I ll ‘1"?
in. .3331.
. .1. Lu . ‘rr .

t.

: fir, ~
J. guy?
Hz.

imp“.
# i,
.rflfidfiul.

fit.
I .s

.lev!
.l 5’... I
19.1.» afin¥€€
.wéin... L. “humus.
» e.uw... JW
l. 1 u.
a...

.
in.)
II '0

:5 Sn
V ' ' 43...,
1. lily-F1 A:
(.1 .: t.
‘ .2
5

iv. . .

2...: a. .. .
V .0?)
firwnawfit ,4?
I . Yin Ezuooiliniit
$3.,ieh... . I?!
.unammwuatx...
:2 . z s
.fivflzty 29
. t. .5.
2c 1.
9‘ ll 3...;
. . iii...
.3. a...

lair .. I,
“nah...“fi
x!!.)|43\.-.olnn .“u in;
I! :Li3_§ .- .flt.
25.3.] 5!...[113‘3 b!»

Xv:- . Ivfi‘t‘ll‘

 

Thesis

2- llllllllllllllllllllllllliillllllll.

3 1293 01688 0548

This is to certify that the

dissertation entitled

The Impact of Campus Public Safety Operations and Crime
Prevention Program Implementation upon Serious Criminality;
at U.S. Colleges and Universities : An Application of the
Situational Perspective

presented by

Donald C. Hummer, ll

 

has been accepted towards fulfillment I
of the requirements for . - ’

Ph.D degreein Sociai Science - Crim. Jus.

 

 

N' Kw; M

Major professor

 

Date 613'“, 'qq?

 

MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771

PLACE m
to remove this
TO AVOID FINES re

 

 

Michigan State

LIBRARY
Unlverslty

‘7

 

 

V

RETURN BOX
checkout from your record.
turn on or befor

e date due.

 

MTE DUE

DATE DUE

DATE DUE

 

{A519

 

serfs 1§99

 

 

 

 

 

 

 

 

 

 

 

use Wm“

THE IMPACT OF CAMPUS PUBLIC SAFETY OPERATIONS AND CRIME
PREVENTION PROGRAM IMPLEMENTATION UPON SERIOUS CRIMINALITY
AT U.S. COLLEGES AND UNIVERSITIES : AN APPLICATION OF THE
SITUATIONAL PERSPECTIVE

By

Donald C. Hummer, II

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

College of Social Sciences

1998

ABSTRACT
THE IMPACT OF CAMPUS PUBLIC SAFETY OPERATIONS AND CRIME
PREVENTION PROGRAM IMPLEMENTATION UPON SERIOUS CRIMINALITY
AT U.S. COLLEGES AND UNIVERSITIES : AN APPLICATION OF THE
SITUATIONAL PERSPECTIVE
By
Donald C. Hummer, II
The study which follows was an attempt to apply the emerging criminological

perspective of Situational Crime Prevention to incidences of serious violent and property
crime at larger colleges and universities in the United States. The Situational perspective
is an integrated theoretical approach which, in contrast to most Positive criminological
thought, assumes potential offenders to be a part of society and thus emphasizes strategies
to deter these offenders from carrying out certain acts of criminality. As rates of serious
crime increased on campus throughout the 1980s and early 19905, university
administrators deliberated on and implemented a variety of crime prevention strategies to
offset this rising trend. Also entering into the mix is the U.S. Congress’ passage of the
Student-Right-To-Know and Campus Security Act of 1990 which mandates the gathering
and publication of crime figures and crime prevention policies for campuses receiving
Federal monies. These factors have led to an emphasis on prevention and law
enforcement in the collegiate environment. As such, campus public safety departments

have begun to mirror state and municipal law enforcement agencies in selectivity of new

oflicers, organization, and duties and functions.

This dissertation utilizes data collected by the Bureau of Justice Statistics’ (BJ S)
Law Enforcement and Management and Administrative Statistics (LEMAS) program,
which administered a questionnaire to larger (enrollment > 2500 students) colleges and
universities throughout the United States. The primary focus of the original B] S study
was to ascertain the structure and functions of campus public safety departments,
however data were also collected on a number of variables indicative of the tenants of
Situational Crime Prevention, as well as data on serious (Part I) offending at these
schools during the calendar year 1994.

Using the data from the LEMAS questionnaire, this dissertation examines whether
crime prevention initiatives derived fiom the Situational perspective are successful in
ameliorating serious offending in the campus environment. The data were analyzed in a
multivariate format utilizing ordinary least squares regression, and a host of factors
ranging from specific crime prevention programs to campus public safety department
operations and institutional demographic characteristics were examined as to their impact
upon crime statistics at the 680 schools surveyed.

Results fi'om the analyses are discussed, as are possible alternative explanations
for the findings. Recommendations are put forth regarding administrative procedures and
policies which could be implemented to ensure student safety. Directions for future

research into this topic are also discussed.

Capyright by
DONALD CHARLES HUMMER, II
1998

ACKNOWLEDGMENTS

Completing a doctorate is an endeavor that requires the assistance and support of
many people. I would first like to thank the entire faculties at both Michigan State
University and Shippensburg University. Every professor at these institutions has been
helpful and encouraging in some way throughout my tenure as a graduate student.
Specifically, I would like to thank Thomas Austin, Mahesh Nalla, and Donna Hale; one
could not ask for better mentors or fi'iends. Thanks also to Vincent Hoffman, Jacob
Climo, and Alan Posner for their work as members of my guidance/dissertation
committee.

I would also like to thank my family, extended and nuclear, living and deceased,
for their encouragement, support, and simply being a connection to reality when it felt
like I was being immersed in the academic life. I wish to thank my mother for her love
and hard work that helped to send me through ten years of college. Last, but certainly not

least, I want to thank my wife, Beth, for her love and sacrifices.

TABLE OF CONTENTS

Page
LIST OF TABLES .................................................................................................. ix
Chapter
I. INTRODUCTION ................................................................................ 1
The Context of Campus Crime ........................................................ 1
The Campus Crime and Security Act of 1990 ................................. 3
The Impact of, and Response to, Campus Crime ............................ 6
The Perpetrators and Victims of Crime on Campus ........................ 10
Purpose of the Present Research ...................................................... 12
II. THEORETICAL FRAMEWORK ........................................................ 14
Environmental Criminology ............................................................ 16
Opportunity Theory or The Routine Activities Approach ............... 20
The Rational Choice Perspective ..................................................... 23
Situational Crime Prevention as Integrated Theory ......................... 26
Criticisms of the Situational Perspective and its Component Parts . 28
Conceptualization Problems ................................................. 28
Applicability Problems ......................................................... 29
Crime Displacement ............................................................. 30
The Situational Perspective and Campus Crime .............................. 33

Literature Pertaining to the Nature and Scope of Campus Crime 34
Literature Pertaining to the Role and Function of Campus Public

Safety ...................................................................................... . ..... 3 8

Literature Pertaining to the Prevention of Campus Crime .............. ‘44
Crime Prevention Policies Involving the Entire Campus

Community ...................................................................... 44

Target Hardening ................................................................. 46
Increasing the Visibility of Campus Police Officers and

Implementation of Community Policing ........................ 47

Volunteer-Driven Strategies ................................................ 50

The Present Study ............................................................................ 51

Research Questions .............................................................. 53

Chapter Page

HI. METHODS ........................................................................................... 55
Research Design and Data Collection ............................................. 55
Selection of Variables ...................................................................... 56

Dependent Variables ............................................................ 56
Independent Variables ......................................................... 58
Limitations of the Data .................................................................... 67
Missing Data ........................................................................ 68
Validity of Using Existing Data .......................................... 69
Criticisms of Law Enforcement Statistics as an Indicator
of Overall Criminality ..................................................... 71
Units of Analysis ................................................................. 73
IV. ANALYSIS AND RESULTS ............................................................... 75
Descriptive Statistics ........................................................................ 75
Extent of Serious Criminality ............................................... 75
Institutional Characteristics .................................................. 85
Crime Prevention Strategies ................................................. 90
Campus Safety Department Characteristics ......................... 97
Multivariate Analyses ...................................................................... 107
Factor Analysis and the Construction of Scales ................... 108
Crime Prevention Strategies ..................................... 109
New Officer Selection Techniques ........................... 1 13
Correlations Among Factors Influencing the Numbers of
Crimes Occurring at Post-Secondary Institutions ............ 113
OLS Regression Analyses - Partial Models .......................... 114
Campus Public Safety Department Selection
Techniques for New Officers ............................... l 18
Campus Public Safety Department Structure and
Function ............................................................... 122
Crime Prevention Programs and Strategies .............. 125
CharacteristiCs of the Institutions ............................. 130
OLS Regression Analyses - Full Model ............................... 134

vii

Chapter Page

V. DISCUSSION AND CONCLUSIONS ................................................... 139
Summary of the Findings .................................................................. 139
Limitations of the Study .................................................................... 142
Utilization of the Survey Research Method .......................... 143
Conceptualization and Operationalization ............................ 144
Limitations of the Data Analysis .......................................... 145
Policy Implications ........................................................................... 147
Recommendations for Future Research ............................................ 152
APPENDICES ........................................................................................................... 155
Appendix A - Definitions and Coding of Variables Used in OLS
Regression Analyses .......................................................... 155
Appendix B - Items Comprising Scales Used in OLS Regression
Regression Analyses ............................................................ 159
LIST OF REFERENCES ........................................................................................... 161

viii

LIST OF TABLES

Table Page

1. Part I Offenses and Rates of Offending at Participating Institutions, 1994 77

2. Comparison in Crime Rates of the Largest Colleges/Universities

between 1991-1992 and 1994 ..................................................................... 81
3. Demographic Characteristics of Institutions Included in the Study ................ 86
4. Physical Characteristics of Campuses Included in the Study ......................... 89
5. Crime Prevention Program Strategies and Implementation ............................ 92
6. Characteristics of Campus Public Safety Departments ................................... 99
7. Duties and Responsibilities of Campus Public Safety Departments ............... 105

8. Factor Loadings for Crime Prevention Items fi'om the LEMAS
Questionnaire .............................................................................................. 111

9. Factor Loadings for New officer Selection Techniques from the LEMAS
Questionnaire ............................................................................................. 1 12

10. Correlations Among Factors Influencing the Number of Campus Crimes ..... 115

ll. OLS Regression Analyses of Campus Public Safety Department
Selection Techniques upon Crime Figures Standardized by Total
Number of Constituents ............................................................................. 119

12. OLS Regression Analyses of Campus Public Safety Department
Selection Techniques upon Crime Figures Standardized by Resident
Population ................................................................................................... 121

13. OLS Regression Analyses of Campus Public Safety Department
Structure and Function upon Crime Figures Standardized by Total
Number of Constituents .............................................................................. 123

Table

14.

15.

16.

17.

18.

Page

OLS Regression Analyses of Campus Public Safety Department
Structure and Function upon Crime Figures Standardized by
Resident Population .................................................................................... 124

OLS Regression Analyses of Campus Crime Prevention Programs and
Strategies upon Crime Figures Standardized by Total Number of
Constituents ............................................................................................... 127

OLS Regression Analyses of Campus Crime Prevention Programs and
Strategies upon Crime Figures Standardized by Resident POpulation ...... 128

OLS Regression Analyses of Campus Demographic Characteristics
upon Gross Campus Crime Figures ........................................................... 133

OLS Regression Analyses of the Full Model upon Gross Campus Crime
Figures ....................................................................................................... 136

CHAPTER I

INTRODUCTION

The Context of Campus Crime

Concerns about the safety of campus constituents and preventing criminal activity
at institutions of higher learning have become hot topics both in the academic press, as
well as among university administrators throughout the country. As violent criminal
victimization rates increased nationally throughout the 1980's and early 90's, the campus
environment has not enjoyed the traditional immunity that it had in the past (e. g. Lunden,
1983; Roark, 1987; Sloan, 1992a; Fisher, 1993; Lederman, 1993; Palmer, 1993; Nichols,
1995; Sloan & Fisher, 1995; Richards, 1996). In fact, an examination of crime rates at
America’s institutions of higher learning reveals them to be microcosms of the larger
society from which they draw their students; no longer are they peaceful enclaves into
which one may escape the harsh realities of everyday life (Stormer & Senarath, 1992;
Sloan, 1994; Fisher, 1995).

The reasons why college and university campuses in the United States have been
slow to come to grips with crime and violence are varied, but most observers agree that to
some extent, administrators, parents, employees, and students simply did not want to
acknowledge that problems existed in places that should perhaps be resistant to such

social malaise. Even though it has been documented that post-secondary institutions have

had instances of crime and deviance throughout the centuries (e.g. Sloan, 1992b; Smith
and Fossey, 1995), some schools have seen crime rates become a serious and sizable
problem within only the past decade or so (Smith and Fossey, 1995). In many cases,
today’s reality of collegiate violence, larceny, rebellion, and criminal mischief and its
antecedents in the campus uprisings of the late 1960’s. Smith and Fossey (199528) sum
up the radical changes of thirty years ago, which have perhaps irreversibly altered the
American post-secondary institution:

“The decade of the 1960’s was a watershed in American history; the

cultures at both ends of that lively period were dramatically

different. The crime and violence that were becoming common fare in

the parent society were evident on campus as well. In addition, the

phenomenon of crime in that period served as a catalyst for other sorts

of profound changes in campus life and administration...[A]s

beleaguered administrators sought to retain control of their campuses,

...[t]he concept of the campus as sanctuary, unfettered in governing

its own affairs and unsullied by the world, was challenged on many

fronts during the 1960’s and 1970’s.”

Even though current rates of violent and property crime remain, on average, lower
than the towns and cities which surround university campuses (Lizotte & Fernandez,
1993; Bromley, 1995a), these on—campus rates have grown steadily in pr0portion with
national crime rates, and the opportunities for violent offending are virtually limitless
given the nature of the collegiate environment (Roark, 1993). In his recent text, F elson
(1998375) discusses criminality at the post-secondary level in rather simple, yet very
logical terms:

“A college camWS is a giant delivery system. It funnels students,
faculty, staff, and various products into one general location and
moves them from building to building to deliver education. It also
delivers crime opportunities. Students and their property may play

several roles in crime. University employees fit the same bill. The
university itself owns many items worth stealing and provides

many comers where illegal action can occur.

By studying crime on college campuses, we can see how

locality and crime interact. We have low-rise and high-rise
campuses. Some are compact and some are spread out. They range
from colleges with fewer than 1,000 students to large universities
with more than 30,000 students. Some are in urban settings,...some
are far more rural.

This great variety of colleges and universities differ in crime
vulnerability because of how they structure life. All crime is local--
every offender goes fiom some local area to another. The

minimal elements of crime converge locally. The physical and
social processes generating crime therefore are intricately related.”

While campus crime figures at present are not reported prOportionate to their
populations as are national statistics in the Uniform Crime Reports (U CR), when some
researchers have converted campus statistics to rates, they have found no significant
differences in crime rates between schools and the parent society (e. g. Smith and F ossey,
1995), while others report substantially higher rates off-campus as opposed to on-campus
(Bromley, 1995a). Similarly, while some scholars (e.g. Fernandez and Lizotte, 1993;
Fernandez and Lizotte, 1995) see the overall rates of campus crime decreasing, it is
undeniable that the gross number of crimes is growing, as are the numbers of students
attending and residing at the nation’s institutions of post-secondary learning, thus

increasing the likelihood of convergence in time and space of potential offenders and

victims.

The Campus Crime and Securig Act of 1990

Although the rise in crime to parallel the larger society had been evident to

campus administrators and law enforcement personnel for several decades (see Smith and

F ossey, 1995), the general public was not privy to such information until the federal
statute known as "The Student Right-to-Know and Campus Security Act of 1990" (also
the Crime Awareness and Campus Security Act of 1990, the Campus Crime and Security
Act of 1990) mandated colleges and universities which receive federal funds, or as a
condition of participating in Federal student aid programs, to annually compile and
publish instances of violent and property crime occuning within their borders (Bromley,
1992; Seng & Koehler, 1993; Sigler and Koehler, 1993; Seng, 1995; Reaves and
Goldberg, 1996; Sloan, Fisher, & Cullen, 1997). Further, although the tenants of the
Campus Crime and Security Act of 1990 do not require them to do so, many colleges and
universities voluntarily report crime statistics to the FBI for inclusion in the gym
gum; Rpm (Seng, 1995). In sum, those interested parties wishing to utilize such
information have been able to obtain it since the Campus Crime and Security Act of 1990
became effective in 1992 (Fisher and Sloan, 1993). Supporters thought the legislation
would uncover the “dark secret” of violent crime at America’s colleges and universities,
and would provide students and parents with a useful piece of information to ponder
when deciding which school to attend (Sloan et a1., 1997). Although the underlying
purpose of the Campus Crime and Security Act of 1990 was to forewam present and
future students and their parents about instances of violence at particular institutions, the
breadth and scope of the legislation was much broader. As Sloan et a1. (1997:150-151)
explain:

“A key provision of this unfunded mandate is the requirement that all

post-secondary educational institutions receiving federal funds

prepare, publish, and distribute annual reports of on-campus crime

statistics and campus security policies. additionally, the legislation
mandates that the reports contain counts of on-campus arrests for

alcohol, drug, and weapons violations. The legislation calls for
definitions for all offenses to be consistent with the Federal Bureau
of Investigation’s Uniform Crime Reports.”

The Student Right-to-Know and Campus Crime and Security Act of 1990 is
important for the present discussion of campus crime prevention because, in addition to
providing information on the numbers of crimes committed on a particular campus, a
document must be drafted which informs students and employees about campus security
features and crime prevention policies (McEvoy, 1992). Specifically:

“The annual security report should contain information about

procedures for students and others to report crimes, security and

access to campus residence, and the nature of the campus security

personnel including their relationship with state and local police

(McEvoy, 1992:147).”
The legislation “lays bare” the school’s security plan for the campus and, most
importantly, highlights potential weak points in emergency response and the prevention
of crime (Fisher and Sloan, 1993; Sloan et al., 1997). At the time Congress passed the
Campus Crime and Security Act in 1990, many smaller schools, as well as larger
institutions in less criminogenic locations did not have such plans and procedures, did not
staff a professional public safety department, nor did they know how to develop
prevention strategies appropriate for their particular size, location, nature of facilities, etc.
This uncertainty remains a problem in 1998, and has recently been the fuel for lawsuits
against universities whose safety measures or responses to criminal activity in the past
have been less than adequate in the court’s eye (McEvoy, 1992).

