsuccessful where they were expected to
perform best (Moore, Trojanowicz, and Kelling 2000,49).
What is Serious Crime?
As if the attack on the assumptions underlying the professional models tactics was
not enough, questions arose regarding what constitutes serious crime. That debate led
to serious questions about how and where the police should deploy their limited
resources. Victims define serious crime as crime that happens to them, but the police
must differentiate between various levels of seriousness to ration their resources. As
a result, the determination of the seriousness of crime is a social judgment and not an
individual one. It is also a value judgment. Factors that may be included in the

judgment of a crimes seriousness include physical violence or violation (i.e., assaults,
murders, and rapes), the size of the victims loss, and the relationship between the
offender and the victim (Moore, Trojanowicz, and Kelling 2000,43).
These factors, however, also contain four assumptions. The first assumption is that
crimes committed by strangers are more serious than those committed by intimates
because the stranger crimes pose a larger threat to general society. This view
overlooks the terror that is experienced by the victim. Victims may well be
victimized by someone they know or love. Often they are in a position that does not
permit them to terminate the relationship and therefore repeated victimizations can
occur (Moore, Trojanowicz, and Kelling 2000,44; Sparrow, Moore, and Kennedy
Second, the focus on violent crimes obscures the effect of fear on citizens. Many
incivilities incite more fear than violent crimes. Citizens buy guns, locks, and dogs
more frequently in response to fear than in response to being victimized. As the
police focus on violent crimes to the exclusion of incidents that engender fear, they
may miss an important opportunity to contribute to the solution to crime. Fear, even
if the source is non-criminal, can transform a neighborhood into an isolated camp
(Moore, Trojanowicz, and Kelling 2000,44; Wilson and Kelling 2000,4).
Third, defining the seriousness of a crime based on its absolute financial impact
misses the victims ability to absorb the loss. Is a $10,000 insured loss more serious
than a stolen $300 social security check representing the victims entire net worth?

Here a serious potential for injustice and ineffectiveness in the deployment of police
resources based on aggregate financial costs as opposed to individualized costs are
evident (Moore, Trojanowicz, and Kelling 2000,44; Sparrow, Moore, and Kennedy
1990, 176).
Finally, the focus simply on individual victimizations overlooks the symbolic
significance of certain crimes. For example, our response to the crimes of child
abuse, spousal abuse, and political corruption is important because it strengthens, or
redefines, broad social norms (Moore, Trojanowicz, and Kelling 2000, 45). By
focusing on serious crime under the assumptions of the professional model, the police
may see that their responsibility is only to enforce the law, primarily in situations
brought to their attention, and to do so under the watchful eye of the courts. By
adopting an alternative view of serious crime that includes the symbolic issues of
safety within relationships, the effects of fear, and the impact of crime on the most
vulnerable victims would require the police to expand their role to defend some social
territory in addition to the streets (Moore, Trojanowicz, and Kelling 2000,45).

Missed Opportunities to Prevent Crime
Another way to view this issue is to look at some of the opportunities to reduce
crime that the police have missed over the years. There are at least three possibilities
highlighted in the literature: First, communication with citizens; second, problem
solving; and third, limited citizen participation.
First, although the police have recognized the need for communication with the
public, they have made erroneous assumptions about how to obtain information. The
reactive portion of the professional model assumes that citizens will call the police
and accurately share information with them when crimes occur. In fact, often they do
not. Crime reporting rates have held steady at between 35%-38% for years. These
rates vary by crime type with homicides and auto thefts being more regularly reported
by contrast to sexual assaults and thefts which are more rarely reported (Bureau of
Justice Statistics 1994,123).
Sharing is usually something reserved for an equal partnership and the police have
not been good at creating such partnerships with the public they serve. They have
spent millions of dollars creating 911 systems, but they have rarely invested in
partnerships with the community. To some extent, they have done the opposite. The
police are often insensitive to either the communitys fear of retaliation or their fear
from less serious violations or disorder such as prostitution, loitering, panhandlers,

drunks, or the mentally ill (Wilson and Kelling 2000,4). Partnerships also take time
and personal attention to cultivate, but the police have sequestered themselves in their
cars. Building partnerships is challenging from within a police car traveling at thirty
miles per hour through the neighborhood (Moore, Trojanowicz, and Kelling 2000,
Second, the police respond to many similar calls for years without effectively
applying the knowledge from past experiences to prevent reoccurrences (Goldstein
1990, 33). They understand many of the factors that precipitate crime, such as people
with criminal histories, bad situations, alcohol and drugs, and frustrating
relationships, but they ignore or at best often are ambivalent about using that
understanding to prevent crime. Applying criminal law only to those specific
situations to which they are summoned does not necessarily use resources wisely, nor
effectively cope with local crime (Moore, Trojanowicz, and Kelling 2000, 52).
Finally, the police frequently assume that the communitys self-defense capabilities
are limited to helping the police. Because of that assumption, personal skills are not
fully utilized by law enforcement. The police know they cannot police the
community by themselves, but often imprisoned by their own professional model
perspectives, they have not invited the community to participate with them in
reducing crime (Moore, Trojanowicz, and Kelling 2000, 52-3).
The professional models emphasis on the value of efficiency quite possibly was
ultimately at the expense of the values of responsiveness, equity, and effectiveness.

The reformers accomplished a great deal and improved the police by removing much
of the incompetence, corruption, and lawlessness from which they suffered at the turn
of the century. However, the entire field of public administration has learned a good
deal about competing values since the Progressive era. As Aaron Wildavsky
suggested, efficiency does not tell you where to go, it just requires that you get there
with the least possible effort (1979,131). Some argue that it may not have been a
coincidence that crime rates rose and those that could afford to do so fled the city in
search of relief in the suburbs during the time the police professionalized in pursuit of
efficiency (Trojanowicz 1994,259).
Is the Problem with the Professional Model Police Management?
It is tempting, and perhaps even natural, to think that the reason professional
policing has failed to live up to expectations is due to the poor management of police
resources and not underlying flaws in the model. Although research has shown the
weaknesses in random patrol, rapid response, and retrospective investigation, some
police administrators still believe that if both patrol officers and detectives could
devote more of their time to these strategies, or slight variations of them, or obtain
more resources to achieve these strategies they would prove to be as effective as
originally believed (Eck and Spelman 2000,196).
Research shows nonetheless that patrol officers and detectives could better use
their time (Gay, Schell, and Schack 1977, xii-xiii). For example, different responses

were possible and just as effective in many cases. These responses included handling
calls over the phone, through the mail, or as walk-in traffic to the police station
handled by civilians instead of police officers (McEwen, Connors, and Cohen
1986,10-2). To free up detective time in many departments, cases without leads are
not investigated by detectives. The investigation is considered complete when the
initially assigned patrol officers simply finish their investigation on the street
(Greenwood, Chaiken, and Petersilia 1977, 243).
With additional time available to control crime, police administrators need
information concerning crime and criminals to direct these new resources
productively. The answer is, for many administrators, more crime analysis. Police
departments around the country in recent years formed crime analysis units whose
purpose is to use police records to analyze the nature of crime and criminals.
Departments use this information to develop patterns and disseminate information to
assist patrol officers and detectives in intercepting criminals as they commit crimes.
Although these efforts may improve police operations, it is doubtful that they reduce
crime (Eck and Spelman 2000,196).
Why is this so? The move to better utilize police resources has one major
limitation. Crime analysis is really an attempt to determine where to apply
established police responses. The responses are set before problems are understood;
the same responses are used on widely differing problems. Instead the aim should be
to understand a problem and then determine what is needed to solve it (Eck and

Spelman 2000, 197). From this perspective, management improvements are always
welcome, but it is important to remember that they are only a means of improving
police capabilities to solve problems that a community wants or sees as needing to be
Crime Prevention Under the Professional Model
There is another issue already mentioned in passing that needs to be explicated.
Although our policing heritage is English, the English chose to emphasize crime
prevention while the American police emphasized responsiveness. With the
development and implementation of the professional model, American police
emphasized crime prevention only through efficient crime control that stressed
finding the suspects and bringing them to justice.
John Alderson, who served as Chief Constable for Devon and Cornwall in Great,
Britain, points out that societys preoccupation with the theory of controlling crime
only through punishing or treating offenders is clearly outmoded (1979, 221). He
reaches this conclusion by reasoning that relatively few crimes come to police
attention, very few offenders are caught, fewer are convicted, and fewer still are
punished via incarceration. Some numbers from the United States support his thesis.
In 1994, victimization studies showed that roughly 37% of serious crimes (rape,
assault, theft, burglary, and arson) were reported to the police (Bureau of Justice
Statistics 1997, 3). FBI statistics show that during that same year the police cleared