Interpreting the tenants of the legislation has also been a source of dissension and

confusion within higher education. Some states, such as Connecticut and Pennsylvania,
have stricter regulations regarding campus crime reporting, while others simply require
post-secondary institutions to compile crime data (e.g. Griffaton, 1995). At face value it
would seem that opposition to a piece of legislation which was drafted to help educate
and ensure the safety of campus constituents would be scant to non-existent. However,
some institutions, such as those situated in urban or “high crime” areas, felt that the
Campus Crime and Security Act of 1990 would artificially inflate their crime figures
because incidents which technically occur on-campus but do not involve constituents of
the school would be counted as campus crimes. Also, these same institutions contend that
suburban and rural schools tend to displace crime to the surrounding community, and
thus crimes committed by and against campus constituents in these surrounding
communities do not appear in the reported statistics (Bromley and Territo, 1990; Seng,
1995). These controversies have led to allegations of misrepresentation and
underreporting of crime statistics by colleges and universities (e.g. Haworth, 1996;
Gearey, 1997; Lively, 1997), as well as calls for even more suingent legislation which
would permit public access to campus police incident reports, perhaps leading to the
public disclosure of students who committed or were victims of certain offenses,

including misdemeanors (e. g. Kirtley, 1997)

The Impact of, and Regppnse to, Campus Crime
This groundbreaking legislation, coupled with steadily increasing campus crime

rates, has led to vast expansions of existing crime prevention programs as well as the

implementation of new and innovative responses to incidences of violence by both
campus public safety departments and administrators at U.S. colleges and universities.
For example, Michigan State University instituted a Community Policing program in the
late 1980's in an effort to reduce increasing rates of property and interpersonal crimes
(Benson, 1993), and the University of Illinois at Chicago in 1991 deve10ped a cooperative
initiative with the Chicago Police Department's 12th Precinct to reduce auto-theft rates
for parts of the city surrounding the UTC campus (Frost, 1993). The University of
Maryland took a more direct approach in the early 1980’s in response to increases in
violent crime. The campus increased its number of sworn, armed officers 40 percent, it
implemented “checkpoints” at campus entrances during evening hours, and fortified the
dormitories with heavy-duty security devices (such as stronger locks and shatterproof
door glass) (Middleton, 1981). These examples lead one to believe that the range and
sc0pe of university efforts to address criminal issues are virtually limitless. Indeed, every
school has its own unique crime problems which it must address, and the demands placed
upon campus police departments and administrators can be extreme (Sloan, 1992a; Fisher
& Sloan, 1993; Palmer, 1993). Thus, an examination of specific programs and how they
are implemented by a given school could conceivably tell us a great deal about how
successful these crime prevention programs will be toward achieving their stated goals;
whether those goals be a lowering of crime rates, a reduction in student fear, or a reversal
of a national reputation as a "crime-infested" campus.

In efforts to ensure the safety of their constituents and spare themselves the

potential liability for lax responses to campus criminality, most larger colleges and

universities have adOpted aggressive, proactive strategies to prevent crimes from ever
occurring (e.g. Dodge, 1991; Fisher & Sloan, 1993; Utz, 1993). According to Kessler
(1993), the issues of potential liability and civil litigation are main reasons for the
proliferation of new innovations in campus security. Further, colleges and universities
now have the responsibility to alert students about potential risks of on-campus
criminality, as well as to provide adequate protection services as advertised (McEvoy,
1992; Griffaton, 1993; Kessler, 1993). If the school does not live up to its end of this
bargain, courts have had no problems assessing monetary penalties against these schools
on behalf of crime victims and their families (Smith, 1992).

Many of these programs or policies are adaptations of strategies which are used
by municipal departments, such as community- or problem oriented- policing (see
Trojanowicz, Benson, and Trojanowicz, 1988; Jackson, 1992; Benson, 1993; Phillips,
1993; Riesling, 1993; Johnson, 1994; Pagon, 1995; Powers, 1996). Others are
innovations first utilized by private and industrial security contractors, that focus on
environment and physical design to reduce the opportunity for offenders to successfully
commit an offense (Mayhew, 1991). While many of these crime prevention techniques
are easily adapted to any given campus, each school must assess and respond to its most
pressing security needs. Without doubt, a cooperative, as opposed to adversarial,
relationship between the campus law enforcement body and the constituents of the school
must be in place before any program is initiated. A significant part of any needs
assessment policy-making must include follow-up evaluation of program effectiveness

and feasibility (Fisher & Sloan, 1993). A noted evaluator of crime prevention programs,

Dennis P. Rosenbaum (1986), has written that the “success” of any such program is the
fundamental question that must be asked to both the law enforcement side of the equation
as well as the constituents whom the program or policy was designed to protect.
Invariably, what constitutes success is a function of context and perspective; there are
bound to be differing opinions among and within parties having a vested interest in
campus crime rates. Thus, it is imperative that before any crime prevention initiative
which seems promising is implemented, all involved parties come to a consensus
regarding what outcomes will be deemed successes and which failures.

The reality is that campus crime affects the bottom line for many colleges and
universities. Prospective students and parents of those students, now have easier access to
campus crime statistics since the passage of the Campus Crime and Security Act of 1990.
Decisions about which school the student chooses to attend are based, at least partially,
upon whether a significant risk of victimization on campus or in the surrounding
community exists. Higher than average crime figures and/or a prevailing climate of fear
among the constituents of a campus could have grave implications for the recruiting
process (e.g. Beeler, Bellandese, and Wiggins, 1991; Fisher and Nasar, 1992; Newman,
1996). At a time when the competition for top students is particularly high, an institution
certainly does not want a non-academic factor such as dispmportionately high crime rates
to possibly influence the student’s decision to matriculate. Further, post-secondary
institutions must remain cognizant of the successful civil suits initiated by students and
parents against universities for crimes that could have conceivably been prevented, or for

which the campus response to a given crime was ineffective or insufficient (Smith, 1988;

10

Smith, 1992).
The Pe_rpetrators and Victims of Crime on Campus
Earlier reports of campus crime (e.g. Middleton, 1981) indicated that many law

enforcement oflicers at large universities believed criminal activity, especially crimes of
violence, to be the work of persons not associated with the institutions. However, more
current research into the area of on-campus violence and victimization indicates that
those individuals committing violent crimes tend to be legitimate members of the campus
environment, and usually are fellow students (McEvoy, 1992). These crimes of violence
on campus which involve students as constituents and victims are often impulsive in
nature, the result of an argument or disagreement, and most often occur at night or on
weekends, when resident students are not in classes, but in or near campus housing units
(Palmer, 1993). Such instances of interpersonal violence also fi'equently involve the use
of alcohol by either the perpetrator, victim, or both (Pollard and Whitaker, 1991; Palmer,
1993; Lederman, 1994). As summarized by Siege] & Raymond (1992:24):

"We suggest that campus violence and crime are related to campus

ecology and that factors within college life promote, stimulate, or in

some way allow violent behavior. We propose that the violent act

is a function of the person and the environment. The person in the

equation includes all the habits, attitudes and expectations of campus

life, and prior experience with freedom and responsibility. The

campus climate includes all the changing local, national and

international conditions as well as all the people who are present on

campus, and includes the physical and social environment. It appears

possible that the typical campus crime perpetrator would not be likely

to commit a crime in a non-college environment."

The implication is that something about the collegiate experience and/ or the

campus environment is providing license for otherwise law-abiding young people to

11

commit acts of deviance that they would not otherwise perpetrate, and indeed would
perhaps frown upon in a larger societal context. This presents a dual problem for campus
administrators who are faced with rising instances of interpersonal violence at their
institutions, and the knowledge that their own constituents are perhaps the source of this
upswing. Given this information it is not difficult to understand why campus leaders are
somewhat reluctant to divulge campus crime information to prospective students and
other interested parties (McEvoy, 1992), and why proponents of the Campus Crime and
Security Act had little objection from law-makers in getting the legislation passed and
signed by former President Bush (Sloan et al., 1997).

There are a plethora of feasible responses or strategies that could be utilized by a
school to reduce crime rates and lower the potential risk of violent victimization of
students and other constituents. However, a truly significant step must take place between
the administration building and the larger realm of the campus itself before crime can be
affected. The implementation and sc0pe of proactive strategies are all too often the weak
links in any college or university crime prevention program. As is illustrated in an article
by Paul Lang (1993) regarding the implementation of community-based policing on
college campuses, simply adopting such a design is only the first basic step. The author
puts forth issues pertaining to hiring, training, and public relations that may sometimes
escape discussion during meetings of campus planners. An extremely pressing issue is
that of adapting the program to fit the needs of the campus and to convince campus law
enforcement personnel and constituents that such a strategy will be beneficial to the

community as a whole (Lang, 1993).

12

Pu_rpose of the Present Research

While the trend within the field of campus public safety over the past decade or so
has been for a conversion from the ‘watchman’ style to the proactive method of law
enforcement, yet to be evaluated is the overall effectiveness of campus public safety after
such transformations have been implemented. Therefore, an important aspect of this work
is its status as the first, nationwide systematic analysis of the structure and functions of
campus public safety departments in the law enforcement-era. Secondly, this dissertation
also represents the initial application of an integrated Situational Crime Prevention theory
as a framework for addressing the problem of serious, on-campus criminality. Lastly, as
an exploratory study, the present research serves as a reference point for future empirical
work focusing on serious campus crime. More detailed investigations into this issue are
needed, optimally using a longitudinal research design, and campus crime scholars may
use these results as a benchmark in devising and conceptualizing further evaluations.

The specific focus of this dissertation is to assess the present state of campus
public safety in the United States, post-Campus Crime and Security Act of 1990, as well
as the scope and breadth of crime prevention and secmity programming at colleges and
universities nationwide. Ultimately, assessments of the collective impact upon the
school’s crime rates of variables pertaining to public safety and crime prevention
initiatives will be made. Analyses will evaluate the relative successes of various public
safety departmental operations, crime prevention, and security programs, controlling for

institutional characteristics such as number of students and employees on campus, and

resident population.

13

CHAPTER II
THEORETICAL FRAMEWORK

All too often, empirical research into the areas of private security and law
enforcement efforts at crime prevention commence on an atheoretical note. This is quite
unfortunate given the promise of several interrelated theoretical frameworks which have
been examined and tested in different contexts over the past three decades. Recently, an
applied theoretical perspective which falls under the umbrella heading of “Situational”
crime prevention has been oft utilized in scholarly research dealing with formulations for
and evaluations of crime prevention policies (see Clarke and Mayhew, 1980; Clarke
1992). This theoretical perspective differs markedly from most socio-cultural

perspectives in that Situational crime prevention assumes potential offenders to be a

 

constant in our society, thus the focus is not upon what motivates the criminal from
committing a given unlawful act, but rather examines what conditions or variables would
persuade an otherwise motivated offender to n_ot commit an act against a certain target or
victim (Hindelang, Gottfredson, & Garofalo, 1978; Cohen & Felson, 1979; Clarke, 1980;
Adler, Mueller, and Laufer, 1995).

The Situational perspective, as applied theoretically to crime prevention contains
three specific features. First, explanation is focused more directly on the criminal event
(Clarke, 1980), not primarily on the offender. Secondly, the need to develop explanations

for separate categories of crime is made explicit (Clarke, 1980). In other words,

14

15

preventive techniques must be adapted specifically to the type of offending one wishes to
thwart, and not toward specifically deterring the offender. Third, the individual’s current
circumstances and the immediate features of the setting are given considerably more
explanatory significance than in positivistic theories (Clarke, 1980). Sociologically
speaking, the perspective is by definition micro-level, whereas most positivist theories
stress macro-level variables. Suggestions for actual crime prevention techniques arising
out of the Situational model center on increasing the chances for an offender to be caught

and reducing the physical opportunities for offending (Clarke, 1980).

According to Clarke (199225), there are five stages to any application of
Situational prevention to a research problem:
1. Collection of data about the nature and dimensions of the specific crime
problem.

2. Analysis of the situational conditions that permit or facilitate the
commission of a specific, troublesome form of crime.

3. Systematic study of possible means of blocking opportunities
for these particular crimes, including analysis of costs.

4. Implementation of the most promising and practicable measures in a
way that will permit evaluation.

5. Dissemination of the results.
Given these components, the key aspect of the Situational approach is that it encourages
the inclusion of contextual variables that have been traditionally neglected in prior work
examining crime prevention techniques (Clarke, 1988). To do so, the Situational

perspective draws upon several related areas of criminological theory : environmental

l6

criminology, opportunity theory (specifically the routine activities approach), and the
rational choice perspective (Clarke, 1988). Other scholars have chosen to include a
discussion of Oscar Newman’s (1972) idea of ‘defensible space’ as well, although the
aforementioned theoretical strains assume many of his earlier ideas. Increasingly, these
strains of theory have come to be used somewhat interchangeably (Garofalo, 1987;
Maxfield, 1987), however, the rationale behind each is explained below for purposes of

clarity.

Environmental Criminology

A number of contemporary criminologists do not have a great deal of confidence
in theoretical explanations of the prevention of deviant behavior which rely upon physical
or environmental measures which may reduce the opportunity for an offender to commit
a criminal act (Clarke, 1980). Perhaps rightly so, many scholars in the field choose to
focus on the “inception” of criminal deviance, not the set of responses to deal with those
individuals who, for one reason or another, are motivated to offend. However, as debates
about what “causes” delinquency rage on, with few clear answers in sight, communities,
institutions, and individuals are increasingly concentrating on how to prevent criminal
victimization, assuming that inevitably criminals are “somewhere out there” waiting for
the right opportunity to put into action their deviant plans. Coupled with an
overwhelming atmosphere of apprehension toward becoming the victim of a crime, which
is all too often fed and perpetrated by the popular media, a distinct focus on the

prevention of crime and reduction of deviant opportunities has emerged in the

17

criminological literature. As Clarke (1980:139) has stated:

“Some practical options for prevention do arise...from the greater
emphasis upon situational features, especially from the direct and
immediate relationship between these and criminal behavior. By
studying the spatial and temporal distribution of specific offenses
and relating these to measurable aspects of the situation.
criminologists have recently begun to concern themselves much
more closely with the possibilities of manipulating criminogenic
situations in the interests of prevention.”

Environmental Criminology draws heavily upon the “defensible space” idea
proposed by Newman (1972). His basic premise was that the overall risks associated with
committing a certain act may be increased significantly through environmental design
(Lab, 1992). Out of this idea has emerged a practical method of operation which in
private security circles is known as Crime Prevention Through Environmental Design
(CPTED) (see Wallis & Ford, 1980). Primarily, CPTED is conceptualized as a warning
system and primary deterrent to would-be offenders, utilizing architectural designs that
enhance territoriality and surveillance, target hardening, and the recognition of legitimate
users of an area (Lab, 1992). Initially, CPTED was a component of urban planning (see
Jacobs, 1961), where such designers provided for personal safety by orienting public and
private space such that “guardianship” was easily accomplished (i.e. building entrances
clearly visible, sidewalks designed so that they would be heavily traveled) (Miller, 1990).

The “modern” strain of CPTED is attributed to the work of C. Ray Jeffery (1971)
and Newman (1972). Jeffrey (1971) believed firmly in the notion of the rational offender
(to be discussed below). If, when contemplating a potential criminal opportunity, the

individual perceives the “benefits” outweigh the “costs” of the action and guardianship is

18

at a minimum, then the would-be offender naturally proceeds. Ultimately, Jeffrey (1971)
believed that while environmental design was one part of effective crime prevention, the
human element of guardianship and making oneself a less attractive target need also be
employed to reduce criminal opportunity. Newman (1972) incorporated this need for the
human element described by Jeffrey in his defensible space concept. As he describes in
his opening chapter:

“‘Defensible Space’ is a surrogate term for the range of mechanisms...

that combine to bring an environment under the control of its

residents. A defensible space is a living residential environment

which can be employed by inhabitants for the enhancement of their

lives, while providing security for their families, neighbors and

friends.”
Although the defensible space concept was geared toward urban geographers and
planners, with its emphasis on ‘Territoriality’, ‘Surveillance’, ‘Image’, and
‘Environment’, certain elements appealed to criminologists (such as Cohen, Clarke, and
Felson) who were developing a theoretical explanation for crime which was based on the
everyday movements and activities of people within and outside their homes, places of
work, etc.

Presently, Crime Prevention Through Environmental Design has become
synonymous with the practical techniques used by security professionals and designers to
thwart a potential ofiender’s idea about committing a crime (i.e. placement of doors, use
of video surveillance cameras). However the Defensible Space idea which gave rise to

CPTED is grounded in the same philosophic ideas about human nature that underlie the

Deterrence Perspective and Rational Choice. It assumes man to be a logical thinker who

19

acts in his/her best interest while in the pursuit of self-gratification. Therefore, while
critics may argue CPTED is essentially atheoretical, the actual crime prevention
techniques are a response to distinct social scientific ideas about the human mind, and
what drives individuals to certain behaviors.

A large part of reducing and preventing criminal opportunities through
environmental planning rests on knowing the geographic locations where a majority of
serious crimes take place (Lunden, 1962; Rengert, 1988; Brantingham & Brantingham,
1991). To be sure, prevention techniques may only be successful if they are implemented
in a location for which a serious need has arisen. These areas that endure frequent and
severe criminality have come to be known as ‘hot spots’, and thus receive a
disproportionate amount of attention by crime control strategists (Brantingham &
Brantingham, 1984; Clarke, 1988; Sherman, Gartin, and Buerger, 1989). According to
environmental criminologists, criminal activity and potential victimization are best
understood as occurring in the context of normal everyday movements and activities
(Brantingham & Brantingham, 1991). Also, while people’s movement patterns are the
impetus for criminal activity, those who are so inclined to perpetrate criminal acts choose
from among many such movements which place offender and potential victim in the
same time and space. While this decision-making process does indeed included elements
of rational-choice (i.e. a cost/benefit analysis on the part of the offender) (Cornish &
Clarke, 1986) and routine activities (Cohen and Felson, 1979), there are also numerous
environmental cues “that separate good criminal opportunities from bad criminal risks”

(Brantingham & Brantingham, 1991 :3). Thus, when these environmental cues are

2'0

examined collectively, those agencies specializing in crime prevention have a clearer
picture of needs assessment, and subsequently can more effectively employ resources in

accordance with calculated need and potential risk.

Qpportunig; Theog or The Routine Activities Approach

For at least two decades, criminologists have examined the relationship between
environmental or contextual variables and subsequent rates of criminal victimization in a
given area. Perhaps Hindelang et al. (1978) were the first to link theoretically “lifestyle”
variables and exposure to risk that may or may not result in personal victimization
(Maxfield, 1987). Some of the most notable work to come out of this exercise has been
the “Routine Activities” explanation for differing crime rates. In a seminal piece written
almost twenty years ago, Lawrence E. Cohen and Marcus Felson (1979) illustrated the
variables or circumstances that must come together in space and time for the criminal act
to successfully take place. ’

Specifically, for a deviant act to be successful, a convergence in space and time of
likely or motivated offenders, a suitable target or targets, and an absence of capable
guardians to prohibit victimization, is necessary (Cohen & Felson, 1979; Bennett, 1991).
Rm to environmental criminology, the routine activities approach incorporates
previously underutilized contextual variables such as locan'm, type and quantity of illegal
acts, and community organization as predictors of level and probability of criminal
victimization (Cohen & F elson, 1979). Much of the literature advocating the routine

activity explanation of criminal victimization has stated that today’s more mobile

21

lifestyles, and the prosperity they produce, have facilitated and eased this convergence in
time of the factors necessary for victimization (Felson & Cohen, 1981; Felson, 1987).
Michael Maxfield (1987:275) states the issue quite simply:

“What people do, how they behave, places them at more or less risk of

criminal victimization. Stated so directly this seems like little more

than common sense. In much the same way, routine behavior

‘explains’ how people avoid or suffer many other misfortunes. Careful

drivers are at less risk of auto accidents, careless smokers stand a

greater chance of setting fire to the house, coal miners suffer

disproportionately from pulmonary disease, and mariners are more

likely to drown than, say, mountain climbers.”
The three key elements of the routine activities approach are further specified in
Bennett’s (1991) work. While a convergence of the three is necessary for a crime to
occur, if one or more is not present, or is lacking to some degree, the opportunity for the
crime to occur is negated. Thus, according to Cohen (1981:141), “[T]he concentration of
patterned or routine activities...varies inversely with the rates of criminal victimization.”

Originally conceived, the routine activities approach was a theoretical fiamework

for explaining or predicting the risk probability of an individual becoming a victim of
interpersonal violent crime given several “lifestyle characteristics” (e.g. Cohen and
Felson, 1979; Felson and Cohen, 1980). But the utility of the perspective for both
theorists and practitioners has been expanded to be predictive of the risk of other types of
crime, such as property or white collar, as well (e.g. Cohen and Cantor, 1981; F elson,
1987; Roncek and Maier, 1991; F elson, 1994). For example, Felson (1997) has argued

that technological innovations and an increase in the flow of capital have both contributed

to an increase in opportunities for ofi’enders to commit certain acts, given more suitable

22

targets (i.e., lighter and smaller electronic equipment) and a lack of guardians as goods
and cash flow more heavily, in larger geographic spaces. As Roncek and Maier
(19912727) sum this phenomenon :

“The importance of movement and activities for crime is among the

original themes of routine activities theory, and both are used to

explain part of the varying levels of victimization of people with

different lifestyles. Facilities or nonresidential land use provide some

of the reasons why people go to different places within the city to

carry out their lifestyles.”
The above application of routine activities theory to nonviolent victimization has been
adopted by writers who examine trends and correlates of campus crime (e. g. Sloan 1992a;
" Bromley 1995b; Smith and Fossey, 1995). A common theme is that the university
campus provides the motivated offender with a multitude and variety of possible targets,
given that most students and faculty possess items of value, over which they are not
particularly vigilant guardians. Any campus public safety administrator will describe
instances where unlocked rooms were entered and expensive electronic equipment stolen,
unlocked offices were burgled' for computer equipment or purses and wallets, and
unattended bookbags stolen from libraries and the foyers of classroom buildings so the
thief could sell back textbooks for cash. In all of these cases, the movement of goods with
some monetary or cash value occurs fiequently and the somewhat transient lifestyle of
the owner does not result in “capable guardianship.”

In sum, at the individual level, persons may take cautions which will minimize the

likelihood that they become “suitable targets”. Systemically, police departments, facilities

planners, and in the present case, campus administrators do what they can to ensure that

23

“capable guardianship” is maximized, as the university environment is one where,
typically, students and other constituents of the institution do not concern themselves
with lessening their target suitability (Middleton, 1981). This is not to say that a college
or university cannot aid its constituents in becoming less suitable targets, or even in
trying to “unmotivate” offenders, rather guardianship is one area of the routine activities
equation over which the school has some modicum of control and is within its’ sphere of
influence. Indeed, Fox and Hellman (1985) point out that several variables related to the
physical structure and location of facilities on the university campus correlate

significantly with the schools’ level of crime.

The Rational Choice Perspective

Coinciding with the arrival of complex quantitative methods in criminological
research came an economically-derived theoretical position at the heart of which assumed
that any given offender, being a rational individual with the free will to make his or her
own life decisions, will choose to commit or not commit a criminal offense based on the
ratio of possible benefits of the act weighed against the inherent costs or risk associated
with that act (Cornish & Clarke, 1986; Akers, 1990). Underlying the rational choice
perspective are tenants from the classical criminological thought of Cesare Beccaria,
Jeremy Bentham, and Cesare Lombrosso; all of whom had similar ideas on the nature of
human (specifically criminal) behavior. Within the early classical theories of crime,
humans are seen as independent, free-thinking, and logical decision-makers, who can
control their own destiny by defining their self-interests (Einstadter and Henry, 1995 :47).

The classical perspective defined those who commit crimes as hedonistic or pleasure

24

seeking, motivated only to maximize their personal gain (Einstadter and Henry, 1995).
The rational choice model:

“[D]oes not see decision-making as always being fully rational

or even properly considered; instead the model employs

notions of ‘limited rationality’ in which economic

explanations of crime are modified by taking cognizance of

other motivational and cognitive factors (Clarke and Cornish,

1983 :50)”.

The idea that potential offenders consciously weigh the ‘pros and cons’ of any
given criminal offense is an offshoot of the larger theory of deterrence, which essentially
states that persons will be discouraged from committing a criminal act if the costs or
penalties are raised such that they outweigh any potential benefits or gains of the act
under consideration (for an intriguing discussion of the state of deterrence as a viable
criminological theory, see Paternoster, 1989). Specific to “costs”, the severity, certainty,
and swifiness of punitive sanctions are most oft examined (see Klepper and Nagin, 1989;
Yu and Liska, 1993), however a multitude of factors may be processed (e.g., Nagin and
Paternoster, 1993). The range and scope of variables that the ‘rational’ offender ponders
is illustrated by Akers (1990:655):

“It [Deterrence] encompasses a full range of behavioral inhibitors
and facilitators : rewards/costs; past, present, and anticipated
reinforcers and punishers; formal and informal sanctions; legal
and extra-legal penalties; direct and indirect punishment; and
positive and negative reinforcement...Some of the rational choice
models of crime in the literature have been expanded beyond
the basic expected utility proposition to include family and
peer influences, moral judgments, and other variables.”
Given this choice-structuring process, Akers (1991) has also posited that logical

cost/benefit analysis may not be the prime factor in such choice-structuring, rather it is

25

the offender’s perception of the benefits that is often the critical factor in choosing
whether to carry out a given criminal act. Similarly, Cornish and Clarke (1987) believe
that the aforementioned choice structuring is both offender and offense specific; that is,
individual offenders contemplate different factors when deciding whether to commit a
given type of offense. Thus, efforts aimed at preventing specific criminal acts, must
consider a multitude of possible situational and contextual variables which may be taken
into account as possible deterrents of criminality.

These contextual variables are what makes the theory attractive to advocates of
the Situational approach, as Cornish and Clarke (1986) believe that choice-structuring is
highly contingent on the potential offenders’ perceptions of risk factors in the physical
and social environment. According to Seigel (1989:113), “offense and offender
characteristics are interactive. Each offense has its own properties, including risk and
payoff; each offender has a unique set of needs and skills.” The implication being that,
since man is a rational individual who can assert his or her own free will, a deterministic
approach to choice structuring may uncover behavioral patterns, but the end result of such
structuring is influenced to varying degrees by a multitude of contextual variables. In
Felson’s (1998:24) words, “Offenders are tempted and controlled by tangible factors in

immediate settings.”

Most importantly, offenders respond to the payoff, effort, peer support, and other
similar factors in deciding whether to commit a certain act (Cornish and Clarke, 1986;

Lab 1992). Some advocates of the Situational approach assert that when a potential

26

offender is discouraged from committing one crime at a certain place or time, they see no
alternative to the prevented crime, and thus mold their criminal opportunities around
available choices (Lab, 1992). If this assumption indeed holds true, then crimes may be
effectively prevented if the offenders’ choice structuring is altered to such a degree that
specific crimes become more effort than the perceived payoff is worth (Lab, 1997). Crime
prevention techniques, then, in the embodiment of rational choice theory, become
contextual variables that the, logical criminal must necessarily consider in order to

successfully take a crime to fruition.

Situational Crime Prevention as Integr_ated Theory

Ronald Clarke (1980) presents a compelling argument concerning the weaknesses
of applying sociological-based theories of deviance to practical crime prevention
situations. Many of these positivist theories focus on “dispositional” aspects of criminal
behavior which develop at a young age in the ofi’ender and/or emerge over time.
Techniques which could “prevent” the aspects from emerging are most certainly outside
the sphere of influence of any criminal justice systemic body. The utility of trying to
explain the root causes or contributors of deviant behavior, whether from a positivist or
critical perspective, is indeed a laudable task, however application of the results of such

research have not proven very beneficial to the practitioner.

The attractiveness of Situational crime prevention, particularly for practitioners
and agencies who are responsible for ensuring a certain area is safe for those who venture

into it, is rooted in the versatility of the approach. More specifically, by incorporating the

27

tenants of the above theories, a more sophisticated and effective strategy of crime
prevention can be implemented. The aforementioned Crime Prevention Through
Environmental Design idea, which at present is extremely popular and widely-utilized by
crime prevention specialists, is essentially a hybrid of the three theoretical strains
discussed above, and may be directly linked to the earlier idea of defensible space.
Relating this concept directly to the college or university campus setting, the components
of Situational crime prevention can be incorporated to work cooperatively in efforts to
prevent victimization. The physical design of the campus may reduce the need for
additional guardians, given that facilities planning can incorporate “natural” guardianship
(see Rubenstein, Murray, Motoyama, and Rouse, 1980; Siegel and Harriss, 1992; Tucker
and Stames, 1993), which also raises the “costs” for potential offenders. Not to be
overlooked is the idea that constituents themselves can do their part, in conjunction with
campus law enforcement, to reduce their attractiveness as a potential crime target (Zahm
and Perrin, 1992). Undoubtedly, any director of public safety in a collegiate environment
will report that individuals leaving themselves vulnerable as “targets” is a serious
hindrance to crime prevention.

An integrated Situational Crime Prevention Theory incorporates certain
assumptions about the nature of man, his/her thought processes, cues which provoke
behaviors, and factors which make criminal opportunity seem less attractive to the
potential offender. Clarke (1980) reasons that the subjective state and thought processes
which play a part in the decision to commit a crime will be most influenced by immediate

situational variables, thus the perceived opportunity to offend may be as much of a trigger

28

in taking a criminal act through to fruition as any life circumstances or psychological
variables which serve as the basis of much criminological thought. Felson (1988) argues
that since security, or the protection of oneself, family, and possessions from predators, is
one of mans’ basic human needs, it is illogical to argue that crime prevention is abstracted
from social scientific theory. He goes on to posit that the notion of security as “locks,
bolts, and badges” is a relatively recent definition, and that security should look to its
roots as a basic need for conceptualization purposes (Felson, 1988). If one also includes
the ideas inherent in CPTED, this integrated approach outlines specific techniques and
strategies for security specialists, architects, and planners to consider, thus moving
Situational crime prevention from an esoteric theoretical idea to a concrete, plan of

implementation for policy-makers.

Criticisms of the Situational Pefipective and its Component Parts

Concgptpalization Problems

Perhaps it is necessary at this time to make a few qualifying statements about the
Situational approach as it is presently conceptualized. Many scholars have criticized this
perspective because of the assumptions made by its proponents. For example, F elson
(1987) states that, routine activities specifically, assumes that offenders are not well
informed, foreword looking, or particularly unrelenting. As stated earlier, the Situational
perspective and the theoretical strains fi'om which it is derived assume an ample supply of
motivated offenders to be within society at any given time. Some critics of the
perspective see this assumption as a conceptual leap, in that it does not necessarily

explain why motivated offenders are concentrated in a certain space and time, or why

29

certain societies tend to have greater concentrations of motivated offenders. In essence,
because the Situational perspective does not address the genesis of criminal behavior, the
perspective’s utility is questioned. Metaphorically speaking, without trying to address the
root or cause of the problem, the criminal justice will continually be plugging a leaking
dike that gets weaker and more overwhelmed. Clarke (1980:136), in his piece which
framed what is presently known as ‘Situational’ crime prevention, rebuked this criticism
as follows:

“Though it may be conceded that preventative measures (such as humps

in the road to stop speeding) can sometimes be found without

invoking sophisticated causal theory, ‘physical’ measures which

reduce Opportunities for crime are often thought to be of limited

value. They are said merely to suppress the impulse to offend which

will then manifest itself on some other occasion and perhaps in even

more harmful form.. Much more effective are seen to be ‘social’

measures...since these attempt to remove the root motivational causes

of offending. [I] argue that an alternative theoretical emphasis on

choices and decisions made by the offender leads to a broader and
perhaps more realistic approach to crime prevention.”

Applicability Problems

A second major criticism has to do with the lack of explanation the Situational
perspective has for particular types of criminal acts. In certain instances, say when a thief
is trying apartment doors to see if any are unlocked, a simple deadbolt may be enough
target hardening to deter the would-be offender. However, if the individual wanted to
physically harm the person on the other side of the locked door, there may be no “capable
guardian” short of police protection that would take away the offenders’ motivation.
Thus, one of the major shortcomings of crime prevention in general is applicable also to

this theoretical perspective. Certain ofi'enders’ motivation is not easily deterred, and some

30

offenders’ cost/benefit analysis (if indeed any exists) may be skewed so much so that no
potential “cost” is great enough to offset the perceived benefit. Therefore, Situational
crime prevention is not easily applied to circumstances (such as domestic disputes or
revenge situations) for which an offenders’ motivation is excessive, and the only capable
guardianship would be to meet force with force. This difficulty in applying the
perspective to particular circumstances is an inherent problem with any macrolevel
theory. Proponents of the Situational Perspective (as well as proponents of other macro-
theories) will argue that the perspective is intended to explain broad patterns of deviant
behavior at the aggregate level, taking into account a variety of contextual and
environmental variables. Because of human being’s ability to determine their own destiny
(i.e. free will), the perspective may very well explain the actions of one individual but not
another. But again, as most sociologists will state, social science is most productive when

interpreting behavioral patterns, not individual actions.

W
Much has been written about another major criticism of the Situational
perspective; the idea of crime displacement. Critics who espouse this idea say, quite
simply, that a crime which is prevented at one time or place, will be attempted by the
offender elsewhere or at the same location but at a different time. Lab (1992:67) describes
displacement, or “crime spillover”, as:
“The idea...that actions taken to reduce or prevent crime in one set of
circumstances simply result in the crime ‘spilling over’ to another
locale or potential target. The assumption is that many crime

prevention actions simply move the crime around instead of eliminating
the overall amount of crime. For example, an increase in police

31

presence in one neighborhood may engender a reduced crime rate in that

area. A contiguous neighborhood, however, may experience a

corresponding increase in its crime rate. The net result is no or a very

small decrease in the overall crime rate.”
This viewpoint assumes that crime cannot actually be prevented, rather the impetus of
crime carries on until the offending behavior can be successfully completed. Not
surprisingly, prOponents of positive theory are often quickest to denounce Situational
crime prevention on these grounds, stating that the only true way to prevent crime is to
keep ofienders from becoming motivated to commit crimes, and this best occurs by
improving the social conditions that are thought to breed criminogenic influences.
Advocates of the Situational perspective argue that crime displacement is not a necessary
result of crime prevention. Whether a potential offender whose efforts at committing an
act were encumbered in one instance will try again at a different time or location, depends
on the nature of the crime, the offender’s strength of motivation, knowledge of
alternatives, willingness to entertain them, etc. (e.g. Clarke, 1980; Clarke and Cornish,
1985)

Further, in an qualitative study conducted by Reppetto (1976) over twenty years
ago, the widespread opinion that crime prevention measures inherently lead to
displacement were refuted through interviews with ofl’enders who had previously
committed the Opportunity crimes of robbery and burglary. Reppetto (1976: 177)
concluded that the displacement hypothesis commonly attributed to crime prevention
techniques or policies was flawed for two major reasons. First, some crimes are so

Opportunity-driven that merely preventing their occurrence in one instance is sufficient

enough to remove the motivation of the offender. Unless another as-attractive Opportunity

32

immediately presents itself, ,most offenders will not seek out an alternative target.
Secondly, even if Offenders are blocked in one sphere and wish to move on to another
area, there are real limits or costs to their doing so, and these limitations and costs will
lessen their frequency of operation, and perhaps increase the likelihood of their being
arrested while committing a criminal act.

In placing this argument in the context of a college or university campus, Clarke
(1980) posits that for opportunistic crimes, ones that are elicited by their case of
accomplishment (such as theft and vandalism), the probability of offending could be
drastically reduced by making it more difficult to commit these types of offenses. To be
sure, most campus public safety directors would undoubtedly report that these
Opportunistic types of crime are, while not necessarily capturing a great deal of media
attention, the most problematic for the departments. The research of Bromley (1995,
1997), Sloan (1994), and Sloan et al. (1997) report that larceny/theft is the most prevalent
crime reported to law enforcement agencies on university campuses. If one applies the
prevailing idea that theft is one of the most underremrted of all crimes (Territo, Halsted,
and Bromley, 1995), then one may correctly infer that the problem of larceny offenses on
American campuses goes much deeper than official statistics can gauge. But if
opportunity crimes such as vandalism and larceny/theft are the major crime problems on
campus, and most offenders are students who would not normally commit such Offenses
in another setting (Siegel & Raymond, 1992), then the application and subsequent
evaluation of Situational prevention techniques, including CPTED, most definitely seems

warranted.