21.4% of those cases by arrest (Federal Bureau of Investigation 1995, 208). A study
concerning conviction and incarceration rates shows that 54% of the people charged
with serious crimes are convicted, and 59.3% of those are imprisoned (Boland,
Mahanna, and Sones 1992, 3). So the chances of conviction are roughly 4.1% and the
probability of going to prison is 2.5% or 2.5 out of one hundred crimes committed.
As Alderson points out, where is the deterrence in that?
The bottom line is that the assumptions of professional model of policing contains
serious flaws and limitations. Many people have been searching for a better way
since the mid-1970s. The last twenty years have seen a great deal of discussion and
some experimentation with a new way to view policing. We turn our attention to it
An Introduction to Community Policing
In 1982, James Q. Wilson and George L. Kelling published an article in The
Atlantic Monthly called Broken Windows: The Police and Neighborhood Safety.
Although they did not use the term community policing in that article, they are
credited with starting, or in their own view restarting, the community policing
movement (2000,12).
This article builds on a great deal of ground-breaking work that both authors had
already contributed to the field of police research. Wilson had written extensively in
the field and had suggested that most police departments adopted one of three styles

and the choice of style was influenced by their communitys view of the proper police
role. Some departments generally in small rural or suburban communities, focused
primarily on order maintenance and retained the night watchman style (Wilson 1968,
140). Others, especially big city departments like Los Angles, focused on criminal
law enforcement and resorted to a legalistic style in which the law and enforcement
to the letter was the primary goal (172). Finally, some adopted a service style that
responded to both requests for order maintenance and law enforcement, but used their
powers of arrest very sparingly in both situations. This service style was mainly an
option where there was wide-spread community agreement about what constituted
order and how it could be obtained (200).
While Wilson suggested different types of policing in his early research, George
Kelling worked on studying actual police operations. He was the lead researcher in
the Kansas City study which explored the effects of random patrol on crime and fear
(Kelling, and others 1974). It was the first of several empirical studies during the
1970s that began to question the efficacy of police practices under the professional
model from roughly 1930 through the early 1970s.
In Broken Windows, Wilson and Kelling hypothesize that there is a link between
disorder and crime. In their view, the police have overlooked this link by relying so
heavily on law enforcement as the way to reduce crime and fear. The police have
suggested that if they solve more crimes, make more arrests, and collect better
evidence citizens will be less fearful as a natural result of such professional police

work (2000, 8). However, the article points out that the fear most citizens feel is not
related to being the victim of a serious, or even a minor crime. Rather, it stems from
the sense that the street is disorderly and is a source of distasteful, worrisome
encounters (2000, 5). Therefore, the police are much more likely to successfully
address fear if they focus on those potentially distasteful, worrisome encounters
involving disreputable, obstreperous, or unpredictable people: panhandlers, drunks,
addicts, rowdy teenagers, prostitutes, loiterers, and [the] mentally disturbed (2000,
Wilson and Kelling suggest that there is a link between fear and the eventual
appearance of crime. Unchecked offensive behavior breeds more serious offensive
behavior and criminal activity because people in the area seem to believe that no one
cares. When no one cares, it is more likely that bad behavior will be ignored. As this
untended behavior escalates, neighbors and those who pass through the
neighborhood get a sense that crime is increasing and modify their own behavior
accordingly. As this fear increases, people and neighbors start avoiding one another,
and as a result, informal community controls weaken (2000, 6).
In response to this chain of events, they suggest that the police refocus on
disorderly behavior and help citizens take back their neighborhoods. Some success
using foot patrols in Newark is just one example that resulted in a decrease in crime
and a decrease in levels of citizen reported fear (Wilson and Kelling 2000,14-16).

Defining Community Policing
Wilson and Kelling are largely credited with starting the community policing
movement because they offered the clearest statement of what community policing
and foot patrol should be able to accomplish and why (Greene and Taylor 1988,
201). That is, police working closely with community residents will increase the
residents confidence in both the police and the residents ability to impact crime, at
the same time enhancing social control and reducing the signs of crime. They did not
use the term community policing, but others connect the concept to Wilson and
Kellings emphasis upon the physical appearance of the neighborhood and its impact
on crime. From the street police officers view of community policing, Wilson and
Kellings viewpoint is a convenient starting place for defining and understanding this
idea. Nonetheless, it also is a bit more complicated than merely the broken windows
thesis since today there are many other views about community policing.
The literature describes community policing as an overall philosophy that
encompasses different tactics including fear reduction, order maintenance (that
together are accounted for in the broken windows thesis), community involvement,
and problem solving (Wycoff 1988,104; Oettmeier and Brown 1988,127-8).
Skolnick and Bayley suggest that there are four core principles of community
policing: community-based crime prevention, reorientation of patrol, increased police

accountability, and decentralization of command (1988, 1-19). Bayley modified this
definition by coming up with the acronym CAMPS to identify the recurring themes
that emerge from police departments as they operationalized community policing.
CAMPS stands for consultation, adaptation, mobilization, and problem solving.
Consultation means asking communities regularly and systematically about their
needs and how the police Can meet them. Adaptation means command devolution so
that those closest to the community members are empowered to assign resources as
they deem appropriate. Mobilization suggests involvement of community residents
and agencies other than the police to solve community problems. Problem solving is
a term for identifying and solving issues that generate crime and insecurity (Bayley
However, these definitions are not uniformly accepted. In fact, community
policing is defined very differently by different scholars. Eck and Rosenbaum
suggest that community policing can be viewed through three different lenses:
effectiveness, equity, or efficiency (1994, 5). If seen from its effectiveness vantage
point, community policing will focus on the ends of policing, that is, on successfully
solving community problems. If focused on equity, community policing continues its
focus on means but will also try to reduce conflicts between the department and the
public by hiring more minorities, providing sensitivity training to officers,
encouraging citizen complaints, or creating more stable beat assignments. If focused
on efficiency, departments still concentrate on means, but they may have civilians

handle non-emergency calls over the phone to reduce costs and increase available
officer time. If that time is not devoted to working effectively with the community it
may add little to police effectiveness (1994,18-9). Eck and Rosenbaum clearly favor
focusing on effectiveness when defining community policing, but many view the
changes described above in the areas of equity or efficiency as an equally critical
version of community policing (Skolnick and Bayley 1986, 212-20).
Another definition in the literature focuses more on community then on police. In
this case community policing is ...all elements in a community, both official and
unofficial,.. .conceiving] of the common good and combining] to produce a social
climate and... environment conducive to good order and the happiness of all those
living within it (Alderson 1979,46). In this view the police serve in a supporting
capacity and not in a defining role.
Based on these definitions, community policing suggests an alternative approach to
policing to the professional model of simply relying on law enforcement to control
crime. Numerous departments around the country certainly are implementing the
concept (Bayley 1988,226; Eck and Rosenbaum 1994, 3; Gianakis and Davis 1998,
498; Goldstein 2000b, 71; Kelling 2000, 60; Leighton 1994,209; Moore 1994,285;
Sadd and Grinc 1994, 27; Weisel and Eck 1994, 53; Wilkinson and Rosenbaum 1994,
110). However, these definitions do not tell us in reality what community policing is,
or why or how it operates in practice.

In spite of this lack of an agreed upon theory, departments throughout the
country are changing their operations to embrace community policing (Eck and
Rosenbaum 1994,4). The broken windows thesis sparked the start of the
community policing movement, especially by addressing the possibility of a link
between fear and crime.
Problems with the Broken Windows Thesis
Greene and Taylor reviewed several neighborhood level studies as a way to
evaluate the critical points of Wilson and Kellings broken window thesis. The first
key point in the model was that as physical and social incivilities increase, informal
social control weakens. Based on several studies they reviewed, Greene and Taylor
concluded there is no evidence to support that link. They found there was an
intervening variable that accounted for lower levels of informal social control and
higher levels of incivility in neighborhoods. That variable was whether or not the
resident was a renter. So, there was no direct link between incivility and lower levels
of informal social control (1988,201-2).
The second key criticism of the Wilson and Kelling model concerned the assertion
that as physical and social incivilities increase, so too does fear. Greene and Taylors
review of this element of the model revealed that there is a link between incivilities
and fear, but it was a conditional link and not a unilateral one as Wilson and Kelling
proposed. The studies showed that the link was present in middle class

neighborhoods but not in those that were poor or well-to-do. So, Greene and Taylor
argued that the Wilson and Kelling model was really context-sensitive (1988, 202).
The last problem with the model involved the conclusion that incivilities are linked
with crime. Here Greene and Taylor discovered a set of intervening variables that
explained most of the variation in both incivility and crime. Those variables were
income, education, and race. They concluded that the linkage between incivilities
and crime appears to be largely driven by the linkage of both [incivility and crime]
with social class and does not exist independently (1988, 202). As a result of their
analysis, Greene and Taylor concluded that the broken windows thesis contains
major flaws and deflects attention from the more likely sources of fear, crime, and
incivility, namely social class factors (1988,203). Others add even more major
factors that influence crime: age, sex, racial or ethnic distribution of the population,
and economic conditions (Klockars 1988, 250).
Problems with Other Aspects of Community Policing
Part of the political appeal of community policing is that it deliberately seeks to
involve the community in a partnership with the police to prevent crime and to
identify problems the police should address alone, with other governmental agencies,
or with citizens (Wycoff 1988,105). Wilson and Kelling suggest that this partnership
between the community and the police and can stop the physical deterioration of the
community and thereby prevent crime. They also suggest that the presence of the

police and citizens working together to better their neighborhoods will help to reduce
fear (Wilson and Kelling 2000,4). Others have argued that even if those ends are not
achieved, a closer working relationship between the community and the police is a
worthwhile end in itself (Skolnick and Bayley 1986,214). However, these
conceptions of community involvement have some underlying assumptions and
inherent problems that have fueled criticism of community policing.
First is the assumption that citizens want additional contact with the police.
Certainly most citizens believe that the police should be close-by during an
emergency, but there is little evidence that they necessarily want to interact with them
in non-emergency situations. In addition, many minorities who fear crime the most,
and indeed may have good reason to, also fear the police (Greene and Taylor 1988,
203-4; Sadd and Grinc 1994,44; Weatheritt 1988, 173). Another facet of this
assumption is that if there is not a community out there willing to work with the
police or where the community is not readily identifiable, the police can and should
organize or create one. This view frightens some critics of the police and their power.
Mollie Weatheritt expresses part of this concern rather clearly when she writes (1988,
The idea that there is a consensual community waiting to be discovered and,
moreover, that the police are best placed to discover it and to give practical
expression to it is not only a recipe for failure, it is to misunderstand the
nature of social process and the place of the police in the political order.
Social harmony is not waiting to be discovered and it cannot be engineered
by the police.