33

The Situational Perspective and Campus Crime

The nature of most Part I index crime on college and university campuses
facilitates the application of Situational Crime Prevention as a theoretical framework for
examinations of serious criminality. Post-secondary institutions have become
increasingly concerned with deterring would-be Offenders and eliminating the infinite
opportunities for criminal behavior available to the potential offender. However, because
the development of the perspective and the empirical research into campus crime are in
their infancy, the Situational perspective has not, as an integrated theory, been applied to
any systematic analyses of serious on-campus crime. Scholars who have previously done
research into campus crime highlight its Opportunistic nature and state their belief that
those individuals who engage in such acts can effectively be dissuaded. Because
Situational Crime Prevention focuses on the factors conducive to offending, it is logical
to apply the perspective to campus crime as a method of explaining and eradicating the
behavior.

Several studies have been undertaken which apply certain aspects of the
Situational perspective, but consequently ignore other key components. Much of this
prior research indicates that elements of the perspective are significantly correlated with
on-campus offending and/or have previously been successful in preventing Opportunity
crimes (McPheters, 1978; Fox and Hellman, 1985). This earlier work is discussed below,

as is prior research that has focused on public safety as a component of crime prevention.

34

Literature Pertaining to the Nature and Scope of Campus Crime

 

A significant amount Of empirical research has been published in the past two
decades in response to increasing numbers of criminal victimizations at the nation’s post-
secondary institutions. While the raw number Of crimes reported to campus police
officials is certainly reason enough for investigation of this perplexing issue, the
ascending level of violence associated with these reported crimes perhaps more than
anything has resulted in this scholarly output. Also worth noting is the wide variety of
Offenses that are characteristic of the campus environment; a fact that has resulted in
studies comparing deviance on school grounds to crimes occurring in society at large. As
Smith and Fossey (1995: 10) explain:

“The reality of campus crime is an...anomaly. The campus has been set
aside as the place for intellectual pursuits, discourse, and reflection;
yet today it is the scene of the same sorts of violence, larceny, and
criminal mischief as the parent society. The higher education press
regularly reports such things as professors shot in their classrooms,
arson in campus buildings, rapes in dormitories, thefts of everything
from rare books to luncher tickets, fiaud in grants and student loan
programs, sexual extortion of students by faculty members, forgery of
transcripts and diplomas, and run-of-the—mill vandalism. Some
campuses suffer a great deal of crime, some far less. But in many
places the myth of the safe campus endures while the campus community
inures itself to the undiscussed threat.”

As the frequency and severity of campus criminality increases, more explanatory
assessments are being made about this emerging and complex issue. The present decade
has seen a significant increase in empirical studies assessing the nature of crime at

American post-secondary institutions, and more recently, evaluations of existing

programs and policy directives are published in a diverse group of academic journals and

35

popular presses. With regard to the later, the focus tends to be on the occurrence of
violent crimes which plague schools large and small, urban and rural (e. g. Storrner, 1989;
Pollard and Whitaker, 1991; Smith and Fossey, 1995). The concerns within social
science about campus crime tend to center upon the effects of increasing crime rates upon
constituent fear, and policy alternatives available to administrators and campus law
enforcement personnel (e. g. Bordner and Petersen, 1983; Bromley and Territo, 1990;
Fisher and Nasar, 1992). A vast majority of this prior literature focusing on the nature of
crime on college campuses has reported increases in the levels of criminality and the
overall seriousness of the offenses committed.

Because this rising trend had been evident for some time prior to the attention
paid to the problem in the academic journals, two decades ago Fox (1977) wondered why
scholars were largely ignoring the issue, especially since the rise in gross crime rates was
greater than in the overall population. Although crime data from colleges and universities
was sporadic and often less than reliable, the rising trend was clear quantitatively, and
qualitatively as well, since media reports of large-scale campus unrest were prevalent
throughout the late 1960’s and early 1970’s. But those researching campus crime did not
have a clear picture about the scope and severity of the problem until the mid to late
1970’s. Projects such as those undertaken by Fox (1977) and Lee McPheters (1978)
provided a clearer understanding about crime occurring on campus, and generally found
that the problem was worse than had been feared initially.

Additionally, McPheters (1978) was perhaps the first scholar to employ tenants of

Situational crime prevention (although this perspective was not yet developed to the

36

formulation we know today) to campus criminality and crime prevention. He believed
campus crime rates to primarily be a function of opportunities for would-be offenders
Offset only by the level of security activity at the particular campus. McPheters (1978)
thus employed a cost/benefit equation to estimate the amount of preventable crime, given
various characteristics of the universities and the aforementioned level of campus
security. His results indicated that two key independent variables were most highly
correlated with crime rates; a higher proportion of students living in dormitories, and a
close proximity to urban areas with high unemployment rates (McPheters, 1978; Smith
and F ossey, 1995).

This second key variable is a bit surprising, given that prevailing thought supports
the notion that much campus crime is attributable to constituents preying on other
constituents and on- and off-campus crime rates are not significantly correlated (e. g.
Bromley, 1995). McPheter’s (1978) results suggested that crime rates may be attributable,
to some degree, to nonstudents. Many scholars feel that the campus environment has
changed significantly in the twenty years since McPheters (1978) performed his research,
and most crimes are being committed by persons with legitimate ties to the schools (e.g.
McEvoy, 1992). Regardless of the background of the perpetrator, McPheter’s (1978) and
proponents of his work argue that Situational crime prevention techniques would be
equally efiecfive in “target hardening” campus constituents, making them less vulnerable
to either “insiders” or “outsiders” (Smith and F ossey, 1995: 14).

Because of the multifaceted context of the physical structure of college campuses,

the diversity of constituents, and the criminogenic influence of surrounding areas,

37

explanations about why certain crimes are common, or what factors influence the
prevalence of violence are similarly variant. Some writers have pointed to the dual
problems of alcohol abuse and the carrying of firearms as a way of explaining the
increase in “impulse violence” on campus (Lederman, 1994; Nichols, 1995). Fox and
Hellman (1985) found crime rates to be primarily correlated with the size and scholastic
quality of the institution, and reported the size of the community in which the campus
was located to not be a significant correlate Of crime; an assertion which was not
supported by Sloan’s (1992a) research. Quite obviously, there are as many explanations
for why crime occurs on campus as there are campuses in existence, making institutional
responses all the more difficult, for even if distinct correlates can be ascertained, it is
highly probable that they will not remain static for very long, thus requiring campus
security to make frequent adjustments in their policy decisions.

Sloan’s (1992a) more recent work also examines the role of environmental factors
in predicting campus crime rates. He found that the setting, the selectivity, and the cost
to attend the school were all significantly correlated with the campus crime rate (Sloan,
1992a). His results, while demonstrating an overall rise in the gross number of all types
0f crimes reported to campus law enforcement, indicate that theft/larceny is still the
Prevailing ofi’ense committing on campus, and therefore the fear of violent victimization
may be unfounded. Also, since this national data set was able to take into account the
comel'ttual differences specified above, Sloan (1992a) concluded that a Situational
approach to controlling or preventing certain types of Offenses would certainly be

plausible, since several significant correlates could be identified as influencing the

38

numbers of crimes reported.

Literature Pertaining to the Role and Function of Cflpus Public Sgttv

Any scientific enquiry into campus crime prevention necessarily requires a
discussion of the agency or body responsible for developing and implementing such
strategies. At most colleges and universities, a campus public safety department,
structurally and philosophically similar to municipal police agencies, is responsible for
patrolling school property and responding to calls for assistance. However not all
institutions of higher learning employ sworn law enforcement Officers. Indeed, many
smaller colleges and universities, especially those in rural and suburban “low-crime”
zones, have resisted the trend of their larger brethren toward becoming institutions with
firlly sworn and equipped police forces (Onishi, 1995; Hummer, Austin, and Bumphus,
Forthcoming). Some campuses have stalled on the progression described by Sloan
(1992b), which has seen campus law enforcement emerge from its early days as a
“watch” system to a fully self-contained policing agency, with full arrest powers and
capable of tactical response. Sewell (1993) states that college and university security
plans are as diverse as the schools themselves. In some instances there are hybrid systems
of both full-time security watchmen and fully-swom officers; in other cases fully sworn
Oflicers are employed, but their powers are limited by the school’s administration (e. g.
Hummer et al., Forthcoming). The worst case scenario are the schools, most often small
in size and not in “crime prone” areas, which have either consciously or unconsciously
chosen to altogether ignore the issues of campus safety and security (Sewell, 1993).

Sloan (1992b) documents the progression and modernization of the campus police

39

over a one hundred year period. The first “campus police officers” were city of New
Haven policemen who were hired by Yale University in 1894 to patrol university grounds
during evening and overnight hours (Powell, 1994a; Powell, 1994b; Peak, 1995). Sloan
(1992b) states that early campus security (approximately the first half of the twentieth
century) can best be characterized as a “watchman” system, where schools hired retired
men to serve in a custodial capacity. They had no formal law enforcement training or
policing powers, rather they protected campus property from fire, water, or other damage,
and reported suspicious activity to local authorities (Espositio and Storrner, 1989; Sloan,
1992b). Campus public safety as we know it today did not emerge until the 1950’s. Sloan
(1992bz86) elaborates:

“This decade saw many colleges and universities begin a period of

unprecedented growth in student enrollment and in physical size. With

this boom, campus administrators soon realized there was a need for a

‘police presence’ on campus. As a result, colleges began hiring

retired municipal police officers to serve as ‘directors of campus

security.’ The first thing these directors did was to reshape campus

security into an agency similar to urban police departments. To do

this, they created campus ‘security officers’ and campus ‘security

departments’. They also tried to severe the ties of campus security to

the physical plant department, where security had traditionally been

housed. In doing this, they sought to make campus security more

autonomous.”

Even though these campus security departments had begun to resemble municipal

police departments, their role and fimction remained similar to that of their predecessors
(Bordner and Petersen, 1983; Sloan, 1992b). The seeds of professionalization were not

planted until the turbulence of the 1960’s and early 1970’s, when campus unrest became

a major problem for college administrators. As Bordner and Petersen explain (1983: xi):

40

“The transition from the silent generation of the fifties to the

active generation of the sixties and early seventies left many
colleges and universities totally unprepared to cope with the new
threats to the tranquillity of the campus. With the advent of the era
of student dissent, campus protest demonstrations, disruptive student
activities, violence and increases in reported crime and fear of
crime, an increasing number of educational institutions began
replacing their line security officers with more educated and better
trained police Officers with police powers of arrest and duties to
enforce state statutes on campus...lt was also recognized that if

the university did not govern itself it would be governed by others
who might be less responsive to the campus community. Thus,
professional police departments began to emerge on college campuses
during the 19605 and early 19708, and law enforcement activities
became part of the ever expanding role of the university officer.”

For many larger schools the dual factors Of internal unrest and rising crime rates in
surrounding communities necessitated the above mentioned changes in policy and role
for campus security. The militaristic model of the police, outlined by Klockars (1985),
which had become the norm for municipal and state police agencies, began permeating
the collegiate environment. Campus police, now instead of just protecting persons and
property, were now given license to “fight crime”, and with this new role, campus police
began to structurally resemble municipal police departments (e.g. in terms of training
requirements, organization, etc.) (Peak, 1989; Smith, 1989; Peak 1995) and physically
mirror their local sworn counterparts in dress and in the hardware (firearms, handcuffs,
etc.) that they carried (Sloan, 1992b).

A review of earlier empirical research pertaining to law enforcement in the
university environment shows that schools were primarily concerned with constituent
safety and building security (see Etheridge, 1958; Nielsen, 1970; and Gelber, 1972).

Some writers in the policing arena, while not academics concentrating in the area Of

41

campus safety, saw law enforcement on the college campus as a relatively easy task,
given the homogenous, mostly law-abiding individuals who comprised the student
population in earlier years. It was thought that simply the threat of harsher punishment
would “scare straight” those “naive, starry-eyed undergraduates” who take advantage of
the campus environment and “prevent our colleges from functioning effectively”
(Gardner, 1970:119). Such writings reflect the uncertainty of the times with regard to
campus unrest and rising crime in communities at large. It also reflects the beginning
point of the diversification of the collegiate environment.

However, just as the overall turbulence and social unrest of the 1960’s changed
the face of policing an urban setting, the college campus was also not immune from these
changing social forces, and the social environment at America’s universities is much
different, of come, than in years past. Over a quarter century ago, Sims (1971) published
an edited monograph that contained several articles calling for the professionalization of
campus security forces in response to increasing rates of violent and property crimes.
Although the need for a professionalized force was apparent to law enforcement
professionals, Gelber (1972) recognized a reluctance on the part of the academic
community, especially students, to accept their campus police force as a legitimate law
enforcement agency. Later empirical assessments of constituent satisfaction with campus
police services echoed Gelber’s ( 1972) findings, and further clouded the legitimacy of
law enforcement in the collegiate environment (Cordner, Marenin, and Murphy, 1985 ;
Miller and Pan, 1987; Peak, 1988; Hummer and Bumphus, 1996). Also, Sloan, Fisher,

and Wilkins (1993b) found that a number of wide-ranging variables may contribute to

42

negative constituent attitudes toward the campus police. Some focus on specific police
activities (i.e. crime prevention policies or patrol strategies), others are more concerned
with the more general environment of fear or satisfaction with police service provision on
campus (i.e. overall performance or keeping crime rates low).

This lack of acceptance was somewhat confusing to researchers in this area, as
Scott’s (1976) national study Of college and university law enforcement reported that
campus police departments, as a whole, provided the same services and performed the
same duties as municipal agencies. Further, Scott’s (1976) conclusion was that campus
law enforcement was emerging as an important local producer of police services, not just
to the schools themselves, but also to the surrounding communities. Even though the
progression from the “watch” system to the militaristic and/or community-oriented
approach to police work (see Peak, 1989; Benson, 1993; Lang, 1995; Lanier, 1995) is
clearly delineated by Sloan (1992b), Bordner and Petersen (1983) point out that not every
campus in America has chosen to replace their approach to campus security. They see a
tremendous amount of diversity in the role and function of campus security or law
enforcement departments. Some schools have seen no need, despite the passage of the
Student Right-to-Know and Campus Security Act Of 1990, to change fiom a watch
system to a crime-fighting system. More common are the schools with security systems
which are a hybrid of security and peace officers. Bordner and Petersen (1983: xii)
describe them as:

“...[M]odem security forces in which the watchman function has been

extended to include pseudo-police functions and are generally
involved to some degree with regulation of student conduct.”

43

In some instances, these hybrid systems are comprised of sworn police Officers who do
not possess all of the powers granted municipal police officers (see Hummer et al.,
Forthcoming), but more commonly they are comparable to contract security agencies,
who are granted certain powers (such as detaining suspects), but are not sworn peace
officers.

Regardless of the form or function of campus law enforcement, the agency
responsible for overseeing the safety and security of the schools’ constituents faces an
inherent dilemma in trying to balance freedom of movement and thought (a keystone of
the concept of the “campus”) with the notion of preventing deviant acts from taking
place; knowing that a good deal of such prevention activities could possibly impede upon
the individual rights of certain parties (Folven, 1995; Sloan and Fisher, 1995). Roark
(1993:5) elaborates on the problem of crime prevention for campus Officials as being
threefold. First, the campus is a relatively Open and free place, academically and
physically. It can rarely be closed to the public (ostensibly because most are publicly
funded), and doing so would challenge the founding principles of these institutions.
Second, the campus is not only Open, but transient as well. Students, faculty, staff, etc.
are usually only in residence a few years before moving on. Also, these individuals are Of
differing backgrounds and cultures which may pose problems with regard to target
attractiveness or offending. Third, the lifestyle and routine activities of those constituents
on campus (i.e. movement at late hours, a high density of residents in campus

dormitories) creates almost limitless criminal opportunity (Roark, 1993).

44

Literature Pertaining to the Prevention of Campus Crime

As previously discussed, there are a number of viable crime prevention techniques
at the disposal of campus law enforcement administrators, which can be implemented
taking into account the context or characteristics of their particular institution. Certainly,
these measures have been proposed in global terms in accordance with the macrolevel
perspective of Situational prevention. It should not be inferred that all of these strategies
could effectively be implemented on all campuses. Sloan (1992a239), has broken
potential responses down into four distinct categories: (1) those that involve the entire
campus, (2) “target hardening” buildings, dormitories, and other locations, (3) increasing
the visibility of campus Ofiicers and implementing service-oriented patrols, and (4)
developing other strategies to prevent crime (such as the use of volunteers and student
“mule patrols”). Prior research outlining crime prevention strategies from these four areas

is discussed below.

Crime Prevention Policies Involving the Entire Campus Community

A daunting task for any college or university is to facilitate an atmosphere where
constituents of a school, especially students, take crime prevention and safety seriously,
especially when crime does not appear to be a problem or fear of crime is not a pertinent
issue. Undoubtedly this is the largest obstacle to overcome in putting together a crime
prevention strategy. Campus officials can only do so much to try and ensure peoples
safety while they are using campus facilities, and they can only make recommendations
and instruct people about how to reduce their chances of being victimized. The problem

is that certain behaviors associated with college life mean that these warnings are ignored

45

in the pursuit of daily activities (such as late night partying, walking alone to classes, the
library, or study sessions, the propensity to leave doors unlocked for easy access, etc.)
(e.g., Fisher, Sloan, Cullen, and Lu, 1997). For some of the most promising crime
prevention tactics to be successfirl, more participation from constituents is necessary, and
by extension a radical change in behavioral dynamics may be required for significantly
positive results to become manifest.

This is not to say that certain advances can be made regarding crime prevention
programs that require some constituent cooperation. Bromley (1994) sees promise in
public education initiatives that would focus specifically on women’s safety and
dormitory security. Additionally, students could be organized to watch out for each other
in small groups (similar to a neighborhood watch program), and, along with maintenance
and service personnel, could report suspicious activities in and around campus residential
areas to the campus or local police (Bromley and Territo, 1990; Bromley, 1994). Further,
other offices of the college and university may take a role in crime prevention/personal
safety besides the police department. In addition to law enforcement presentations about
how to protect oneself, information can be dispensed from counselors, the Office of
student afl’airs, and health care officials to name but a few (Nichols, 1995).