Second, the entire idea of community policing is really the brainchild of the police
and not of the community. Some believe it is really nothing more than a different
approach to community relations (Eck and Rosenbaum 1994,19; Weatheritt 1988,
167). Others see it as simply another police strategy aimed at shaping and
manipulating public opinion (Lyons 1999, 166; Manning 1988,40) which is another
way of saying community relations. Taking a more optimistic view, there is still the
problem that the police may be trained to do community policing, but the
community is not. That is, the police are training themselves to interact with the
community, but that training is limited to police employees, there is not an analogous
training session for citizens to teach them how to get involved with the police (Sadd
and Grinc 1994,48).
Third, if we accept the idea of community and its legitimacy in influencing police
operations, who represents the community? This problem is particularly troublesome
in low-income sections of communities since they are often poorly organized and
politically underrepresented (Murphy 1988,186). Richard Feynman, a Nobel Prize
winning physicist, was once asked in open court if a particular drinking establishment
that provided adult entertainment was acceptable to the community. His response
epitomizes the challenge to those considering community policing and public
administrators generally in the freeworld: I need to know what you mean by
acceptable to the community. Nothing is accepted by everybody, so what

percentage of the community must accept something in order for it to be
acceptable? (Feynman 1985, 274, emphasis in original).
Finally, consider the assumptions implicit in the very idea of community.
Sociologically, the concept of community implies a group of people with a common
history, common beliefs and understandings, a sense of themselves as us and
outsiders as them, and often, but not always, a shared territory. Although
community in this sense may exist in many parts of this country, it really does not
exist much of urban life today (Klockars 1988,247-8). There are also other ways to
view community including institutions, social commitments, or land use
characteristics, that is as communities of interest. However, community policing as a
concept does not necessarily differentiate between these different conceptions and
plan police intervention accordingly or take into account the difficulty associated with
either creating or maintaining those communities of interest (Community Policing
Consortium 2000,127; Greene and Taylor 1988,205; Manning 1988, 33; Murphy
1988, 186; Weatheritt 1988,170).
In reality, most community policing is carried out where police can define and
understand such units as neighborhoods, precincts, or districts, whichever is easiest to
apply, instead of on the basis of any generalized notion of community. So the
problem with this causal use of the term community is that it can exaggerate what
an officer can actually do for society. Suggesting that an officer can effectively
police an entire neighborhood, district, or precinct extends the notion of a police

officers sense of territorial responsibility beyond any reasonable limits. The level
and intensity of policing in neighborhoods, precincts, or districts is determined by
police administrators and thus may be practically impossible to achieve on the line by
individual police officers. Arguably then, some conclude, community policing is not
an implementable theory, but just rhetoric that permits police administrators to speak
to organizations or groups in areas that are at once, geographically, too large to be
policed and, politically, too large to be ignored (Klockars 1988,248-9).
Although there is a great deal of disagreement about community policing, there is
also quite a bit of agreement. First, there is enough agreement on the concept within
police circles, that many large and small departments around the country are
implementing this concept1. This fact alone suggests that the idea is worthy of study.
Second, within the academic community there is general agreement that community
policing has brought about a fundamental change in the way American police
agencies operate. Today there is widespread agreement that community policing
must be embraced as a philosophy and not just as a public relations program
(Community Policing Consortium 2000,118; Goldstein 2000b, 73; Greene and
Taylor 1988,196; Kelling and Moore 1988,24; Moore 1994,290; Murphy 1988,177;
Oliver 2000, 81; Oettmeier and Brown 1988, 132; Riechers and Roberg 2000, 350;
Sadd and Grinc 1994,41; Trojanowicz and others 1998, 5; Wycoff 1988, 104-5).
1 The most recent study of implementation finds that 99% of the departments in the state of Florida
employing more than fifty officers claim to have implemented community policing (Gianakis and
Davis 1998,488)

Departments are actively working to find ways to involve the community in
establishing police priorities and in helping the police solve problems (Bayley 1988,
226; Community Policing Consortium 2000,126; Kelling 2000,61; Kelling and
Moore 1988,22, Kelling and Stewart 2000,292; Klockars 1988, 247, Ottemeier, and
Brown 1988, 126; Trojanowicz and others 1998, 5; Wasserman and Moore 2000,219;
Wycoff 1988, 105). The police also are focusing upon solving citizen problems and
not simply responding to calls for service (Community Policing Consortium 2000,
126; Eck and Spelman 2000, 198,205; Goldstein 2000a; Goldstein 1990; Leighton
1994, 218; Weisburd and McElroy 1988,93-7; Wycoff 1988, 105). Police officers
are being tactically deployed in new ways to facilitate their ability to get to know the
local businesses and residents through fixed beat assignments (Community Policing
Consortium 2000, 124,143; Farrell 1988, 76; Goldstein 1990, 27; Goldstein 2000b,
73; Kelling and Stewart 2000, 292; Sparrow 2000,179-80; Trojanowicz and others
1998,13; Weisel and Eck 1994,69). Police organizations are empowering street
level officers and supervisors to solve problems without reliance on the higher levels
of police bureaucracy (Bayley 1988, 226; Community Policing Consortium 2000,
134, 155; Eck and Spelman 2000,205; Goldstein 2000b, 79; Greene, Bergman, and
McLaughlin 1994, 99; Kelling 2000, 69; Kelling and Moore 1988,20; Kelling and
Stewart 2000, 292; Kelling, Wasserman, and Williams 2000, 275; Klockars 1988,
252-3; Leighton 1994, 218; Mastrofski 1988, 59; Meese 2000, 303-4; Moore,
Trojanowicz, and Kelling 1994, 56; Oliver 2000, 211; Riechers and Roberg 2000,

350; Skogan 1994,167; Wilkinson and Rosenbaum 1994, 111-2; Wycoff and Skogan
1994,77). Finally, many police officers are being better trained in the new
community policing methods (Community Policing Consortium 2000,145-6;
Goldstein 2000b, 80; Weisel and Eck 1994, 65; Wilkinson and Rosenbaum 1994,
Community Policing in Practice
So in spite of the perceived weaknesses and flaws in the theory underlying
community policing, in practice it is a new way of policing that has been adopted
throughout the United States, and other countries (Weatheritt 1988; Murphy 1988).
For this reason alone, the concept is worthy of study even if that study is now focused
more on its application than theoretical concerns.
Clearly, the biggest challenge to practical application is the lack of an agreed upon
definition. Community policing seems an elastic term; it means different things to
different people. It includes foot patrol, bicycle patrol, horse patrol, problem solving,
information gathering, victim counseling and services, community organizing and
consultation, education, walk and ride, park-walk-and-talk, knock-on-door programs,
beepers, cellular phones, specialized units, specialized forms of patrol, neighborhood
police offices, traditional criminal investigation, intensified enforcement of drunk-
driving laws, tightened disciplinary procedures, liaison with ethnic groups, victim
support, and rapid response to emergency calls for service (Bayley 1988,278; Kelling

and Moore 1988, 22; Wilkinson and Rosenbaum 1994,120). In other words,
community policing means almost anything the police do including many things they
have always done. Community policing means too many things to different people.
It is also seen as so virtuous that everyone wants to be seen doing it (Bayley
1994,278). Consider for example Murphys summary (1988,178):
[Community policing can be seen alternatively as an ideological response to
the ongoing search for community and order in modem urban society; a
political promise of responsive and responsible police service; a
programmatic set of internal, organizational, and managerial reforms; and
pragmatic, operational strategies aimed at enhancing police effectiveness.
This definitional problem is not likely to be resolved anytime soon and some of the
leading scholars in the field are not concerned by it. Herman Goldstein the Evjue-
Bascom Professor of Law at the Law School at the University of Wisconsin in
Madison, who pioneered problem-oriented policing, recognizes the need, on one
hand, to adequately define community policing, but he equally is concerned that such
effort will lead to oversimplification. He suggests that policing already suffers from
more than its fair share of oversimplification. For example, the terms crime,
violence, and disorder are simple, convenient terms, but they disguise
amorphous, complex problems (2000b, 72).
As a result, he suggests that we should not lose patience over the fact that we have
not come up with the perfect model of community policing. [W]e should not get
stalled trying to simplify change just to give uniform meaning to a single, catchy, and

politically attractive term.... We need to better understand the complicated rather
than search for the simple (Goldstein 2000b, 72-3).
Mark Moore, Chair of the Kennedy Schools Program in Criminal Justice, is also
encouraged by the ambiguity in the concept of community policing. He writes It is
partly the ambiguity of the concept that is stimulating the wide pattern of
experimentation we are observing. In this sense, it is important that the concept mean
something, but not something too specific. The ambiguity is a virtue (1994, 290).
Others are concerned as well. For example, Robert Trojanowicz who has written
extensively on community policing, adopts a stance that community policing is a new
philosophy and is not merely a bundle of tactics (Trojanowicz and others 1998,5,
emphasis added):
[I]t is crucial to understand that community policing is a new philosophy that
offers a coherent strategy that departments can use to guide them in making
the structural changes that allow the concept to become real. Community
policing is not just a tactic that can be applied to solve a particular problem,
one that can be abandoned once the goal is achieved. It implies a profound
difference in the way the police view their role and their relationship with the
community. Just adopting one or more tactics associated with community
policing is not enough.
Without a definition, however, we lack a firm starting point to evaluate living
examples of community policing. Bayley believes the result is that the success of
community policing will never be evaluatedit means too many different things to
different people (Bayley 1994, 278).

In an attempt to further elaborate upon some of these definitional issues, a brief
summary of the experience in New York City is presented as a case study. It
explores the realities of defining variations of community policing.
Community Policing in New York City
In July 1984, the New York City Police Department (NYPD) became one of the
first in the country to use a community-oriented patrol deployment strategy that
they called the Community Patrol Officer (CPO) Program (Farrell 1988,73).
Because NYPD is the largest police department in the countiy and one of the most
notorious, it serves as an important case study. NYPD has been notorious for both its
violence and corruption and has not been particularly well known for its innovation.
Nonetheless, its adoption of community policing and its implementation (referred to
as the CPO program) highlights many of the issues that continue to plague
community policing to this day.
The NYPD elected to use a specialized unit approach to community policing rather
than adopt the alternative. The other choice would have been to implement it
department-wide. Many departments opt for the specialized unit approach as an
incremental way to introduce the concept to the department, then later to adopt it
department-wide (Greene, Bergman, and McLaughlin 1994,103; Trojanowicz and
others 1998,17; Wycoff 1988,106). Some academics argue that unless the strategy is

implemented department-wide it is not really implemented (Trojanowicz and others
The NYPDs specialized unit approach called for ten officers and a sergeant to
cover a particular precinct. The CPO program that started with a single precinct was
deemed such a success by police administrators, citizens, and politicians, they
expanded it quickly. Within three years, sixty-four of the citys seventy-five precincts
were covered under the CPO program (Farrell 1988,74). Each of those precincts had
the same compliment of ten officers and a sergeant assigned to the unit. Yet, even
after three years, only 2.6% of the Departments officers were involved in community
NYPDs approach assigned nine of the ten CPOs in the unit to a relatively small
geographical area known as a beat. The tenth officer was reserved for
administrative duties. As a result, nine beats in each precinct were covered by CPOs.
Some precincts covered a larger geographical area than others and covering only nine
beats meant some beats in a precinct did not get a CPO (Farrell 1988, 81-2).
Assigning officers to specific geographical areas is one of the hallmarks of
community policing, at least from the academic perspective, enabling officers to get
to know the community.
The CPO was responsible for identifying crime and order maintenance problems
within their beat and working to finding solutions with community members and

other city agencies. The CPOs were specifically required to do three jobs to
accomplish that goal: First, they were to conduct foot patrol each day and cover their
entire beat on foot at least twice per week. Second, they were required to meet
regularly, but informally, with residents and business owners on their beat. Finally,
they worked with community groups on crime issues and, where none existed, they
were expected actively to organize them (Farrell 1988, 80).
The initial success of the program was measured by the positive reaction from
both citizens and police administrators. The citizens liked it because they had a
responsive contact in the police department whom they could call when they had a
problem. Apparently, many citizens believed this gave them a new found sense of
control over the quality of their neighborhood life. Police administrators liked it
because it gave them a direct method to address quickly order maintenance issues.
While order maintenance concerns generated a huge call volume of complaints both
to 911 operators and to the precinct commanders office, the commander had little
power to address them before this new program was implemented (Farrell 1988, 83).
However, the popularity of the program turned out to be a double-edged sword.
Once the program started to demonstrate its merit everyone wanted it, even though
NYPDs resources were limited. Many cities have struggled with selecting an
appropriate site for testing their approach to community policing (Lyons 1999,123-
32). Citizens or politicians in other cities heard of its success and soon desperately 2
2 According the FBI Uniform Crime Reports for 1987, there were 27,523 officers in the NYPD.

wanted to adopt it for their own. This sudden rush to enlarge the CPO program
everywhere put police administrators in an awkward political position since it forced
them to expand the CPO program quickly which, in turn, jeopardized the test itself
(Farrell 1988, 83).
Although many citizens, politicians, and precinct commanders wanted the CPO
program for their neighborhoods, constituents, or precincts, it still lacked then, and
continues to lack today, a methodology to accurately evaluate its effects. For
example, police officers, and therefore police departments, are traditionally judged on
factors such as the number of calls they answer, their response times, and the number
of arrests they make. None of these factors were relevant for officers engaged in foot
patrol with responsibilities for identifying and resolving order maintenance problems
in a specific geographical location (Farrell 1988, 85). Evaluating CPO programs
remains a challenge due to the absence of clear agreement on qualitative and
quantitative measures that some serious scholars believe will never be solved (Bayley
1994, 278).
The NYPD CPO program, furthermore, had good intentions to involve citizens in
the identification and resolution of local crime problems. However, the CPOs often
identified problems within a standard police context and addressed them using
traditional police tactics. So, while the community may have helped identify a
problem, its solution was left for police to solve using their old-fashioned methods
Roughly, 704 of them were assigned to the CPO Program and that represents 2.6%.

(Weisburd and McElroy 1988, 92). Existing community groups were often helpful,
but citizens outside such organizations often were unheard, even apathetic. Many of
them were so disorganized in their own lives that they had nothing left to give the
community (Weisburd and McElroy 1988, 100-1). Other agencies within the city
were also of little help. The typical CPO had to work hard to organize other outside
agencies response and cooperation then keep after them to continue to work with the
CPO program (Weisburd and McElroy 1988, 96). Involving the community remains,
like NYPD discovered, perhaps the biggest practical challenge in implementing
community policing (Friedman 1994, 263)
Another potential problem that the NYPD faced with its CPO program was
corruption. The NYPD has a long history of corruption and the common wisdom was
that officers that were close to the community were easier targets of corruption than
those that were professionally detached. As a result, the Department forbade
officers from entering business premises or private residences on other than police
business, and from engaging in unnecessary conversation with citizens (Farrell 1988,
86). This is exactly what CPO officers were required to do, so the NYPDs existing
policies actually contradicted a fundamental aspect of community policing (Farrell
1988, 86).
To address the potential problem of corruption or misconduct the NYPD
implemented four policies in connection with the CPO program. First, they carefully
screened the officers selected to serve as CPOs. Second, they charged all supervisors

with functional supervision of the CPOs instead of limiting it to only the CPO
supervisor. Third, the program was monitored by an inspection unit of the Chief of
Patrols office. Finally, several levels of supervisors were required to conduct weekly
and monthly interviews with residents and business owners in order to determine how
the CPOs were actually conducting themselves on their beats (Farrell 1988, 87).
These issues only highlight the fundamental differences between traditional
policing and community policing as well as point to the dilemmas in their resolution.
Traditional policing focused on law enforcement and with each law enforced the
officer generated a paper trail that allowed management to oversee an individual
officers work. For example, if an officer wrote a ticket it could be counted and
categorized. If the officer made a custodial arrest, that too could be counted and
categorized. The accompanying report could be reviewed for quality, and eventually
the courts would rule on the appropriateness of the officers actions. Even if only a
verbal warning was given, law enforcement management had a statistical summary
that could hold the police officer accountable.
Since community policing is more concerned with order maintenance than law
enforcement, accountability becomes more problematic. Officers assigned to
community policing need to be free of traditional responsibilities. Instead of
spending their time answering radio calls and making the arrests associated with those
calls, they need to spend their time with citizens identifying problems, prioritizing
them, and generating potential solutions. This is relatively free-flowing, creative

work that does not easily lend itself to a very specific work plan, quantitative data
generation, or close supervision. Most police departments across the country are
structured as para-military, hierarchical, bureaucratic organizations with many
layers of management designed to oversee the layer below. In Warren Bennis
vernacular police departments are very concerned with doing things right, but they
are not particularly adept at doing the right things, distinguishing between leaders
and managers, or differentiating between means and ends (Bennis and Nanus 1997,
Summary of the Practical Issues in Community Policing
From this case study we can see many of the issues that perplex police
administrators and scholars about community policing. One of the challenges is to
put these issues into a framework that facilitates analysis and understanding. Again,
Trevor Bennett suggests that the issues concerning community policing can be sorted
into categories: philosophy, principles, goals, organizational strategies, general
operational strategies, and specific operational strategies (1994, 229). Table 2.2
attempts to illustrate the relationship between Bennetts categories and NYPDs CPO
This table may clarify some of the confusion surrounding community policing.
The literature refers to many different issues in community policing but rarely
differentiates between these different categories. For example, the literature refers to

many different programs in different departments that are all implemented under the
rubric of community policing such as store-front police facilities, foot patrol, and
neighborhood surveys, and goes on to ask which is community policing? By looking
at the table above, we can see that each of these elements is a specific operational
strategy or a tactic that might be implemented in a move toward community policing
depending on the departments goals, but none of them alone could be termed,
community policing.
Table 2.2 Community Policing Philosophy and Practices for New York City
General Specific
Organizational Operational Operational
Philosophy Principles Goals Strategy Strategy Strategies
Consent Community Foot patrols,
Consultation Accountability Contact Informal meetings, Meetings with Community
groups, Establish
Community Decentralization community
Policing Problem Groups
Ideology Collaboration Identification
Problem solving Proactive Fixed beat
policing assignments, Problem- oriented Policing, Relationships with city agencies Policy changes
Table 2.2 also helps clarify the key problem with community policing theory. The
entry in each box represents an idea or application of community policing while the
categories themselves represent the theory. That is, the logic of community policing
should flow from left to right to reflect an integrated theory. There is plenty of
discussion in the literature about each box, but there is very little discussion about the