Most schools begin designing, or re-vamping their existing, crime prevention
plans with a task force that evaluates the crime problem as it presently exists and what
tools they have at their disposal with which to respond. As Richards (1996:49-51) states:

“The initiation, analysis, and implementation of a security
survey should be considered the first step in a progressive
campus crime prevention program... Ideally, a campus

security survey should be designed to investigate the levels
of existing security at the site, assess vulemability to

46

criminal opportunity, and to make appropriate
recommendations upon analysis of the findings.”

All too often, this step of assessing the needs of the institution are not taken into account
by all involved in administering security on campus. Policy initiatives are frequently put
in place by overzealous university administrators and campus law enforcement
professionals who overgeneralize positive findings from other universities, without taking
into account aforementioned contextual variables, thus resulting in less than stellar

OUICOITICS.

6

Target Hardening’

Referencing the above discussion, many scholars believe the Situational approach
holds great promise in preventing crime on campus, and have thus called for prevention
measures which utilize CPTED. Of course, the purpose of such strategies is to reduce the
perceived or real physical Opportunities for offending or to increase the chances of an
Offender being caught (Clarke, 1980). Bromley (1994) asserts that the first step in
preventing crime through target hardening is to design buildings and facilities in such as
way as to make targets less attractive to would-be offenders, utilizing the principals
deve10ped by Jeffrey (1971) and Newman (1972). Campus buildings, pathways, and
common areas can be constructed in such a way as to ensure they are well traveled and
clearly visible to passersby, thus providing some measure of guardianship. Fisher and
Sloan (1993:75) provide examples of these designs, which include blocking off hiding
spots or cutting down overgrown shrubbery to reduce physical opportunity, and

implementing foot patrols in potential “hot spots” to increase the chances of catching an

47

offender during the commission of a criminal act.

Additionally, CPTED techniques which are specific to deterring criminal events
include, but are not limited to, internal and external lighting systems, locking devices on
external building doors and resident rooms, alarmed and/or electronic access to
dormitories, staffing of front desks and registration of all guests in dormitories, and
emergency communication systems in strategic locations around campus that are wired
directly to campus police headquarters (Bromley and Territo, 1990; McEvoy, 1992;

Sloan, 1992a; Fisher and Sloan, 1993; Bromley, 1994).

Increasing the Visibilim of Campus Police Officers & Implementation of Comqu
Policing ,

Perhaps the most obvious tool of colleges and universities to deal with campus
crime is the professionalization of campus law enforcement described above. Indeed,
Fernandez and Lizotte (1993) give this as a primary reason for their argument that
schools are safer today than in decades past. Without doubt, law enforcement training has
greatly benefited departments that are responsible for large campuses or those in crime
prone areas. The knowledge gained by the officers who undergo academy training has
proliferated down so that university policy is based on the latest crime prevention
techniques currently utilized by municipal police departments and private security
contractors. Thus, safety advances which may have bypassed the college environment in
the past are now implemented in an environment that is arguably more controlled and
conducive to these techniques.

Whether the administration of safety on a college campus is entrusted to peace

48

Officers or non-swom personnel, lines of communication between the department,
constituents, administration, and local police are critical (Nichols, 1995). Bromley and
Tenito (1990:19) explain:

“The role chosen by the institution’s leadership for its campus law

enforcement department should be clearly defined for both the campus

community and the city or town in which the campus is located.

Regardless of the specific duties of campus law enforcement personnel,

there should be a close working relationship between the campus law

enforcement agency and the local police department which services the

surrounding community. A clear understanding of which department is

responsible for which kinds of activities on campus is very important.

Many campuses and local police departments have clearly defined

‘mutual aid agreements’ which help to ensure good cooperation in

critical situations. Communication between the two departments should

be frequent at both the formal and informal level.”
Because crime prevention strategies are most Often developed and certainly implemented
by the campus public safety agency, it is vital that those whose well-being depends upon
these policies be informed about their role in ensuring a safe environment. Therefore, in
addition to keeping the lines Of communication open between policing agencies, those
with a vested interest in campus safety should be included in policy analysis and program
implementation. These lines of communication, Of course, flow in both directions, as
some of the prevention programs initiated by the campus police depend upon
administrative approval and funding in order to come to fruition. As an example, Sloan’s
(1992a) recent piece found a correlation between crime rates and the Officer to student
ratio on campus. The more full-time public safety officers in relation to students, the

lower the amount of overall campus crime (Sloan, 1992a). Of course, this finding implies

that the lowering of this ratio, all things being equal, should produce a reduction in crime.

49

Since decreasing the number of students enrolled at the institution is typically not an
option, the hiring of more officers appears to be the appropriate response in this instance.
Such hiring, in turn, can only be accomplished with administrative approval and funding.

Campus crime experts see great promise in adapting community policing
principles to the collegiate environment (e.g. Trojanowicz et al., 1988; Jackson, 1992;
Lanier, 1995). Considering that a major ‘hot spot’ for criminal activity on campus is on or
near student dormitories (e.g. McPheters 1978), the assignment of campus police officers
to a particular residence hall or complex may be beneficial. Benson (1993) reports that
Michigan State University has seen a reduction in crime since such a policy was
implemented in late 1987. Bromley (1994:43) states the positives of such a policy :

“The Officers assigned to the dorm areas would become well known to
students and provide protection as well as psychological reassurance.
This step would reflect the concept of community-oriented policing
now being practiced by many municipal police agencies and in a few
university settings. Even if officers could not be assigned to the dorm
areas one hundred percent of the time, they could be required to make
routine checks in and around the student dorms.”

Further, results from larger universities indicate at least moderate support or
promise for a variety of programs created and implemented by community policing
officers. Bromley (1997) identified a number of correlates regarding auto theft at the
University of South Florida in Tampa, and made recommendations which could be put
into action by an officer assigned to this problem, such as coordinating a Lot Watch
program or placing signs in lots reminding people to lock their vehicles and keep

valuables out of sight. At Michigan State University, community police officers

coordinate various prevention initiatives such as campus watch programs in residence

50

halls, mountain bike patrols, and office watch programs designed to reduce theft in
university buildings (Benson, 1993). Jacksonville State University in Alabama instituted
a preventive patrol strategy in 1992 to reduce violence and “nuisance crimes”, which the
school identified as precursors to violent victimizations, both on and off campus
(Nichols, 1995). At the University of California, Los Angeles, campus police oversee an
escort and transportation service that had previously been staffed by seven volunteers, but
now employs two hundred student workers (McEvoy, 1992) Lastly, campus police
Officers at the University of Alabama at Birmingham coordinate volunteer watch groups
which are “eyes and ears” that report suspicious activities on campus grounds and in

buildings (Sloan, 1992a).

Volunteer-Driven Strategies

With any volunteer—driven initiative, the most difficult logistical concerns center
on procuring and retaining participants, and with eStablishing leadership to coordinate
activities. On a college campus, the available pool of volunteers is usually within the
student body population. For such programs to take shape, those overseeing them must,
initially, have a sufficient number of interested persons who are willing to participate in
and be committed to a crime prevention program (e. g. Huff, 1990). Therefore, a plan of
action should be in place prior to recruiting volunteer participants, so that those who want
to be involved can see how the policy is designed and how it will take place.

A second concern, once enough volunteers are committed to the project, is
reducing mortality of participants and/or recruiting other volunteers to replace those who

withdraw from the program. Since a college student’s time is of the essence, and

51

volunteer work is perhaps low on their priority list beneath classes and paid employment,
if time constraints become a problem and activities are to be cut from their schedule, it is
logical that volunteer work is first eliminated. Further, if the program is one large in
scope, or requires a great deal of people to work at varied hours, then there will most
likely never be a sufiicient number of participants for the program coordinator.

Liability problems could also be a significant impediment to getting volunteer
crime prevention programs up and running. Since there is no guarantee that volunteers
will not come into contact with an offender, or be in danger while performing their duties,
there is always the possibility of litigation against the institution or program coordinator
if a volunteer is injured. Programs such as Lot Watches, Campus Escort Services, and
Safety Coordinators in residence halls, put their participants in potentially dangerous
situations by the very nature of the work. Even though volunteers often sign a waiver that
excludes the institution from civil liability, there are instances where such waivers do not
hold up in court, especially if the school is found negligent in some manner. The specter
of susceptibility to litigation then must also be taken into account, ideally in cost/benefit
terms. The decision has to be made if the program’s potential good is outweighed by the
detrimental effects of lawsuits should an unfortunate and unlikely situation occur where a

volunteer is harmed while participating in the program.

The Present Study
To summarize, earlier research has laid the foundation for an application of the
Situational perspective to the problem of campus crime, however the lack of utilization of

the full framework remains a glaring omission to the literature on serious campus

52

Offending. Little is known presently about how much serious crime is occurring on larger
college and university campuses nationally, and even less regarding the effectiveness of
the many crime prevention initiatives implemented by schools as a result of increasing
crime rates on their campuses or due to positive findings from research conducted at other
institutions. A great deal of information about the nature of campus crime comes from
research conducted at one institution or from a small sample of institutions in a given
geographic region. None of the studies presented above utilized a nationwide sample as
large as that used in this dissertation, nor have any of these studies collected data on as
many variables or facets of the campus environment and public safety as has the present
research. Much like the UCR is a measuring rod for criminality nationwide over the
course of a calendar year, this dissertation may be seen as a barometer of serious campus
crime at larger post-secondary institutions, and therefore fills a large void in the empirical
research pertaining to collegiate criminality.

Specifically, this dissertation will assess the operations and professionalism of
campus public safety departments, as well as the sc0pe and breadth of crime prevention
and security programs at colleges and universities nationwide, and evaluate their
collective impact upon the school’s crime rates respective to their implementation.
Relative successes of various crime prevention and security programs will be evaluated,
controlling for institutional characteristics such as the number of constituents affiliated
with the campus, geographic location and resident population. Ultimately, the results of
prior research into this topic and the work of scholars who have written conceptual and

practical pieces about the campus crime problem and potential responses to it, have been

53

used to develop a set of research questions to be examined.

Research Questions
The primary objective of this dissertation is to interpret the relationship between

campus public safety department operations, the various crime prevention strategies
implemented by institutions and the overall amount of criminal activity as reported by the
schools in compliance with the Campus Crime and Security Act of 1990. Therefore, the
initial research question, which is derived from the theoretical tenants of Situational
crime prevention, is:

1. Do institutions of higher learning who have incorporated crime

prevention techniques espoused by scholars who adhere to the

Situational perspective, have lower overall rates Of offending than

those institutions with fewer, or without, such measures in place?

Secondly, as discussed above in the review of literature pertaining to campus

police, the movement toward employing professional safety officers to patrol college and
university campuses is well documented. This is particularly applicable to larger, urban
schools, although smaller suburban and rural schools have also begun to embrace this
trend. Additionally, an examination of institutional procedure with regards to screening,
hiring, and training campus law enforcement personnel is possible. A key reason why
administrators have given the go-ahead to switch from a security-oriented to peace
officer- oriented system of social control is the argument that highly trained police
officers are more “professional” and therefore better able to handle safety needs as well as
emergency response (e.g. Gelber, 1972). There is a widely held assumption that sworn

police officers, who are often more educated and trained, will be able to implement crime

54

prevention policies into the specific campus environment. In order to determine the
nature of the relationship between professionalism and overall campus criminality, other
questions to be examined are:

2. Do institutions of higher learning, irrespective of student
enrollment or resident population, have less criminality
when serviced by their own campus public safety officers, as
opposed to institutions which contract out for protective
services?

3. DO institutions of higher learning, irrespective of student
enrollment or resident population, have lower overall
criminality when professionalism in terms of the duties
performed by campus public safety officers is accounted for?
4. Does the selection and training of campus public safety
personnel, irrespective of institution size or crime prevention
strategy, influence overall criminality at the institutions
served?

Finally, to examine the nature of crime in general on college and university
campuses, the dissertation will examine the relationship between crime rates and
institutional characteristics including, but not limited to, size of the constituency, location
of the campus, resident population, acreage of the campus, and the number of buildings
on campus. Therefore, the last research question to be addressed will be:

5. What impact to institutional characteristics have upon crime

rates and/or the number of crimes reported to the university or
college police?

CHAPTER III
METHODS
Research Desigp and Data Collection
The data utilized in this dissertation were Obtained from the United States

Department of Justice, Office of Justice Programs, Bureau of Justice Statistics (BJ S) Data
Archives (Bureau of Justice Statistics, 1997). To ascertain the nature and scope Of
protective services at larger U.S. colleges and universities, BJ S surveyed 4-year
institutions of higher education with 2,500 or more students during 1995 (Reaves and
Goldberg, 1996). The co-coordinators of the initial study describe the scope of the
project:

“This information comes from the 1995 Survey of Campus Law

Enforcement Agencies, the largest study of police and security services

at institutions Of higher education ever conducted. The eight-page, mail

survey asked about a wide-range of topics including agency functions,

hiring practices, employee characteristics, types of equipment used,

computers and information systems, expenditures, salaries, policies,

and special programs” (Reaves and Goldberg, 1996: iii).
The survey was administered under the auspices of the BJ S Law Enforcement
Management and Administrative Statistics (LEMAS) program, which had gathered
similar information on state and local law enforcement agencies in previous years

(Reaves and Goldberg, 1996). The questionnaire was distributed to campus public safety

agencies due the lack of representation of such agencies in the sample of municipal police

55

56

departments surveyed by the B] S LEMAS program. Further, as described above, in the
years since LEMAS had been conducting surveys of state and local police departments,
campus public safety has shifted from a security focus to a more proactive law

enforcement focus, thus necessitating an independent evaluation model.

Public safety departments at 680 college and university campuses were included
in the study population. Eighty-five percent, or 581, schools returned the questionnaire
with all or most questions answered, including 91 percent of the agencies at public
institutions and 76 percent of those at private institutions (Reaves and Goldberg, 1996).
Almost all safety departments responded at least in part. About three—fourths of these
agencies employed sworn police officers with arrest powers equivalent to municipal
policing agencies, while the other quarter relied primarily on nonswom security personnel
(Reaves and Goldberg, 1996). In this dissertation, for reasons of convenience, the term
“campus public safety” has been utilized to describe both groups, irrespective of whether

the officers are sworn peace officers or non-swom security personnel.

Selection of Variables

De endent Variables

The measurable outcomes of the LEMAS questionnaire are crime statistics,
similar to the Uniform Crime Rpmrt, collected by campus public safety agencies and
subsequently kept on file at the schools for public inspection upon request. In the present

study, gross numbers of Part 1 Index crimes (felony offenses which include

57

murder/manslaughter, forcible sex offenses, robbery, aggravated assault, burglary,
larceny/theft, motor vehicle theft, and arson) occurring on campus were reported for
1994, the latest year for which complete data were available when the questionnaire was
administered in late 1995 by the Bureau of Justice Statistics.

Most researchers realize that gross numbers of crimes occurring in a particular
jurisdiction are contingent upon a number of factors. Most obviously, the population of
the campus (or in the present case, the size of the constituency or resident population)
directly impacts the actual number of offenses reported to campus public safety. When
presenting the overall numbers of crimes for comparative purposes, the FBI standardizes
gross index crime statistics utilizing U.S. Census Bureau population data. The total
population is divided by 100,000, and the gross number Of crimes is standardized using
the resultant denominator.

In partial model regression analyses to follow, crime rates will be standardized
using both total constituency and resident population as denominators. This is done so
that size of the campus population is taken into account, so as not to produce spurious
results when examining only certain independent variables. Two population figures are
used due to an ongoing debate among those conducting research in the area of campus
crime. One school of thought is that all students enrolled and all employees of the
institution should be used as a pOpulation parameter, because all such individuals come
together in time and space on the campus, and all are potential offenders and victims.
Another argument is that since crime rates in the larger society are standardized using

permanent residents as a base, rates on campus should be calculated similarly for

58

comparative purposes. Therefore, in an effort to show possible differences when using the
two denominators, both will be presented in the partial analyses. The six outcomes thus
become: total Part I index crimes/ # of constituents, total Part I index crimes/ resident
population, Part I violent offenses/ # of constituents, Part I violent crimes/ resident
population, Part I property crimes/ # of constituents, and Part I property crimes/ resident
population.

Crime figures are left unstandardized in one partial model analysis as well as in
the complete model, and population figures (total constituents and resident population)
are included as controls. In these analyses, three dependent variables will be examined as
outcomes: gross numbers of overall Part I index Crimes, gross numbers of Part I violent
crimes, and gross numbers of Part I property crimes. Collectively, these are referred to as

“The Extent of Serious Criminality on Campus.”

Independent Variables

The key independent variables in the following analysis are the same as the
variables that scholars examining the overall effectiveness and organizational structure of
state and municipal law enforcement agencies, given that the survey instrument is
identical to those utilized in research on police departments and their crime prevention
initiatives. Therefore, comparisons may be made between campus law enforcement
agencies and those agencies that serve the population at large, in terms 'of personnel and
crime prevention initiatives. While most of these indicators are self-explanatory and

common knowledge to those who study criminal justice and/or law enforcement, a

59

discussion of the most pertinent correlates of campus crime in the present study follows.
Because the LEMAS questionnaire contains of multitude of variables, only some
of which apply to the research questions being investigated in this dissertation, it is
necessary to discuss those variables to be later utilized in explanatory analyses. The
independent variables which are to be included in subsequent linear regression analyses
can be subdivided under three distinct sub-headings: Characteristics of the Institutions
surveyed, Crime Prevention Strategies, and Campus Safety Department Characteristics
and Functions (see Appendices A and B for a full descriptions of dependent and
independent variables, coding, scales, and abbreviations used in subsequent analyses).
As stated in the theoretical fiamework, environmental criminologists hold that
factors associated with the physical structure of a location or facility are key to
preventing, or encouraging, criminal acts to take place. Newman (1972) held that an area
may only be truly “defensible” if the opportunities for a potential offender to commit a
deviant act are identified and eliminated by design. Of course, putting the necessary
components into place to achieve “defensible space” becomes a more difficult endeavor
when the space to be defended is large in terms of physical area. Fmther, the task is made
even more complex when the types and number of facilities on campus is great. For
example, a campus with a large area in acreage or with a large number of buildings
presents a challenge for the director of campus safety in planning and implementing
strategies to make such space “defensible”. Because the difliculty inherent in reducing
criminal opportunities over a vast territory, it is expected that campuses with large

numbers of and widespread facilities (in the present study, operationalized by the

60

LEMAS questionnaire as number of buildings, land area in acres, and miles of campus
roadways) will have crime rates which are correspondingly large, and may perhaps
respond to these rates with more detailed crime prevention programming.