individual categories, how they fit together and, most importantly, about why these
should be the essential categories instead of others and whether or not they are
logically related.
Table 2.3 on the next page expands Table 2.2 from the implementation of
community policing in a single department to many other theoretical issues
concerning community policing. Note how these entries are entirely different than
those on Table 2.1 that described the professional model. Most striking about the
contrast between Tables 2.2 and 2.3 is that every department which chooses to adopt
community policing has numerous options since there remains no single definition of
community policing or a single defining characteristic. Instead there is a general
philosophy that, at this point, perhaps we think should move from left to right on this
table, without mandating specifically what should be done or how it should be done.
Therein lies the frustration. Within the context of community policing we do not
yet understand what it is, how it should be done, and why. The literature
concerning the practical considerations of community policing generally takes one of
two views.
First, police departments are embracing the idea of community policing and things
are happening. Departments are building storefront police facilities closer to the
neighborhoods they serve. They are devoting resources to community policing units
and they are asking residents for more input concerning problems in their
neighborhoods. Given the hierarchical, bureaucratic nature of police organizations,

some authors are simply impressed by the fact that anything is happening at all
(Skogan and Hartnett 1997, 237).
Table 23 Community Policing Philosophy and Practices Discussed in the Literature
General Specific
Organizational Operational Operational
Philosophy Principles Goals Strategy Strategy Strategies
Consultation Citizen Openness Community Establish
participation meetings community groups,
Consent Meet with community groups
Accountability Program Consultative
(Dept, to evaluation groups,
community) Policy changes Measure patrol
Collaboration Citizen participation Decentralization Participative Mgt. Policy changes
Openness Community contact Foot patrols
Problem solving Reorientation of Recruiting Fixed beat
patrol practices, assignments,
formal training, Relationships
Promotional with
practices other agencies
Specialized unit
Crime reduction Community crime Neighborhood
prevention watch,
Reduce fear property marking,
Reduce reactive security surveys,
Calls for service target hardening, victimization
Accountability Performance
(Officers to Dept.) systems
Community Problem-oriented Problem Decentralization Proactive policing Policy changes
policing policing identification
ideology Citizen Openness Community Foot patrols.
participation contact Informal meetings
Reduce reactive Alternative police Response by
Calls for service responses appointment, delayed response, walk-in reporting, mail-in reporting, telephone reports
Problem solving Reorientation of Participative mgt., Fixed beat
patrol recruiting assignments.
practices, Targeted patrol,
formal training, Relationships to
Specialized unit promotional practices other agencies

Table 23 (ConL) Community Policing Philosophy and Practices Discussed in the Literature
General Specific
Organizational Operational Operational
Philosophy Principles Goals Strategy Strategy Strategies
Crime reduction Community crime Neighborhood
prevention watch, property marking, security surveys, target hardening, victimization
Citizen surveys
Community satisfaction Fear reduction Foot patrols,
policing Accountability Performance surveys
ideology (Officers to Dept.) Job satisfaction system
Democracy Citizen Openness Community Consultative
participation meetings groups, Establish
Meaningful community
citizen groups,
input Meeting with
Consent community
Responsiveness groups
The second view is that the police are using the label, community policing, to
simply conduct business as usual. Many police officers have expressed that
community policing is nothing new. Rather it simply describes what the police have
done all along (Bayley 1988, 225). Lyons articulates this view clearly when he
suggests that community policing is really about managing citizen inputs by
constructing who the community the police serve will be, whose fears will become
most salient, and where control over problem-solving resources will be located
(1999, 184-5). Klockars also expressed this view when he suggested that the purpose
of the police is to control situations that emerge in society. Sometimes that requires
the use of force, including lethal force. Even though the right to use force is

fundamentally offensive to [the] core values of modem society... it is
unchangeable. Klockers believes the police must shield this purpose in a more
palatable form. Yet, community policing achieves this purpose by wrapping the
police in the powerful and unquestionably good images of community, cooperation,
and crime prevention even though the [p]olice can no more create communities or
solve the problems of urban anomie than they can be legalized into agents of the
courts or depoliticized into pure professionals (1988,257).
So we have come full circle. In the early history of policing, police departments as
night watchmen were responsive to the needs of those that controlled themthe
political bosses. Those bosses were responsive to their constituentsthe community,
therefore suggesting that police were, if only indirectly, responding to community
needs. The professional model sought to change this drastically in order to eliminate
corruption and political influence by removing the police from being responsive to the
community and requiring them to be responsive to professional codes, rules, and
the hierarchical authority of the police administrator. Consider, for example, NYPDs
proscription against engaging in unnecessary conversation with citizens.
Community policing suggests that we return to the old days of responsiveness to the
community and engage in necessary conversations with citizens. This time, however,
the police will be directly responsible to the community and indirectly responsive to
their administrators as well as politicians. Also early police possessed wide discretion
and chose which laws to enforce, whose peace to keep, which public to serveeven

whether or to exercise curb side justice instead of taking suspects to jail. The
officer on the beat largely decided those issues. To some extent, community policing
is a return to that old-fashioned night watchman style.
Finally, community policing focuses American police on crime prevention by
encouraging them to partner with the community to solve problems. The English
police focused also on crime prevention in 1829, yet the American police tended to
focus on it narrowly, to serve their own purposes. By the 1930s Americans developed
the professional model to focus on preventing the effects of crime. Now police have
come full circle and have started to understand the benefits of focusing on the goal of
crime prevention instead of strictly specifying the means to get there. Many of the
Peelian principles, outlined earlier, are now being embraced with fervor.
An Aperture in the Literature
The foregoing discussion focused, to a large extent, on describing what community
policing looks like today in practice, what it should look like in theory by discussing
best practices the experts argue implementing it, and now discuss evaluating it. The
problems associated with evaluating the success of community policing also emerge
in the literature such as using citizen surveys (Cordner 1988, 141; Lurigio and
Rosenbaum 1994,154; Skogan 1994,176; Wycoff and Skogan 1994, 83), trying to
gauge changes in levels of fear (Cordner 1988,141; Skogan 1994,176), or measuring
how much more job satisfaction officers assigned to community policing have over

their traditional counterparts (Cordner 1988,139-40; Lurigio and Rosenbaum 1994,
152; Wycoff 1988, 111; Wycoff and Skogan 1994, 82-3). However, these views of
the efficacy of community policing seem to miss an important point. Community
policing, as described in the literature, requires profound philosophical changes. We
need to evaluate whether or not such changes are being made and, if so, whether the
changes are having the desired effect on the police organizations. Chris Murphy
eluded to this problem when he wrote as community policing is an attempt to reform
both the philosophy and practice of contemporary policing, assessment of its impact
must address changes in police philosophy and ideology and its impact on actual
organizational or operational practice (1988, 183; emphasis added).
Gianakis and Davis began this evaluation process when they studied the effects
community policing had on organization structure and operating policies in Florida
police departments from the perspective of the top executive (Gianakis and Davis
1998) instead of simply describing the implementation process (Murphy 1988,185).
During their study they surveyed all the local law enforcement agencies in Florida
employing at least fifty sworn personnel. Their survey included police departments
with appointed heads and sheriffs departments with elected heads (1998,488).
The first step in their analysis was to distinguish between minimal and maximal
implementers. The minimal implementers reported that they had adopted the
community policing philosophy usually by adding a specialized unit, but they had not
made any effort to restructure their departments in the process. The maximal

implementers, on the other hand, had achieved some restructuring, most frequently
restructuring patrol operations and decentralizing into substations. They expected that
the maximal implementers would report more extensive organizational changes to
incorporate the new philosophy (1998,494).
Their results were somewhat predictable, given the elasticity of the community
policing concept. All but one of the 77.6% of those departments responding to the
survey claimed to be implementing community policing. Seventy-two percent of the
responders fell into the maximal implementer category while the remaining 28% were
minimal implementers. However, there was not much impact on structure or policies
regardless of whether the department was a minimal or maximal implementer. The
most frequent change these researchers found was in training. Many police
executives reported community policing simply requires changing the police officer
(1998,491). There were very few departments recording changes concerning the
involvement of the community in police operations. Thus Gianakis and Davis
commented The public is polled, but kept at arms length from the formal
organization (1998,496-7). Gianakis and Davis concluded that It does not appear
that the essential elements of community policing as described in the research
literature are finding their way into the Florida programs (1998,496). Is it too soon
to expect much change? And is this phenomena that these two researchers discovered
consistent in departments across the country?

Gianakis and Davis work right now is worthy of expanding to a national study for
four reasons. First, their study appears to be headed in the right direction and worthy
of emulating since what departments are doing, rather than what they are saying, may
be the most important part of evaluating community policings impact (Murphy 1988,
Second, Gianakis and Davis studied only police departments in Florida. While
those departments may adequately represent the entire country, that supposition needs
to be tested nationwide.
Third, we need to begin to explain why particular departments are doing what they
are doing. Many police departments are implementing what they are calling
community policing, but it may be that they are incorporating this new philosophy
or strategy very narrowly within their existing traditional philosophies and
organizational structures. For example, many departments may remain committed to
crime control and rapid response to calls for service as their basic operational
strategy, but they simply add community policing strategies to a specialized unit that
remains separate, physically and philosophically, from the heart of their traditional
approach to policing. Bennett refers to this process as bolted on community
policing to emphasize the point that this approach misses the underlying premise that
community policing should become an overriding philosophy department-wide not
simply a new program (1994,234). Lack of the management skills and sophistication

that is necessary to facilitate the fundamental organizational change community
policing requires may well be the cause (Murphy 1988,183).
Finally, and perhaps most importantly, we need to get a sense of what top police
executives think about community policing and its underlying tenets. Skolnick and
Bayley suggest that in order to innovate successfully police chiefs must believe in
what they are doing. Their leadership is particularly important because of the para-
military character of police agencies throughout this country (1986, 220).
The chief is in the best position to observe and decide on what is being
implemented under the rubric of community policing, what kinds of problems are
being encountered, where support for community policing is coming from, and
evaluate why particular departments have chosen to implement community policing
in the way they have.
In the next chapter we will discuss using Gianakis and Davis approach combined
with the information in Table 2.3 as an ideal type model to evaluate organizational
change in police departments across the country that are implementing community
policing. Hopefully, such a comparison will add to our understanding of how
community policing is defined today and actually being practiced.

Science is built up with facts, as a house is with stones. But a collection of facts is no more
a science than a heap of stones is a house.
Jules Henri Poincare
The purpose of this research is to explore the effects of community policing on
police departments policies, structure, and officer responsibilities, and to examine the
departments top executives attitudes concerning the underlying tenets of community
policing. The first task is challenging due to the lack of an agreed upon definition of
community policing. Without a clear definition we can not authoritatively prescribe
the best practices for community policing programs, policies, structure, or attitudes.
We can, however, look to see if there are changes. Depending upon what we find, we
can then make some tentative judgments as to whether these changes seem to support
community policing or seem to contradict the concept.
Research Questions and Hypotheses
This study will address six primary questions concerning police departments in the
United States and their experience with community policing:

Research Question #1
What are police departments around the country implementing under the heading of
community policing and does their implementation vary by department demographics
such as department type (police or sheriff), department size, accreditation status, and
commitment to community policing (minimal or maximal implementer)?
Null Hypotheses3 #1
There are not significant differences in what is being implemented as community
policing by different kinds of departments.
Rationale for Hypotheses #1
The rationale for this research question is that community policing can mean different
things to different people (Bayley 1994,278; Kelling and Moore 1988, 22; Murphy
1988, 178; and Wilkinson and Rosenbaum 1994, 120; Wycoff 1988,103-4). As a
result, we would expect to see some variation in what is being implemented, but we
need to be sure that any variation we see in the sample actually represents what is
going on in the population of police agencies throughout the country. Therefore, this
null hypothesis, that we hope to reject, is that there is no difference.
Research Question #2
Has the implementation of community policing significantly impacted the policies,
structure, or the responsibilities of patrol officers in local law enforcement agencies?
3 Null hypothesis is a statistical term referring to the assumption that the relationship found in the
sample data is different than what exists in the population. Said in a different way, this means that our
sample of police agencies and their top executives may not be representative of the underlying
population of police agencies with more than fifty officers and their heads. The benefit of statistical
analysis is that it allows us to quantify the chance that the sample is not representative of the
underlying population. So, we state our hypotheses as null hypotheses and then try to reject them.
If we cannot reject the null hypotheses then we are saying that we are not confident that the sample
results are consistent with what is really going on in the population. On the other hand, if we can
reject the null hypothesis that means we are confident that the results from our sample are consistent
with what is going on in the population. On one level, that is the rationale for each of the six null
hypotheses, although it will not be repeated in the sections that follow.

Null Hypotheses #2
There is no significant impact on the policies, structures, or responsibilities of patrol
officers in departments that implement community policing.
Rationale for Hypotheses #2
Community policing is seen as a fundamental change to the way the police conduct
their business (Community Policing Consortium 2000,134; Manning 1988, 31-2;
and Skogan 1994,167). So, it seems reasonable to expect that those changes would
manifest themselves in changes to policies, structure, or responsibilities of patrol
officers. The null hypothesis suggests that this change has not yet taken place.
Research Question #3
From the perspective of the top police administrator, is there a difference in the level
of support for community policing between citizens and police employees?
Null Hypotheses #3
From the perspective of the top police administrator, there has not been a significant
difference in the level of support for community policing between citizens and police
Rationale for Hypotheses #3
The purpose of this hypothesis is to get a sense of commitment from outside the
executive suite to community policing. It has been argued for example, that police
culture is very closed and resistant to change (Kelling and Bratton 2000, 258 and
Reuss-Iianni, 1993, Skolnick and Bayley 1986,211). If that is true in this case it may
have important ramifications for the potential success of community policing, the
directions it may take, or the limits of its practicality.
Research Question #4
From the perspective of the top police administrator, are there still significant
organizational problems associated with the implementation of community policing?

Null Hypotheses #4
From the perspective of the top police administrator, there are not still significant
organizational problems associated with the implementation of community policing.
Rationale for Hypotheses #4
Because community policing is such a fundamental shift for American policing, it
still seems necessary to gauge the progress of its implementation by getting a sense of
the problems being encountered. To date, most of the work in this area has been by
case study and not by looking at the experience of police agencies all around the
country (Lyons 1999; Skogan and Hartnett 1997) and this question hopes to address
that shortcoming.
Research Question #5
Do police executives who head departments that have adopted community policing
differ in their opinions concerning citizen participation, citizen oversight, command
devolution, and problem solving from those executives who have not implemented it
and do views vary based on department demographics?
Null Hypotheses #5
There is no significant difference in the opinions of police executives who head
departments that have adopted community policing concerning citizen participation,
citizen oversight, command devolution, and problem solving from those executives
who have not implemented it and views do not vary based on department
Rationale for Hypotheses #5
The rationale for this hypothesis is simply to start searching for possible explanations
in the variety of approaches to community policing. This part of the research is
exploratory and this hypothesis and the next one are due to that exploratory nature.

Research Question #6
Are there certain external factors (such as a unionized workforce) or personal
characteristics (such as educational level) that help to predict which top police
executives will be successful in their implementation of community policing?
Null Hypotheses #6
There are not external factors or personal characteristics that help to predict which top
police executives will be successful in their implementation of community policing.
Rationale for Hypotheses #6
The rationale for this hypothesis is the same as that for the previous one; we are
simply searching for relationships among the data since so many different data
elements were collected during the study.
Addressing six research questions in a single project appears daunting, but it is
appropriate in this case because of the exploratory nature of this research. Also
because the first five questions share department demographics as their independent
variables, the research is easier than it may first appear. The sixth question adds
additional independent variables, but does so in a manageable way.
Each of these hypotheses will be analyzed statistically allowing a precise
description of the confidence in the results. The results of the research will either
allow us to retain or reject the null hypothesis for each research question. Retaining
the null hypothesis means that there is no statistically significant difference between
the defined groups, while rejecting the null hypothesis means there is a difference.
Research is almost always more interesting when it finds differences (that is, can

reject the null hypothesis), but statistical analysis must start with the null and be
prepared to show why it can or cannot be rejected. There is always the possibility of
error in this process. We may reject the null hypothesis when in fact it is true or we
may retain it when it is false. These errors, however, are what the statistical analysis
will allow us to quantify. This research will use a 95% confidence level. Therefore,
assuming our sample is large enough, that is enough of the surveys are returned, there
is only a 5% chance that our final choice to retain or reject each null hypothesis is
Statistical significance is what allows us to retain or reject the null hypothesis and
is determined by using sophisticated statistical tests or analysis. Equally important
however, is the concept of substantive significance. Using statistics small differences
between groups can be statistically significant. However, some of these small
differences may not be of any practical value. In this sense, it is possible that one or
more of the null hypotheses may be rejected statistically, but contain such small
differences as to be of little or no use in the real world.
Quantitative Research Design
To answer the six research questions, the research design compared the views of
randomly selected top police executives throughout the United States on a variety of
topics related to community policing. Following the approach employed by Gianakis
and Davis, only departments with more than fifty sworn employees (excluding

corrections and court protection personnel) were included in the sample frame. The
University of Colorado, at Denver, Human Subjects Research Committee approved
this project.
The primary data collection method in this research was a survey. The survey
instrument used by Gianakis and Davis was adapted to this project with their
permission (see Appendix A). The survey data was augmented with a qualitative
approach using interview data from police executives who agreed to be interviewed
by telephone. The thirteen police executives who were interviewed responded to a
series of open-ended questions to explore more in-depth their experiences and views
concerning community policing. This portion of the research is discussed in more
detail later in this chapter.
The study design was comparative, descriptive, and exploratory. Again, the
research design followed closely that used by Gianakis and Davis. Their choice of
independent variables was mimicked and included the type of department (police or
sheriff), the size of department (147 is the cut off between large and small), the
departments accreditation status, and the departments commitment to community
policing (minimal implementers or maximal implementers). These variables will be
referred to as department demographics from this point forward to facilitate
Table 3.1 summarizes the variables for each of the research questions. Department
demographics are used in each case as the independent variable. The survey