In addition to the variables collected by BJ S using the Law Enforcement
Management and Administrative Statistics questionnaire administered to college and
university public safety agencies, additional data were collected on the nature of the
campus environment as well as the state of criminal justice programming at schools in the
sample. The latter of the two was collected for exploratory purposes, as the relationship
between the development of a criminal justice curriculum and crime prevention programs
at the institutions will be examined. Information on the campus environment was
collected because of the differing nature of crime in, and the policing of, urban, suburban,
and rural environments (e.g., Sims, 1988; Crank, 1990; Monkkonen, 1992; Weisheit,
Wells, and Falcone, 1994). Lunden (1983) attributes the rise in campus crime, at least in
part, to the rapid growth of metropolitan areas, and the associated increases in population
turnover and decreases in informal social control in the areas surrounding post-secondary
institutions. From the definitions of campus environments provided below, it is fairly
clear that an assumption is being made with regard to levels of community atmosphere at
the different types of campuses. In fact, one may view “informal controls” and
“community ties” on a continuum, where both decrease as campuses are more urbanized.

Information on this variable was gathered from the College Entrance Examination
Board’s (1997) The College Handbook, 1998. This source was chosen because it provides

information on the nature of the campus environment, and does not simply characterize a

61

campus solely based on the size of the community in which it is located. Following are
the College Entrance Examination Board’s (1997:46) definitions of urban, suburban, and
rural campuses:

“Urban Campuses : Can be as different as the cities in which

they are located. They include some of the nation’s oldest colleges,
with historic buildings and parklike quadrangles, as well as
colleges located in a single high-rise building with no traditional
campus at all. What most urban campuses share is ready access to
the extended cultural life of the city itself- museums, theaters,
professional sports, concerts, and so on.”

“Suburban Campuses: Can be quite varied, ranging from converted
shopping malls to traditional, ivy-covered buildings. Suburban
campuses are by definition located near cities; students who have

transportation into the city can take advantage of a diversity of
cultural events.”

6‘

ma] Campuses: Generally some distance from metropolitan
areas, and students must rely almost entirely on the social and
cultural events offered on campus. The ‘sense of community’ may
be increased because the college is more self-contained than in
larger communities.”

These definitions of nature of the campus environment are preferred because they
take into account the individual physical characteristics of the campus and do not simply
rely solely upon the latest census statistics for classification. Therefore, a campus which
is technically within the boarders of a major city may still be classified as suburban if the
campus has defined boundaries, or is primarily residential in nature. Conversely, a
campus may be classified as urban according to the definition above, even if the
community in which it is located is classified as a minor metropolitan area, or even a

large town. It is thought that the nature of the school’s environment may be correlated

with the extent of crime prevention policies, and subsequently, with the reported crime

62

rate at the institution.

The College Entrance Examination Board’s (1997) The College Handbook, 1998
was also used to collect data regarding the nature of the criminal justice program at each
of the 680 institutions. Specifically, this guidebook was used to determine if a school had
a separate baccalaureate program in criminal justice. Those colleges and universities who
had no criminal justice program, or who offered a criminal justice concentration in
another discipline were coded as not offering a bachelors’ degree in criminal justice,
while those who specifically reported having an undergraduate major in criminal justice
were so coded. For information regarding graduate programs in criminal justice,
Peterson’s (1998) Graduate Proms in the Humanities, Arts, and Social Sciences was
utilized. This guidebook contained information on whether an institution offered a
masters’ or doctoral program in criminal justice, or if the university offered a
concentration in criminal justice or criminology within another graduate curriculum.

Although it may not be viewed as a direct correlate affecting campus crime, the
possibility exists that the presence of a separate criminal justice curriculum at a college or
university, and thus by necessity a group of criminologists in residence, may impact the
development and implementation of crime prevention strategies on campus which can
directly influence rates of violence and property crime. More often that not,
criminologists and criminal justicians consult on projects for criminal justice agencies in
the region surrounding the campus at which they hold faculty positions. ISsues of crime
and safety on campus, because of the publicity and immediate proximity of the

“problem”, are naturally Of some interest to criminal justice faculty. These faculty

63

members can work with the campus safety department to develop strategies aimed at
crime prevention, victim assistance, and institutional response to criminal incidents.

An institution which Offers the criminal justice baccalaureate degree has within it
a pool of majors who could benefit campus safety departments as volunteers, interns, and
student employee help. While students of any academic major can (and do) participate in
such activities for campus safety departments, the hOpe is that students majoring in
criminal justice who participate are more committed to the work, and posses an
underlying knowledge of the criminal justice field that aids them in their tasks. For these
reasons, the presence of a separate criminal justice curriculum is seen as an institutional
characteristic which may potentially correlate with the development of specific campus
safety and crime prevention strategies. As such, the nature of the criminal justice program
at the schools under study will be utilized as an independent variable in the linear
multiple regression analysis.

Also included in this subsection of independent variables are the demographic
characteristics of total constituency and resident population which have been discussed
previously. Again, these two variables serve as standardization figures for crime rates,
and thus may be seen as control variables in the multivariate analyses.

The LEMAS questionnaire contained a number of items which enabled an
analysis of the relationship between crime statistics and crime prevention initiatives
presently in place at the institutions included in the sample. A few of the items which fall
under this heading have been so designated with some degree of subjectivity. For

example, the item pertaining to the arming of campus public safety officers may not

64

strike some. as a crime prevention technique. However if one views the arming of officers
as aiding “guardianship” as in routine activities theory, or as upping the potential “costs”
inherent in a given criminal act as in rational choice, then providing sidearms to officers
may be viewed as another, perhaps logical, preventative strategy employed by post-
secondary institutions to ensure constituent safety.

Other items, such as the presence of an emergency blue light system, an
emergency 911 system, and the various programs offered by institutions and administered
by the public safety department clearly fall under the heading of crime prevention. This
subset of independent variables is representative of the routine activities segment of
Situational crime prevention, as discussed in the preceding chapter, and show that
colleges and universities are best able to impact the “guardianship” element of routine
activities theory. However, some Of these crime prevention program strategies, such as
date rape prevention and selfldefense training, also help constituents reduce their “target
suitability”. Thus, the two elements of routine activities over which institutions have
some sphere of influence are indicated by this subset of independent variables.

The professionalization of campus public safety departments in the United States
has been well-documented (Gelber, 1972; Sloan, 1992b; Bromley, 1995b; Lanier, 1995).
Indeed, the data set utilized for this dissertation was borne out of the idea that, since many
schools are choosing to employ sworn law enforcement officers in place of traditional
private security guards to protect their campuses and many of these departments are
separate entities recognized by municipal departments near the institutions, there should

be a comprehensive analysis of the characteristics and structure of these departments.

65

Thus, the LEMAS questionnaire was distributed by B] S for the purpose of gathering
summary data on larger campus public safety departments nationwide. Because the
survey instrument was nearly identical to that given to municipal law enforcement
departments, items pertaining to structure, function, and role of these campus departments
model those of other policing agencies. For the following analyses, certain variables were
selected as indicators of the theoretical tenants of the Situational perspective and as
gauges of proactive order-maintenance activity at the participating institutions.

One variable which is unique to campus public safety departments is that of
outsourcing. Some schools, especially those smaller institutions and/or those in areas
with lower crime rates, will contract out to either local or state law enforcement agencies
or to contract security companies to provide routine patrol for campus property. Most
Often this practice is seen a savings in revenue, and may informally help enhance school—
community relations by supplying revenue to the local municipality. In this dissertation,
possible differences in violent and property crime rates will be examined between those
schools outsourcing campus protective services and those with such services in-house.
Similarly, criminality at institutions whose campus safety departments (either in-house or
outsourced) possess arrest powers within their respective jurisdictions will be compared
with those institutions whose departments are without such powers. This variable “arrest
power” is a demarcation between those institutions utilizing sworn officers (either in-
house or contracted) and those using private security contractors, for private security
officers normally possess only the same arrest powers as the average citizen.

An area that has become a central focus for colleges and universities is that of

66

dealing with the flow of illicit narcotics on campus. As is the case in larger society, the
underground economy of the drug trade flourishes in an environment where individuals
are receptive to the idea of experimenting with controlled substances and have the
resources to support such experimentation. The post-secondary institution certainly
typifies this description, and thus the personal use and distribution of narcotics on campus
is a major concern for both public safety chiefs and administrators alike. Many schools
with fully sworn officers are responsible for the enforcement of narcotics laws on
campus, and some participate in multi-agency drug task forces on campus and in the
surrounding communities. These two items will be included in later analyses to see if the
enforcement of drug laws is significantly related to overall crime rates on campus.

For descriptive purposes, several other characteristics of campus public safety
departments are presented in the following chapter (such as patrol types, number of
department employees, etc.) to demonstrate the growth and evolution of campus public
safety as described by Sloan (1992a). However these variables are not included in the
explanatory analyses to follow for one or more of several reasons. First, the distribution
of some of these variables is such that there is little variance within the item. For
example, virtually all departments employ one-officer motorized patrols, and almost none
employ two-officer bicycle patrol. thus, because of homogeneity within the item, it is
seen as ineffectual within a larger model. Second, some variables, such as number of non-
sworn public safety department employees have no theoretical purpose within a
Situational framework to be included in explanatory analyses. Because those variables

previously discussed at some length in this section are deemed more germane to the

67

research questions being examined, these later descriptive items are not included in later

multivariate analyses, but are rather presented points of reference.

Limitations of the Data
The possible dilemmas associated with utilizing existing data are well

documented in the literature on research methodology in the social sciences in general,
and in criminal justice specifically (e.g. Babbie, 1992; Monette, Sullivan, and DeJong,
1994; Taylor, 1994; Maxfield and Babbie, 1998). Most such problems arise when agency
data is utilized in a scientific context, for which the original data collection procedure was
not intended. Monette et a1. (19941191) explain:

“When using available statistical data, remember that most such

data were not collected for research purposes--or at least not for the

specific research questions for which you now intend to use them.

They were collected to meet the needs of whatever agency,

organization, or researcher originally collected them, and you are

limited by the form in which the data were collected.”
This problem is especially acute in evaluation studies or studies of agencies/organizations
who compile data for descriptive or record-keeping purposes only, and without the
intention of such material being analyzed scientifically. In the present research, because
the data set was obtained from an agency within the United States Department of Justice

specializing in empirical research, the assumption is made that the data will lend itself

easily to advanced statistical analysis.

68

Missing Data

A problem that is somewhat related to using data compiled by non-researchers is
that of missing data, or an “unclean” data set resulting from poor record-keeping,
improper coding, or human error in data processing. Indeed, many research projects have
fallen victim to the aforementioned issues by producing spurious results because an
unclean data set was employed. Even seemingly trivial errors within data sets may
become magnified when the data is re-analyzed or used for a purpose other than initially
intended (Kiecolt and Nathan, 1985). Some studies have been salvaged through efforts to
fill in missing data as completely as possible or by checking data codes to insure
correctness, but often a poor data set requires a re-assessment of the entire project at
hand. It is rather easy to understand why a poor data set would lead to inaccurate results
by face value alone, but even marginal data sets may not provide needed results, as most
advanced statistical applications assume complete, error-fi'ee data, thus producing results
which are not indicative of the research hypotheses at hand, rather they are reflective of
poor data management.

Again, because of the nature of the agency which first collected the data, it is safe
to assume that, since experienced researchers produced and complied the data being used
in this dissertation, the amount of error within the data set has been held to a minimmn.
Indeed, while a few cases have missing data, the amount of missing responses is minimal,
and as such should not affect subsequent analyses or the validity of conclusions drawn
from these analyses.

However, there are instances where institutions did not respond to certain sections

69

of the LEMAS questionnaire if the items posed were not germane to the institution. For
example, if a school outsourced for its protective services, the section of the LEMAS
survey dealing with equipment provided for Officers would not be relevant, and thus
would be counted as non-response (see Reaves and Goldberg, 1996). The co-coordinators
of the original BJ S project differentiate between non-response and missing data, as the
two forms of “incomplete response” differ qualitatively. It may be argued that missing
data termed non-response because the items are not applicable to the respondent do not
compound sampling error, as do missing responses for which no explanation is given.
Therefore, in this dissertation, for multivariate analyses non-response vs. missing data is
taken into account. The proportion of items simply left blank is significantly less than
those responses left blank because they were not applicable to the respondent institution.
The amount of data “missing” is not significant enough to affect the accuracy of results
generated from the data set.

Although the possibility of human error is present in all such data collection and
processing, data sets produced from research agencies are normally considered to be
much more valid than data sets produced by non-researchers. Additionally, because the
data being used in the present research has already been minimally analyzed (by
generating summary statistics) with success, the potential problems inherent with using

an unclean data set do not seem to apply to the present research.

Validity of Utilizing Existing Data
While the advantages to using existing data are thought by many researchers to

outweigh the negative aspects, one of the more troubling aspects of secondary analysis is

70

the compatibility of the data set with research questions or hypotheses specified by the
researcher. Kiecolt and Nathan (1985:14) provide an excellent summary of this dilemma:
“Even [data from] surveys of high quality may have measurement
problems. Invalidity is of concern to the extent that survey items
are imprecise measures of the concepts a secondary analyst has in
mind, or that the variables have been poorly operationalized...
[S]econdary analysts must frequently make do with measures that
are not precisely those desired. Often this results in criticism from
peers for lacking hypothetically perfect indicators, or for proceeding
atheoretically with research.”

The debate over whether secondary data or Official statistics may be applied to
later research questions has been ongoing for many years in social science. In
criminological research, reliance on such information bases has become commonplace,
and as a result has fueled the fire concerning the validity of results generated from
existing data sets. More commonly, some of these criticisms hold true for those projects
employing available statistics for related, but dissimilar, research questions other than the
original. Because a fair proportion of existing data are not collected for scientific
purposes, many scientific research questions are immediately suspect in terms of their
construct and external validity. An unfortunate, albeit frequent, pitfall for many
secondary analysts occurs when the available data demand re—definition of key concepts
or hypotheses in order to fit the data as it exists. Monette et al. (1994:195) state:

“The more the definitions are changed, the more the validity of
the measures is called into question. The operational definitions
may be changed so drastically that they no longer measure the

theoretical concepts they were intended to measure.”

Perhaps the most cogent synopsis of validity issues with secondary data comes

71

from Monette et a1. (1994: 194-195), where they state that, in utilizing other primary
research methods, issues of possible validity problems may be addressed before or during
the actual collection of the data, and modifications may be attempted to head off threats
to validity. However, when available data are used, validity problems cannot be
considered and resolved before the data are collected, and thus the problems become

entrenched.

Criticisms of Law Enforcement Statistics as an Indicator of Overall Criminali

The shortcomings of official crime data, such as the Uniform Crime Reports, are
well documented, and have been written about and debated for decades (e. g. Chambliss
and Nagawasa, 1969; Skogan, 1974; Jackson, 1990). The most vocal criticism of the
F BI’s measuring rod for crime is that it is sorely lacking in content validity, especially
when it is considered that, for less serious offenses, only the number of crimes solved or
“clear ” is reported. By the most optimistic estimates, the UCR is indicative of only
about half of less serious offenses (Part II index crimes) that are perpetrated in any given
calendar year.

The picture for more serious crimes seems brighter. Gove, Hughes, and Geerken
(1985) found that a majority of Part I Offenses are reported to law enforcement agencies,
and subsequently to the FBI, because the nature of these offenses often requires police
intervention. Because police departments are required to submit to the FBI all reported
Part I offenses, it appears that overall numbers of such offenses are a fairly valid indicator

of the extent of serious crime occurring in any given jurisdiction. Specific to this

72

dissertation, because the LEMAS questionnaire sought information about only Part 1
offenses, the unreported or “dark side” of crime is not of central concern is it might be if
data for less serious crime were the focus.

Another common criticism, especially in years past, regarding the UCR and other
police department statistics is that of misinterpretation and inadvertently erroneous
reporting of crime figures (e. g. Selke, 1977). When the UCR program was begun by the
FBI, definitions of certain offenses were established. In those early years of reporting,
police departments conunonly defined some crimes differently than the FBI, thus leading
to the production of inaccurate statistics. Over time, the standard definitions of offenses
have become commonplace, and such errors are less common among state and municipal
police departments. However, the Campus Crime and Security Act of 1990 has mandated
the publication Of crime statistics for interested consumers beginning in 1992, and these
reporting errors may become manifest again with agencies/institutions that may be
unfamiliar with criminal definitions put forth by the FBI. Because many institutions have
not collected crime data consistent with the guidelines for the UCR, there exists the
possibility that some crimes may be under- or over- reported as a function of
misinterpretation of definitions. This problem does not seem to be as serious in this data
set, as the definitions of the eight Part I Offenses are, ostensibly, clear to anyone affiliated
with public safety on a college campus. Misinterpretation cannot, however, be completely
discounted as a potential source of measurement error.

A third problem associated with official crime statistics compiled by law

enforcement agencies is that of intentional under-reporting of criminal activity occurring

73

within a particular jurisdiction. Although the Campus Crime and Security Act of 1990
was seen by most observers as a commonsense piece of legislation which would benefit
college students, parents, and others affiliated with post-secondary institutions, there was
vocal opposition from administrators at institutions in urban and/or high crime areas who
asserted that a good portion of crime occurring on their grounds did not involve students
either as victims or perpetrators, and thus the mandatory publication of gross crime
figures could wrongly paint these institutions as patently unsafe to attend. Such a label
might sway students to attend a different institution for safety reasons, thereby
introducing a “non-academic” criterion into the student’s decision as to which school to
attend.

As a result, there have been allegations at institutions such as MIT and the
University of Pennsylvania (schools which are highly prestigious and compete with other
similar institutions for the best students, but which are located in rather crime prone
areas) that the campus public safety departments have under-reported criminal activity on
campus so as not to dissuade potential students and parents away from their institutions.
Such intentional misrepresentation of crime statistics would be in violation of the
Campus Crime and Security Act of 1990, and would perhaps be difficult to prove. But as
is the case in any type of survey research, the possibility of respondent dishonesty as

potential measurement error exists.