Table 3.1 Research Questions and Variables
Research Question
1. What are police departments around the country implementing
under the heading of community policing and does their
implementation vary by department demographics?
2. Has the implementation of community policing significantly
impacted the policies, structure or the responsibilities of patrol
officers in local law enforcement agencies?
3. From the perspective of the top police administrator, is there a
difference in the level of support for community policing between
citizens and police employees?
4. From the perspective of the top police administrator, are there still
significant organizational problems associated with the
implementation of community policing?
5. Do police executives that head departments that have adopted
community policing differ in their opinions concerning citizen
participation, citizen oversight, command devolution, and problem
solving from those executives that have not implemented it and do
views vary based on department demographics?
6. Are there certain external factors (such as a unionized workforce)
or personal characteristics (such as educational level) that help to
predict which top police executives will be successful in their
implementation of community policing?
Department type
Department size
Accreditation status
Commitment to CP
Department type
Department size
Accreditation status
Commitment to CP
Department type
Department size
Accreditation status
Commitment to CP
Department type
Department size
Accreditation status
Commitment to CP
Department type
Department size
Accreditation status
Commitment to CP
Implementation Status
Elements of Comm. Policing strategy
Changes in policies and procedures
Changes in structure
Changes in officer responsibilities
Police employee support
Police manager support
Citizen support
Non-police manager support
Perception of organizational problems
Opinions on citizen participation
Opinions on citizen oversight
Opinions on command devolution
Opinions on problem solving
Department type Success in implementing CP
Department size Success with CP in community
Accreditation status Projected future success with CP
Commitment to CP
Personal characteristics

Table 3.2 Research Questions and Associated Survey Items________
Research Question
Survey Question
1. What are police departments around the country implementing under the heading of community policing and does their
implementation vary by department demographics?
Subscale #1.1 Elements of community policing strategy
2. Has the implementation of community policing significantly impacted the policies, structure or the responsibilities of patrol
officers in local law enforcement agencies?
Subscale #2.1 Changes to policies
Subscale #2.2 Changes to structure
Subscale #2.3 Changes to officer responsibilities
Subscale #2.4 Goals associated with community policing
3. From the perspective of the top police administrator, is there a difference in the level of support for community policing
between citizens and police employees?
Subscale #3.1 Police employee support for community policing
Subscale #3.2 Police managers support for community policing
Subscale #3.3 Citizen support for community policing
Subscale #3.4 Non-police governmental officials support for community policing
4. From the perspective of the top police administrator, are there still significant organizational problems associated with the
implementation of community policing?
Subscale #4.1 Perception of organizational problems
5. Do police executives that head departments that have adopted community policing differ in their opinions concerning
citizen participation, citizen oversight, command devolution, and problem solving from those executives that have not
implemented it and do views vary based on department demographics?
Subscale #5.1 Top executives opinions concerning citizen participation
Subscale #5.2 Top executives opinions concerning citizen oversight
Subscale #5.3 Top executives opinions concerning command devolution
Subscale #5.4 Top executives opinions concerning problem solving
6. Are there certain external factors (such as a unionized workforce) or personal characteristics (such as educational level) that
help to predict which top police executives will be successful in their implementation of community policing?
Subscale #6.1 Top executives perception of departments success in implementing community policing
Subscale #6.2 Top executives perception of departments success of community policing in the community
_______________Subscale #6.3 Top executives perception of department's future success in community policing__________________
Question 8; 24 items
Question 9; 16 items
Questions 3, 8
Question 10; 15 items
Question 7; 20 items
Question 13; 3 items
Question 13; 2 items
Question 13; 5 items
Question 13; 4 items
Question 14; 17 items
Question 15; 6 items
Question 15; 3 items
Question 16; 7 items
Question 17; 9 items
Question 4
Question 5
Question 6

Table 3.3 Research Questions and Proposed Statistical Analysis
Research Question Statistical Test
1. What are police departments around the country implementing under the heading of community policing and does their implementation vary by department demographics? ANOVA
2. Has the implementation of community policing significantly impacted the policies, structure or the responsibilities of patrol officers in local law enforcement agencies? ANOVA
3. From the perspective of the top police administrator, is there a difference in the level of support for community policing between citizens and police employees? ANOVA
4. From the perspective of the top police administrator, are there still significant organizational problems associated with the implementation of community policing? ANOVA
5. Do police executives who head departments that have adopted community policing differ in their opinions concerning citizen participation, citizen oversight, command devolution, and problem solving from those executives who have not implemented it and do views vary based on department demographics? ANOVA
6. Are there certain external factors (such as a unionized workforce) or personal characteristics (such as educational level) that help to predict which top police executives will be successful in their implementation of community policing? ANOVA and Multiple Regression
instrument was designed to capture the information on each of these variables. Table
3.2 summarizes how the data on the dependent variables was captured using the
survey instrument. Finally, Table 3.3 summarizes the statistics that will be used to

make comparisons. ANOVA, which is an acronym for analysis of variance, breaks
the variation in the dependent variable into two components; the variation within the
group and the variation between the groups (Fox 1992,174). Using this type of
analysis we can evaluate differences between police departments in different groups.
For example, we can look to see if there is a statistically significant difference
between the elements of community policing adopted by police departments versus
those adopted by sheriffs departments.
Each of the dependent variables in the research questions comes from the literature
on community policing. Unfortunately, as chapter two outlines, the theory underlying
community policing is undevelopedor in the process of developingso reliance on
the literature for these variables is speculative, or exploratory, at best instead of
rigidly defensible on theoretical grounds.
The first research question confronts this problem head-on. The literature, as we
have seen, clearly articulates that community policing means different things to
different people (Bayley 1994, 278; Kelling and Moore 1988,22; Murphy 1988,178;
Wilkinson and Rosenbaum 1994,120; Wycoff 1988,103-4). As a result, one of the
priorities for the research community is to sort out what it means to whom.
Therefore, the first question tries to get a handle on the variety of tactics that top
police executives are comfortable referring to as community policing.
The second research question picks up on one of the more recent themes in the
community policing literature. That is, due to the fundamental difference between the

professional model and the community based model, we should be seeing changes in
the way police departments that adopt the community based model conduct business
(Gianakis and Davis, 1998). Some changes should manifest themselves by reforms in
policies, structure, and police officer responsibilities (Community Policing
Consortium 2000,134; Manning 1988, 31-2; Murphy 1988; Skogan 1994,167).
The third research question acknowledges indirectly that change is often difficult
and especially more difficult in an environment opposed to it (Kelling and Bratton
2000, 258; Reuss-Iianni 1993). On the other hand, support for change can be
interpreted as either intrinsic support for the idea or strong leadership encouraging
different interest groups to get on board (Peters and Waterman 1982, 102).
The fourth research question is also from Gianakis and Davis survey and tries to
sense what police executives are facing in the way of challenges to see if that sheds
any light on how they function. The literature addresses the problem with freeing up
resources to re-deploy them to community policing (Kennedy 2000,224) and this
question will give us a sense of whether police executives are making progress on this
front. Again, most of the exploration in this area has been through case studies and
not by way of a nationwide survey.
The fifth research question tries to determine how committed police executives are
to some of the basic tenets that underlie community policing. As chapter two
suggested, the literature documents that most departments are committed to
community policing (Gianakis and Davis 1998,494). However, it also points up that

police officers and detectives are more skeptical than their command staffs toward the
idea (Lewis, Rosenberg, and Sigler 1999, 577). Is this skepticism because more time
is needed to convince line level personnel of its merits? Or are police executives
saying one thing and doing another and therefore, confusing the rank and file? This
issue is not addressed in the literature. This fifth research question explores this issue
by assuming that citizen participation, citizen oversight, command devolution, and
problem solving are the basic tenets that underlie the community policing philosophy
and, then in turn, looking to see how supportive the executives are of these tenets
(Leighton 1994, 218; Meese 2000, 299-300, 304; Goldstein 1990).
Finally, the sixth research question provides a mechanism to explore relationships
between the data collected to answer the other questions involving the successful
implementation of community policing. For purposes of this question, successful
implementation of community policing is defined as the executives perception of his
or her success based on three separate questions in the survey. The first asked them to
evaluate the success of the departments implementation efforts. The second sought
to gauge the effects of community policing in their community and the third, to
quantify their anticipated future success in the community.
Sample Characteristics and Sampling Strategy
The sampling strategy used in this research followed that used by Gianakis and
Davis. They included all local law enforcement agencies in the state of Florida that

employed at least fifty sworn personnel (1998,488). This study also included only
agencies that employed at least fifty sworn personnel. However, it expanded the
scope by including agencies from the entire country instead of limiting the survey to a
single state. In contrast to Gianakis and Davis, this research used a sampling
approach instead of surveying every department that met the criteria. The sample was
randomly selected. The sampling frame was the 1999 edition of the National
Directory of Law Enforcement Administrators published by the National Public
Safety Information Bureau. Of the 13,533 local police and sheriffs departments
1,827 of them met the criteria for inclusion in the study. From this group, 557 were
randomly selected and their chief executive received the survey questionnaire.
The sample size of 557 was calculated using the formula suggested by McCall
(1982, 192) assuming a 95% confidence level and a tolerable error of 5%. That
sample size was adjusted to reflect a projected 70% response rate. The study hoped
for 383 completed surveys and received 388.
Although desirable to stratify the sample to ensure enough variation in responses to
support the proposed statistical analysis, that information did not exist. For example,
stratifying police departments between those that have implemented community
policing and those that have not would have ensured the ability to make comparisons
between these two groups. However, no information exists on which departments
implemented community policing. In addition, using Gianakis and Davis experience
as a guide, nearly all departments claim to have implemented it (1998,494).