Units of Analysis

Sometimes even data sets which seem methodologically sound, have no problems

with missing data, and appear to fit nicely with the researchers’ hypotheses, still may not

74

lend themselves to utilization in a given project because of problems associated with units
of analysis. This problem is especially acute when researchers wish to be more detailed in
a study, and make assumptions about specific cases, when the data is actually measured at
the group, or aggregate, level. Besides the Obvious problems of committing the ecological
fallacy (see Maxfield and Babbie, 1998), the results gleaned from such analyses are
suspect to the same shortcomings as data which do not fit specific research questions. In
this dissertation, the research questions pertain to institutions of higher learning in the

aggregate, and the unit of analysis of the available data set is the same.

CHAPTER IV
ANALYSIS AND RESULTS
Descriptive Statistics
A fundamental element of any research project is to provide summary statistics of
variables pertinent to subsequent explanatory analyses. As mentioned in the preceding
chapter, the LEMAS questionnaire contains of multitude of variables, only some of
which apply to the research questions being investigated in this dissertation. The
variables that are to be utilized can were subdivided into four distinct sub-headings:
extent of serious criminality at the schools, characteristics of the institutions surveyed,
crime prevention strategies, and campus safety department characteristics and functions.
Frequency distributions of these variable groupings are presented below, beginning with

the dependent variables.

Extgpt pf Seripus Criminality

The LEMAS survey collects information on UCR Part I offenses occurring on and
reported to campus public safety departments. As previously noted, prior research has
indicated that relatively minor Ofienses such as alcohol violations and vandalism are the
most prevalent offenses officers deal with on campus, and that campus crime rates are

generally significantly lower than those in the surrounding community (e.g. Bromley,

75

76

1992; Bromley 1995a). But these studies acknowledge that the gross numbers and rates
of more serious crimes seem to be increasing at a faster rate proportionate to the
surrounding community, thus warranting an investigate of the factors influencing Part I
Offending nationwide at larger colleges and universities.

Table 1 presents the gross number of Part I crimes reported by the participating
schools, the average (mean number of offenses per institution), the rate of the various Part
I Offenses using mean enrollment as a standard, and rate of Part I offenses using mean
resident population of the campus as a standard. The mean number of students enrolled at
the institutions surveyed is 10,783, and the average number of on campus residents is
2,538 (these statistics are presented below in Table 3). Using these figures, two different
rates of serious criminality may be calculated for comparative purposes. The table below
clearly indicates that the nation's colleges and universities were not battling an
extraordinary amount of serious violent crime in 1994. Six hundred-seven responding
institutions reported only 16 homicides during the calendar year, and only one institution
had more than one homicide occur on campus (one institution reported 2 homicides).
Further, when a murder rate is calculated using either student enrollment or resident
population as a standard, this figure indicates almost zero risk of homicide while on
campus, either as an enrolled student or resident of campus housing. The most prevalent
type Of serious violent offending reported to campus safety department is aggravated
assault. Unfortunately, the UCR does not disaggregate domestic assaults, stranger-on-
stranger-assaults, etc. fiom the more generally category, thus we are unable to determine,

from this data, the context of these assaults. Regardless, schools in the study averaged

77

Table 1. - Part I Offenses and Rates of Ofl'ending at Participating Institutions, 1994

 

 

 

 

 

Offense # of crimes reported (mean # per campus) # crimes # crimes
enrollment res. pop.

(mean = 10,783) (mean = 2,538)

Murder/Manslaughter 16 (.03) .001 .006

(N=607)

Forcible Sex Offenses 831 (1.37) .077 .327

(N=608)

Robbery 1213 (2.00) .112 .478

(N=606)

Aggravated Assault 2471 (4.06) .229 .974

(N=609)

Burglary 17,370 (28.43) 1.61 6.84

(N=61 1)

Larceny/Theft 124,681 (221.46) 11.56 49.13

(N=563)

Motor Vehicle Thefi 5,017 (8.28) .465 1.98

(N=606)

Arson 784 (1.43) .073 .309

(N=550)

All Violent Offenses 4531 (7.48) .420 1.79

All PIOperty Offenses 147,852 (268.82) 13.71 58.26

All Part I Offenses 152,383 (277.06) 14.13 60.04

 

 

78

only about four aggravated assaults on campus. This translates into rates of less than one
assault per student enrolled or campus resident (.229 and .974 respectively).

As expected, the rates of Part I property offenses were correspondingly higher
than those for violent crime on campus. Larceny/Theft continues to be the most oft-
reported offense to campus safety departments. In 1994, a total of 124,681 such crimes
were reported. Interestingly, only 563 schools reported the number of overall larcenies to
B] S in the survey. This could perhaps reflect a reluctance on the part of campus police
chiefs to divulge the true number of thefts their agency receives, as such information is
required to be tabulated as stipulated by the Campus Crime and Security Act of 1990. Of
course, another possible explanation is that the requested information was not readily
available. Either way, it is somewhat surprising that the response rate for this particular
offense was significantly lower than for the other seven Part I offenses (563 vs. 606-611).

The mean number of larceny offenses was slightly higher than 221, and the
corresponding rates totaled just more than eleven per enrolled student and just over 49 per
campus resident. Schools reported raw offense numbers for the other Part I property
offenses (Burglary, Motor Vehicle Theft, and Arson) to be significantly lower than those
for larceny, and in fact, the calculated rates for the remaining three property offenses
seem to be comparable to the rates of serious violent criminality, although the rates of
burglary and motor vehicle theft were slightly higher.

Taken together, the total number of violent offenses occurring on the participating
campuses in 1994 was 4531, with a mean rate of about seven and one-half serious violent

offenses per school. This translates to a rate of .420 violent offenses per enrolled student,

79

and 1.79 violent offenses per campus resident. The total number of serious property
crimes occurring at the institutions in 1994 was 147, 852, approximately 32 times greater
than the total number of serious violent crimes. Campuses averaged almost 269 serious
property crimes in 1994; a rate of 13.71 per enrolled student and 58.26 per resident. In
sum, the schools reported 152,3 83 total index crimes for 1994, an average of slightly
more than 277 per campus.

These rates were calculated using the most conservative reporting figure for each
of the four serious violent crimes. For example, 607 institutions reported the number of
homicides occurring on campus. This was the lowest number of institutions reporting for
either of the four violent crimes, thus this number was used as the divisor in calculating
total violent crime rates. Similarly, the most conservative estimate was used to calculate
property crime rates as well. Also, 550 institutions reported the gross number of arsons
occurring within their borders, therefore this statistic was used as a standard for both
Total Property Crimes as well as Total Part I offending . Of course, using the most
conservative figure slightly inflates the overall crime rates, however, given the difference
in the number of schools reporting figures for each crime, it is perhaps the most valid
method of calculation.

Using data fiom Bromley’s (1995a) study which compared college and university
crime rates to crime rates in the community in which the school is located, enables a
comparison of campus crime rates in 1991 and 1992 with crime figures gathered by B] S
in the LEMAS questionnaire during 1994. Bromley (19953) collected UCR data on Part I

offenses occurring on campus grounds for the two or three largest schools in the

80

contiguous lower 48 states for 1991 and 1992. At that time, compliance with the Campus
Crime and Security Act of 1990 was not yet mandatory, and, as is the case presently, not
all institutions reported crime data to the federal government for inclusion in the UCR.
Thus, data for certain institutions was not available. For other institutions, some reported
their crime statistics for 1991 and not 1992, and vice versa Therefore, Bromley (1995a)
used data that was available for 1991, 1992, or averaged the two years crime figures
(when available) to use in calculating the overall index crime rate, the violent crime rate,
and property crime rate for each campus (see Bromley, 1995a: endnote 1).

Similarly, because the LEMAS survey administered by B] S was not a formal
request for crime data in compliance with the Campus Crime and Security Act of 1990,
some institutions in the present data set also failed to supply requested data, therefore
direct comparisons for a handful of institutions was not feasible. What follows in Table 2
is a comparison of index crime, violent crime, and property crime rates for those
institutions supplying complete data for both the BJ S LEMAS questionnaire and to the
FBI, which was subsequently used by Bromley (1995a). All rates are standardized by
dividing the raw number of offenses per 1,000 enrolled students. Total index crime rates
were rounded to the nearest whole number by Bromley (1995a), and for comparative
purposes, the calculated total index crime rate for 1994 was also rounded in the same
manner. Rates for violent index crimes and property index crimes remained as calculated
to the hundredths decimal in both Bromley’s (19953) study and the present research.
Although comparisons may be made between individual crime rates for each institution

represented in Table 2 (see Bromley, 1995azl35-142), such a procedure was not used in

Table 2.- Comparison in Crime Rates of the Largest Colleges/Universities between
1991- 1992 and 1994. (N=73)

 

 

 

Institution Total Index Violent Index Property Index
‘91-’92 ‘94 ‘91-’92 ‘94 ‘91-’92 ‘94

U. Alabama 29 26 1.01 .51 29.09 25.36
U. of Arizona 36 36 .45 .45 35.56 35.18
Arizona State U. 31 30 .45 1.28 30.32 29.16
U. Arkansas-Fayetteville 26, 19 .35 .42 25.76 1 8. 14
U. Arkansas-Little Rock 8 13 .19 .75 7.70 12.00
Boston U. 35 31 1.24 1.75 34.06 29.04
Brigham Young U. 17 16 .19 .16 16.50 15.39
U. Califomia-Los Angeles 41 46 1.17 .44 40.12 45.69
California-Long Beach 15 10 .39 .22 14.77 9.79
Central Connecticut State 11 8 .14 .00 10.50 7.50
U. of Cincinnati 37 27 .74 .58 35.89 26.65
Clemson U. 34 34 .85 1.02 33.04 32.81
U. of Colorado 35 32 .47 .32 34.72 31.99
Colorado State U. 30 11 .96 .95 29.44 10.08
U. of Delaware 41 35 .74 .78 40.43 34.60
U. of Florida 42 39 .81 .96 41.04 38.02
U. of Georgia 27 21 .30 .52 26.96 20.17
Georgia State U. 18 23 .09 .47 18.30 22.79
U. of Illinois 25 23 .42 .93 24.60 21.72
U. of Indiana 32 30 .53 .79 31.77 28.72
Indiana U. of Pennsylvania 14 15 .90 .50 14.38 14.72
Iowa State U. 27 30 .83 .28 25.91 29.32
U. of Kansas 34, 36 1.05 .46 32.64 35.56
Kansas State U. 26 18 .27 .15 25.32 18.10
Kean College (NJ) 19 18 .45 1.35 18.87 16.60
U. Nebraska-Kearney 14 17 .00 .00 13.64 17.03
U. of Kentucky 42 34 1.14 1.01 41.28 33.38
Louisiana State U. 29 32 .64 1.84 28.45 30.59
U. of Louisville 21 26 1.20 1.23 19.55 24.66
U. of Maine 35 28 .16 1.84 35.22 28.21
Marshall U. (WV) 12 12 .12 .08 11.71 12.19
U. of Michigan 67 54 1.57 1.33 65.73 52.73
Michigan State U. 47 33 1.17 .83 45.94 31.75
Middle Tennessee State U. 20 17 .39 .35 19.97 16.22
U. of Minnesota 26 27 .29 .39 25.44 26.72
Mississippi State U. 28 28 .19 .07 27.88 28.14
U. of Missouri 29 29 .56 .45 28.46 28.35
U. of New Hampshire 24 24 .95 .57 22.94 23.45

Table 2. (cont’d.)

82

 

 

U. of New Mexico 36 29 .66 1.01 34.80 27.71
New Mexico State U. 39 33 .55 .43 38.05 32.24
U. of North Carolina 34 34 1.43 .95 32.06 33.49
North Carolina State U. 34 32 .52 .32 32.88 29.99
U. of North Dakota 28 20 .38 .77 27.55 19.68
Northern Illinois U. 28 17 .57 .89 27.09 16.58
Northeast Louisiana State U. 24 25 .42 .60 23.70 24.37
Northeastern 17 16 .81 .60 16.17 15.06
U. of Nevada 31 27 3.14 .83 28.28 26.00
U. of Nevada-Las Vegas 19 23 .52 .32 18.31 22.23
Oklahoma State U. 16 13 .17 .37 15.39 12.64
Ohio State U. 38 32 .80 .71 36.68 31.39
Purdue U. 34 25 .72 1.59 32.99 23.24
Rutgers U. 27 23 .65 .48 26.50 22.97
U. South Alabama 4 16 .10 .24 4.02 15.58
U. of South Carolina 28 24 .96 .52 27.01 23.21
South Dakota State U. 18 18 .06 .52 18.05 17.33
U. of Southern Maine 10 11 .00 .10 10.96 10.50
SUNY-Albany 29 35 .97 .72 27.78 34.25
SUNY-Buffalo 27 30 .45 .70 26.96 29.30
Texas A&M 20 17 .20 .31 19.84 17.10
U. of Tennessee 36 23 .82 1.06 31.70 21.90
U. of Texas 20 17 .23 .08 19.92 16.46
U. of Utah 34 24 .53 .41 33.07 23.94
U. of Vermont 50 24 .54 .28 49.66 23.55
U. of Virginia 30 35 .34 .61 26.39 34.86
Virginia Commonwealth U. 37 37 .83 .55 36.53 36.79
Virginia Tech 23 16 .40 .69 22.71 15.25
U. of Washington 32 32 .33 .44 32.09 31.94
Washington U. (MO) 17 38 .17 1.54 17.09 36.07
Western Michigan U. 24 21 .72 .45 23.25 21.01
West Virginia U. 23 17 .05 .34 23.13 16.81
U. of Wisconsin 26 21 .22 .28 25.93 21.05
U. of Wisconsin-Milwaukee 18 20 .31 .25 17.57 19.78
U. of Wyoming ‘ 28 23 .30 .08 28.17 22.89
Mean 27.60 25.14 .607 .623 27.07 24.52
Standard Deviation 10.47 8.89 .481 .431 10.15 8.71

Test of Significance F= 4.27 p=.000*"'* F= 1.30 p=.290 F= 1.84 p =.537

 

 

at“ [K .01

*p<.10

83

the present research. Because the factors which may be responsible for a difference in
crime rates at individual institutions are beyond the scope of this dissertation and thus not
included in any subsequent analyses, tests of statistical significance for the difference in
crime rates between 1991-1992 and 1994 would only demonstrate if that difference was
significant, not why it was significant. Therefore, if a measure of the difference in crime
rate is desired, a percentage increase or decline can easily be computed by utilizing the
algorithm [a + b/ a] x 100. For example, the calculated percentage difference between the
total index crime rate, the violent index crime rate, and the property crime rate for the
University of Alabama would be -10.3, -49.5, and -12.8 respectively.

A test of significance is possible, however, when examining the cumulative
difference in crime rates at the 73 schools. Because the rates in both Bromley’s (1995a)
study and those from the LEMAS data in 1994 were both calculated using the same
standardizing denominator (# crimes/ per 1,000 enrolled students), the difference in rates
may be compared directly, similar to the comparison of overall mean scores. Because just
two means are subsequently compared, the F statistic (F raenkel & Wallen, 1993) is
interpreted to show significant differences between means of each of the three crime
rates. It is interesting to note that when examine the aggregate of overall index crime
rates, there is a significant difference in serious on-campus crime between 1991-1992 and
1994. Table 2 shows that the mean number of serious offenses occurring at this sub-group
of large institutions decreased by more than two offenses dming this time span (27.60 vs.
25.14 serious ofl‘enses per campus; F= 4.27, p=.000). However, when the offenses are

dichotomized into violent and property offenses, there does not appear to be an

84

appreciable change in the rates of Part I offending for either type of serious criminality.
Such results can be difficult to interpret. While overall crime rates appear to have
decreased, it seems that neither the level of violence nor the amount of property offenses
has decreased sufficiently enough, as subgroupings, to have registered as an overall
significant decline within categories. Therefore the results from Table 2 must be read
with caution, as it would most certainly be premature to suggest that the amount of
serious crime at these colleges and universities is decreasing. As mentioned previously,
the purpose of this dissertation is not to determine which factors correlate with levels of
crime at individual institutions. The above table simply represents the second wave of
mapping crime rate trends at the universities first examined by Bromley (1995a).
Furthermore, because it is normal for crime rates to fluctuate on a yearly basis,
and in different directions depending on which indicator is chosen to measure criminality,
it is very difficult to ascertain which factors (demographics, impact of crime prevention,
policing tactics, etc.) are responsible for the rising and falling over crime rates over time
(Cole, 1995). Lastly, a precipitous drop or increase in the levels of serious crime would
not be expected given that the time lag in between measurements is a relatively scant
two- to- three years. Even supposing a radical shift in institutional philosophy (i.e.
proactive law enforcement, design initiatives to prevent criminality) regarding criminal
activity on campus, the results of such initiatives would not likely be reflected in actual

crime statistics over such a short period of time.

85

Institutional Characteristics

Table 3 outlines the demographic characteristics of the 680 schools selected by
BJ S for inclusion in the study in terms of student enrollment, size of the resident
population, number of employees, and the nature of the criminal justice curriculums at
each institution. Sloan’s (1992a) earlier study of campus criminality found the variables
student enrollment and on campus resident population to be significantly associated with
gross campus crime statistics, thus such demographics are presented here and included in
subsequent regression models, as either standards of crime statistics or control variables.

When the Bureau of Justice Statistics originally surveyed the institutions, only
those schools with 2500 or more students enrolled (including full- and part-time) were
selected. As the original authors state:

“The 1995 Survey of Campus Law Enforcement Agencies was
mailed to the campus law enforcement agency at each U.S. 4-year
university or college that had 2,500 or more students. The U.S.
military academies, graduate or professional schools, and schools
operating on a for-profit basis were excluded.” (Reaves & Goldberg,
1996).

It is worth noting that the re-grouping of the attributes of the variables utilized in
this study was done for purposes of clarity in presenting information in tabular form.
Although this re-grouping lowers the level of measurement to ordinal in the summary
tables, data, where applicable, remained at the original ratio level in subsequent
regression analyses. Referring to Table 1, of the campus safety directors who provided
complete information, just under half (47%) reported that their school employed less

than one thousand individuals. One-third of all respondents indicated that between 500

and 999 persons were on the college or university payroll.