Therefore, these two groups may not even exist. Another example is related to the
problem of distinguishing between minimal implementers (those that have only
adopted the philosophy, but have not made any structural changes) and maximal
implementers (those that have made structural changes). Again, this differentiation
was impossible to discover because the information did not exist until after the
respondents answered that question on the survey. Both groups were defined by the
survey results. Yet if they had been known before hand, groups of minimal and
maximal implementers could then have been represented in other subsets of the data
such as size, type, and accredited status.
Pilot Study and Instrument Development
Gianakis and Davis developed a survey instrument to conduct their study. They
pilot tested it with police executives in one police department and one sheriffs
department (1998,488). They gave the researcher in this project permission to use
their instrument. The instrument used in this project drew heavily from that
instrument, but it was modified to meet the needs of this project. As a result, the new
instrument was pilot tested again.
The pilot test consisted of six police executives completing the survey and
answering a series of questions concerning the survey itself. Of these six executives,
three were from police departments and three were from sheriffs departments. The
suggestions for improvement from the pilot test were incorporated into the final

survey instrument before it was approved by the University of Colorado, at Denver,
Human Subjects Research Committee.
Reliability Testing
There are two types of errors possible in survey research: random error and
measurement error. Random error occurs in all research relying on sampling. Those
picked to participate in the research may not represent the population from which they
were chosen. This type of error is best addressed by large sample sizes and statistical
analysis that quantifies the possibility that the surveys results occurs by chance
instead of due to real differences among survey respondents (Litwin 1995, 5). In this
project problems of random error are prevented by those methods. The sample size
was calculated statistically, adjusted to account for the projected response rate, and
the results were analyzed statistically to quantify the possibility they occurred by
Measurement error refers to how well an instrument measures what it is supposed
to measure. Consider, for example, a wooden yardstick compared to a plastic tape
measure. The wooden yardstick simply needs to laid on the table to measure fabric
while the plastic tape measure needs to be stretched along the fabric. Different people
will stretch the plastic tape measure in different ways resulting in different
measurements than those obtained using the wooden yardstick. Neither instrument is

perfect, as both are subject to error, but the plastic tape measure produces more
measurement error than the wooden yardstick (Litwin 1995,7).
In this research measurement error is a real possibility because it attempts to
measure attitudes (concerning citizen participation, citizen oversight, command
devolution, and problem solving, for example) and behaviors (which community
policing strategies have been implemented) not concrete units such as inches or
weight. To minimize measurement error in this research the survey tried to measure
the same attitude several times. For example, there were six questions concerning the
chief executives attitude toward citizen participation, three related to citizen
oversight, seven involved command devolution, and nine on problem solving. In
general, a measures reliability increases as the number of questions increases
(OSullivan and Rassel 1989, 84).
Reliability itself is a statistical measure of how reproducible the survey
instruments data are (Litwin 1995, 6). Reliability comes in three forms: test-retest,
alternate-form, and internal consistency. Test-retest reliability requires the same
groups be tested at two different times and that was not part of this projects design.
Alternate-form reliability requires that questions be asked multiple times in the same
survey, but in different forms. In this project the survey was too short and focused to
accommodate this requirement. As a result, the form of reliability relied upon in this
project was internal consistency.

This study uses Cronbachs coefficient alpha to measure internal consistency of the
surveys subscales. It is a reflection of how well the different items complement
each other in their measurement of different aspects of the same variable or quantity
(Litwin 1995,24). When Cronbachs coefficient alpha is 0.70 or greater, it is
generally accepted that there is good reliability between the items (Litwin 1995, 31).
Table 3.4 summarizes the Cronbachs coefficient alpha calculations for the pilot
survey data. The calculations for the final instruments calculations are discussed in
the next chapter along with all the other results from the survey. All values are near
or above .70 suggesting good reliability between the items in this survey.
Table 3.4 Cronbachs Coefficient Alpha Results for Pilot Survey
Survey Cronbachs
Question Coefficient
Number Question Topic Alpha
15 Chief executives view of citizen participation .7324
15 Chief executives view of citizen oversight .9512
16 Chief executives view of command devolution .7934
17 _______Chief executives view of problem solving______________.6824
Validity Testing
Validity concerns how well [the survey instrument] measures what it sets out to
measure (Litwinl995,33). Returning to the yardstick example above, in order to be
valid the wooden yardstick and the plastic tape measure must both be thirty-six inches
long. The wooden yardstick is reliable and can be used consistently each time. Yet if

it is really thirty-nine inches long, the yardstick does not measure thirty-six inches,
truly. In other words, reliability is concerned with consistency and validity is
concerned with truth (Litwin 1995,33-4).
The form of validity here involves content validity. Does the survey include what
it should include and exclude what it should exclude to allow for the data collection
necessary to answer the research questions? Unfortunately, there is not a statistic
analogous to Cronbachs coefficient alpha to allow us to precisely measure the answer
to this question. Instead we must rely on the judgment of the researcher that
developed the survey instrument together with others that reviewed it and offered
their opinions.
The survey instrument was subjected to three phases of review. First, the survey
was reviewed by three different professors, two of which spend much of their time
teaching and consulting on matters relating to research methods in general and survey
methods in particular. Two of the three professors were asked to complete a written
evaluation of the survey instrument (see Appendix B).
Second, the survey was submitted to law enforcement professionals, including
chiefs, and police researchers to get a viewpoint closer to the surveys topic. Six of
these people reviewed the survey. Two of them are chiefs, one a deputy-chief, and
three are researchers. One of the researchers is a retired chief.
Third, the survey was pilot tested. Five police executives in the state of Colorado
and one from the state of Washington completed the pilot test survey. Finally, those

that piloted the survey were also asked to complete an evaluation of it (see Appendix
The feedback, received from all three sets of reviewers suggested that the survey
would be effective in providing the data needed to answer the research questions.
There were some excellent suggestions for improvement and those were incorporated
into the final version of the survey.
Qualitative Research Design
The survey collected quantitative data concerning community policing, but that
methodology necessarily leaves gaps in understanding the executives thought
processes regarding the how and why of their approach to community policing. In
order to gain further insights into such issues, this research design added qualitative
methods (Creswell 1994, 145; Lincoln and Guba 1985, 259; Strauss and Corbin 1998,
11). Thirteen police executives were chosen for half-hour telephone interviews.
The police executives who participated in the telephone interview portion of this
research were selected from those that participated in the survey research. One of the
questions in the survey asked participants whether or not they would be willing to
participate in the telephone interview process. From the 162 that agreed, thirteen
were selected. Table 3.5 describes the executives who were interviewed. The
participants were not randomly selected, but were judgmentally chosen not to

Table 3.5 Summary of Telephone Interview Participant Characteristics
Department Type Size (# of Officers) Level of Implementation Sex
Title Police Sheriff 50-146 147+ Minimal Maximal None Male Female Age Race Education
Chief A X X X X 51 White Bachelors
Chief B X X X X 54 Asian Masters
Chief C X X X X 58 White Bachelors
Chief D X X X X 52 White Masters
Chief E X X X X 47 White Bachelors
Chief F X X X X 46 White Masters
Chief G X X X X 46 Hispanic Masters
Sheriff H X X X X 49 White Bachelors
Chief I X X X X 51 White Masters
Chief J X X X X 51 White Masters
Sheriff K X X X X 72 White Some College, No Degree
Chief L X X X X 53 White Bachelors
Chief M X X X X 56 White Some College, No Degree
Totals 11 2 11 2 3 8 2 12 1 52.8

represent the population of police agencies but the variation among departments for
obtaining as much different information as possible (Lincoln and Guba 1985,265).
Table 3.5 indicated eleven police chiefs and two sheriffs were interviewed. Two
chiefs were from large departments, while the other eleven were from small ones.
Three minimal implementers, eight maximal implementers, and two non-
implementers were interviewed. One of the thirteen executives was a woman and, on
average, the telephone interviewees were 52.8 years of age.
The design of this portion of the research was necessarily broad and recognized the
unpredictability of what would arise (Lincoln and Guba 1985,259). Three primary
components to qualitative research include: data, procedures to interpret and organize
the data, and written or verbal reports (Strauss and Corbin 1998,11-2). In this case
they came from interviews that were tape-recorded and then transcribed.
The interview questions (see Appendix D) were developed based on a preliminary
review of the survey data. This review was conducted after receiving approximately
250 of the surveys. Although the interview questions were used as a guide, the
conversation with individual executives flowed in whatever direction the executives
responses took (Lincoln and Guba 1985,41; Rubin and Rubin 1995,2).
The procedures used to interpret and organize the data, including conceptualizing,
reducing, elaborating and relating, are referred to as coding (Strauss and Corbin 1998,
12). In this case that process was completed with the help of a computer using
NUDIST (Non-numerical Unstructured Data, Indexing Searching and Theorizing)

software. Some 111 categories were generated and roughly ninety-six pages of text
were categorized into those categories. Those categories were summarized into
seventeen groups and then memos were prepared for each group (Lincoln and Guba
1985,42). Those memos formed the basis for the results outlined in chapter 5.
One of major challenges of qualitative research in Lincoln and Gubas words is
establishing its trustworthiness, that is, persuading audiences that the findings of
an inquiry are worth paying attention to (1985, 290). In quantitative research the
issues effecting trustworthiness are internal validity, external validity, reliability, and
objectivity. Lincoln and Guba suggest the anomalous terms for qualitative research
are credibility, transferability, dependability, and confirmability (1985, 300).
To achieve credibility they suggest prolonged engagement, persistent observation,
triangulation, and member checks (1985, 301). In this case the observer was a police
officer for six years and a police supervisor for three years suggesting prolonged
engagement that permitted ample time to learn the culture. Persistent observation
in this case was achieved only by completing thirteen different interviews. That was
not a magic number, but rather just a workable quantity providing enough exposure to
achieve interesting findings.
Triangulation suggests obtaining information from different sources, methods, or
investigators in order to find similarities (1985, 307) and in this case to the extent that
participating police executives (different sources) agreed with one another
triangulation occurred. By submitting chapter five in its entirety to all thirteen