86

Table 3. - Demographic Characteristics of Institutions Included in the Study.

 

 

Enrollment (N=581) ' %
2500 - 4999 31
5000 - 9999 30
10000 - 14999 17
15000 - 19999 8
20000 - 24999 5
25000 - 29999 5
30000 + 5

Resident Population (N=357)

0 - 999 37
1000 - 2499 31
2500 - 4999 20
5000 + 13
Number of Employees (N =43 0)
0 - 499 14
500 - 999 33
1000 - 1499 13
1500 - 1999 9
2000 - 2499 7
2500 - 2999 4
3000 - 3499 2
3500 - 3999 3
4000 - 4999 3
5000 + 12

Nature of Criminal Justice Program at the Institution (N=663)

None 51
BS/BA in Criminal Justice Only 29
MA/MS in Other Discipline w/CJ concentration 6
MA/MS in Criminal Justice 12
PhD in Other Discipline w/CJ concentration 9

PhD in Criminal Justice 12

 

 

87

With respect to the size of the student body, few institutions could be
characterized as large “mega-universities”. Only fifteen percent of the schools in this data
set enrolled more than twenty thousand students. More typical were smaller universities
and larger colleges which enrolled between twenty-five hundred and ten thousand
students (61% of the total). Less campus law enforcement directors responded to the
question pertaining to the size of the resident population at their respective institutions.
This seems to indicate that either the school had no resident pOpulation (i.e., was a
“commuter” school), or the respondent did not know and/or was unable to locate the
information requested. Of the three-hundred and fifty-seven responses, most schools
report a resident population numbering less than 2500 (68% of the total), while one-third
of the total have relatively large on-campus populations of 2500 residents or more.

A final variable included in Table 3 regarding the extent of a criminal justice
curriculum at the schools was collected specifically for this dissertation using the
secondary data sources outlined in the preceding chapter. Somewhat surprisingly, almost
a full one-half of all schools included in the study were found to offer at least a bachelors’
degree in criminal justice. This figure is all the more impressive when one considers that
this figure does not include those schools which offer an undergraduate concentration in
criminal justice within another academic major. In other words, only those schools listing
criminal justice studies or criminology as an academic major were categorized as offering
the ‘BS/BA in Criminal Justice’. It is quite likely that a number of other institutions offer
courses pertaining to criminal justice in other disciplines such as sociology, political

science, public administration, or urban affairs to name a few. The relatively large

88

number of schools offering degree programs in criminal justice does, however, seem to
reflect the growing trend of interest in crime and justice as an academic field; one that
had previously been relegated to institutions of lesser quality and community/junior
colleges (e.g., Sullivan, 1994).

Table 4 presents descriptive statistics pertaining to the physical/spatial
characteristics of the campuses included in the BJ S study. The variables presented here
are those which correspond most directly with the theoretical tenants of environmental
criminology and routine activities theories. Scholars who have published research on
campus crime in the past twenty years or so have found correlations between campus
crime rates and physical characteristics such as acreage and campus setting (e. g., Fox and
Hellman, 1985; Sloan, 1992a).

According to the definitions provided by The College Handbook (1997), there is a
fairly equitable split in the numbers of campuses which could be classified as rural,
suburban, or urban in nature. It is worth repeating that the nature of the campus
environment is determined solely by the definition provided by The College Board and is
not necessarily a reflection of the size of the community in which the school is located.
Therefore it is poSsible to have a rural campus that is technically within the boundaries of
a large city, and conversely, an urban campus may exist within a suburban area or large
town. The ‘suburban’ definition of the college campus is perhaps the most identifiable
with the traditional vision of the college campus, that is a self-contained locale adjacent

to or within a larger town or municipality. This could explain why almost half of the 666

89

Table 4.- Physical Characteristics of Campuses Included in the Study.

 

 

Nature of the Campus Location (N=666)

Rural
Suburban
Urban

Size of the Campus in Acres (N=57 8)

0 - 49

50 - 249
250 - 499
500 - 749
750 - 999
1000 - 1249
1250 - 1499
1500 +

Number of Buildings on Campus (N=563)

0-24
25-49
50-74
75-99
100+

Miles of Campus Roadways (N=435)

0-9
10-19
20-29
30+

%

21
45

35

19
31
19

23

 

 

90

on campus, this is also not necessarily the case institutions for which information was
available through The College Handbook (1997) were classified in this manner.

It should be noted that the definition of campus location differs significantly from
that utilized by Sloan (1992a), as in that study the campus location was synonymous with
the size of the population of the area in which the school was situated. The other variables
in Table 4 which describe the physical environments of the institutions included in the
study are fairly self-explanatory. But it should be emphasized that the variables ‘Number
of Buildings’ and ‘Miles of Roads’ are not necessarily correlated with the student
enrollment at that particular school. One only need visualize the tightly compact, urban
campus to verify this assertion. Furthermore, while it is expected that schools with more
students, employees, etc. will also have a correspondingly high number of buildings

given that some buildings can be larger and contain more facilities than others.

Crime Prevention Strategies

One provision of the Campus Crime and Security Act of 1990 is for colleges and
universities to compile and provide information pertaining to crime prevention strategies
in place at the institution. As such, there has been a growing number of crime control
innovations put into action at schools around the nation. It is safe to say that, given the
specialized populations served by campus public safety deparunents and the specter of
civil liability which is always present, campus police departments are perhaps the most
proactive and prevention-minded policing bodies in existence. Table 5 presents the range

of crime prevention strategies which are in place at schools around the country and also

91

shows the rate of participation/implementation of such strategies at the responding
institutions. Eighty-five percent of all institutions from which information was requested
reported that their in-house campus public safety department performed routine patrol
functions on school property. Furthermore, those institutions whose public safety
departments provided patrol functions also more than likely provided this service on a 24-
hour basis (83% of 581 institutions providing patrol services). While it is not possible to
compare these figures with earlier reports, as this is the first time the LEMAS survey was
distributed nationally to college and university departments of public safety, the statistics
do indicate that a vast majority of institutions have placed a high priority on providing
security and order-maintenance fimctions for their constituents.

Table 5 also presents information regarding the type of patrol strategies which
have been adopted by the institutions in this study. Motorized patrols, usually with one
officer only, continue to be the preferred method of patrolling campus grounds. However,
a significant number of campus safety departments utilize foot patrols on campus as well,
with 62 percent assigning at least one officer to such a function on weekdays. This may
be reflective of the popularity and initial successes of community policing at many
schools nationally (e.g. Trojanowicz et al., 1988; Benson, 1993; Powers, 1996). Or
perhaps foot patrols are more suitable for certain campuses, such as those without many
automotive thoroughfares or at those primarily urban in nature, where facilities are more
densely packed, and areas of pedestrian traffic are not visible fi'om roadways. The same

argument could also be made for the almost 30 percent of schools reporting at least one

92

Table 5.- Crime Prevention Program Strategies and Implementation

 

 

Program Strategy/Policy

%

 

Campus Safety Departments Performing Routine Patrol (N=680)
Yes
Campus Safety Departments Providing 24-Hour Patrol (N =581)
Yes
Types of Patrol Implemented by Campus Safety Departments (N=578)

Auto, 1 Oflicer, Weekday
Auto, 1 Officer, Weekend
Auto, 2 Officer, Weekday
Auto, 2 Officer, Weekend
Foot, 1 Officer, Weekday
Foot, 1 Officer, Weekend
Foot, 2 Officer, Weekday
Foot, 2 Officer, Weekend
Bicycle, 1 Officer, Weekday
Bicycle, 1 Officer, Weekend
Bicycle, 2 Officer, Weekday
Bicycle, 2 Officer, Weekend

Campus Oflicers Armed (N=680)
Yes, Always
Yes, Sometimes

No

Emergency 911 System (N=581)

911 Expanded/Basic
Emergency Blue Light System (N=581)

Yes
Avg. # = 33.5 (SD = 50.66)

85

82

87
86
17
16
62
59
11
ll
28
22

60

36

77

93

Table 5. - (cont’d)

 

Communications Equipment (N=581)

Portable Radios
Mobile Vehicle Radios
Cellular Phones

Base Station Radios

Computer Usage/Access (N=5 78)

Mainfiame Computer
Mini-Computer

PC

Laptop

LAN System

Car-Mounted Digital Terminal
Hand-Held Digital Terminal
Other

Agencies Overseeing Specific Crime Prevention Initiatives (N=567)

Victim Assistance

Crime Prevention Education
Bias-Related (Hate) Crimes
Student Security (Mule) Patrol
Date Rape Prevention
Stranger Rape Prevention
Self-Defense Training
Alcohol Education

Drug Education

Other

85
68
52
78

37
85
24
60
68
60
40
53
50
12

 

 

94

officer assigned to bicycle patrol, and thus campus safety departments seem to be open to
methods of patrol other than traditional squad cars cruising the streets.

Somewhat surprisingly, the amount of patrol at the participating schools seems to
diminish slightly on weekends; a time that, traditionally, sees the most criminal activity
in towns and cities. While the differences are not necessarily substantial, it is a curious
notion to consider. At many schools the amount of constituents on campus is not as great
as during the week, since usually classes are not in session. Therefore, faculty, a majority
of employees, commuter students, and some resident students are not on campus in
numbers as great as one would see on a weekday when a full schedule of classes were
taking place. This could very well explain the slight declines in weekend patrol.
However, earlier investigations into campus violence and serious criminality have
identified alcohol as a major contributing factor to such incidents (e. g. Pollard and
Whitaker, 1991; Smith and Fossey, 1995), and because students undoubtedly tend to
consume more alcohol over the weekend than during the week, it would be expected that
overall incidents of violence would peak between Friday night and Sunday morning.

One variable included in Table 5 has perhaps been the most hotly contested
campus security policy over the past several years, that is the arming of campus public
safety officers (e.g., Tillinghast, 1996; Hummer et al., Forthcoming). While arming
policies are fairly standard practice at larger universities and in urban areas,
administrators at smaller schools and those in suburban or rural locales have resisted the
adoption of general issue sidearms for officers. In the present study, sixty percent of the

institutions surveyed reported that their ofiicers are armed at all times while on duty. an

95

additional four percent report that officers are issued firearms in certain situations. Only
slightly more than one-third of the institutions indicated that guns were not issued to
campus safety officers at all.

Of the 581 institutions which provided complete information regarding the above
items, nearly two-thirds reported having in place either an expanded or basic 91 1
emergency system. Similarly, over three-quarters of these schools maintained an
emergency blue-light system on campus. Typically, such a system consists of a number
of call-boxes placed at strategic (e.g. areas where criminal activity has been/may be
prevalent) locations about the campus. These call-boxes are wired directly to the
dispatcher at the department of public safety on campus. Most of these systems, which
enable the caller to talk directly with the dispatcher, also have a tracking system whereby
the dispatcher can trace the exact location the call is coming from, even if the caller is
unable to speak or is unsure of their location on campus. Of those schools maintaining a
blue light system, the mean number of call-boxes distributed around campus was about
thirty-three, however the calculated standard deviation of 50.66 indicates a fairly
substantial dispersion or range. That is, more than likely there are many schools with a
fairly small number of blue phones and many schools with a large number.

Concerning technology utilization at the participating institutions, the above table
indicates that 85 percent of reporting departments utilize portable radios, slightly more
than two-thirds use mobile vehicle radios, 78 percent of the departments have base station
radios, and just over half make use of cellular phones. Additionally, with respect to

computer technology, ninety percent of the departments have at least one personal

96

computer in use, slightly less than two-thirds utilize a mainframe computer, and one-third
are connected to local area law enforcement computer networks. Around twenty percent
of reporting departments utilize laptop and mini-computers, and a small number of
agencies have begun using hand-held and vehicle-mounted digital computer terminals.
These devices allow the officer to access records and information directly, thus saving
time because they are able to by-pass the dispatcher who would access information for
them. In addition, with officers being able to access information directly, the amount of
radio traffic is diminished, and dispatchers are able to focus specifically on incoming
calls for service. Hand-held and vehicle-mounted terminals are commonplace in many
municipal departments, but perhaps because of expenditure concerns, have not yet
become widely used on campus.

The final portion of Table 5 deals with the percentage of agencies who participate
in the various crime prevention initiatives Specified in the LEMAS survey. Regarding
specific programs, eight-five percent of schools providing information on crime
prevention initiatives indicated that they oversee a general crime prevention education
program. Sixty-eight percent sponsor a program aimed at preventing date or acquaintance
rape, 60 percent supervise the student volunteer, or “mule”, patrol and provide non-
acquaintance rape prevention classes. About halfof all reporting departments provide
drug and alcohol awareness seminars, forty percent supervise self-defense training, just
above one-third offer victim-assistance initiatives, 24 percent offer hate crime prevention
programs, and twelve percent oversee a variety of other prevention initiatives for the

campus community. Again, because there is not a standard with which to compare these

97

statistics, it is unclear as to whether this rate of participation is significantly higher or
lower than in past years. However, given the possible ramifications of not providing
adequate prevention strategies as they pertain to the Campus Crime and Security Act of
1990, it is probably safe to assume that the number of, and participation in, crime

prevention programs is increasing rather than diminishing.

Campus Safeg Dgpartment Characteristics

Most recent empirical work which has focused on campus policing reports
growing professionalism among departments, a proactive strategy which often includes
problem-oriented policing, and a general evolution of these departments such that they
now tend to be duplicates of municipal departments in form and function (Bordner &
Petersen, 1983; Trojanowicz et al., 1988; Sloan, 1992b; Johnson, 1994; Bromley, 1995b;
Lang, 1995). Of course, the needs of institutions of higher education are quite diverse and
the characteristics of the campus safety departments at larger colleges and universities
reflect this divergence. Table 6 presents campus safety departmental characteristics. The
selection criteria that are utilized by campus public safety departments appear to be very
similar to those selection methods utilized by municipal and state law enforcement
agencies. Over ninety percent of campus departments require interviewees to come in for
an oral interview and conduct backgrormd and criminal records checks on prospective
officers. Nearly seventy percent of the departments have prospective employees submit to
a medical exam, fifty percent to a psychological examination, and just over forty percent

of reporting institutions screen for narcotics usage. These statistics indicate that campus

98

public safety departments take careful measures in selecting the individuals who may one
day be responsible for patrolling campus grounds and buildings. Undoubtedly, the
selection process has become more advanced, perhaps because of the move fiom security
guards or watchmen to sworn law enforcement officers.

The professionalization of campus public safety is also indicated by the amount of
schools which, as part of their overall departmental policies and procedures, have written
policy directives for different situations which may arise during the everyday activities of
their campus officers. At least half of all reporting departments have written directives
regarding the use of deadly force, handling citizen complaints, handling domestic
disputes, motor vehicle pursuits, and officer codes of conduct and appearance. Also, at
least fifty-two percent of all schools possess written directives on relations with on- and
off-campus agencies such as residence life officials, victim assistance services, and law
enforcement agencies from surrounding communities. It stands to reason that these policy
directives may be a part of an institutions overall crime prevention/response plan
stipulated by the Campus Crime and Security Act of 1990, as many schools have begun
to utilize a multi-departrnental response to disorder and violence on their campuses
(Sigler and Koehler, 1993).

Although campus public safety departments are continuing to move from the
“watch” system to a firll law enforcement model (Sloan, 1992b), significant differences in
, terms of personnel remain. Increasingly, the educational requirement for entry into the
law enforcement profession is shifting fiom a high school diploma to the two- and four-

year college degree (Egan, 1991). However, as of 1994, over two-thirds of the institutions

99

Table 6.— Characteristics of Campus Public Safety Departments.

 

 

Department Characteristic

 

Average Number of Department Employees (N=678)

Full time sworn
Part time sworn
Full time nonswom
Part time nonswom

Officer Selection Criteria (N=5 76)

Written aptitude test
Oral interview

Board interview
Background interview
Criminal record check
Physical agility test
Psychological screening
Medical exam
Polygraph exam

Drug screen
Other

Educational Requirements (N = 578)

4 year college degree

2 year college degree

Some college required

High school diploma or equivalent
No educational requirement

Other educational requirement

Classroom Training (N=497)
Yes

No
Avg. # of hours if yes = 326.6 (SD = 248.5)

15.7 (SD = 17.0)
.3 (SD = 6.5)
3.9 (SD = 19.3)
11.9 (SD = 21.6)

% Yes

39
94
61
95
94
36
56
69
10
41
4

ll
16
68

88
12

100

 

Table 6.- (cont’d)

Field Training (N=497)
Yes 91
No 9
Avg. # of hours if yes = 270.2 (SD = 311.8)

Arrest Powers of Campus Public Safety Officers (N =678)
Yes, state level 9
Yes, local level 66
Citizen arrest 1
Other 0
No arrest powers 23

Outsourcing of Campus Protective Services
Yes 25
No 75
Avg. % outsourced if yes = 31.7 (SD = 36.98)

Type of Outsourcing Contracted (N=l67)
Private security 70
Local police 22
Sheriff 4
State police 2
Other 2

Written Policy Directives for Special Circumstances (N=574) % Yes
Use of Deadly Force 81
Handling juveniles 60
Handling the mentally ill 51
Handling the homeless 16
Handling domestic disturbances 58
Citizen complaints 70
Pursuit driving 69
Off-duty employment of officers 63
Code of conduct and appearance 94
Employee counseling assistance 50
Relationship with other law enforcement agencies 70
Relationship with student judicial officers 55
Relationship with residence life ofiicials 56

101

Table 6.- (cont’d)

 

Relationship with victim/counseling services 52

 

 

sampled required new recruits to posses only the high school diploma or its equivalent,
while only thirteen percent required a two- or four-year college degree. These statistics
may indicate that municipal departments and campus departments are drawing new
recruits from different applicant pools, with the more educated recruit gravitating toward
the municipal positions. Further, the educational requirements of campus safety officers
may also be a function of organizational culture of the individual departments (e.g.,
security-oriented vs. sworn officers) or the size of the institution. In either case, it is still
somewhat surprising that institutions of higher education would not have entrance
requirements for their public safety departments equal to those of municipal or state
police departments.

Irrespective of educational requirements for new recruits, campus safety
departments, overall, seem to allocate a substantial amount of time to classroom and field
training. Of course those departments employing fully sworn officers have the benefit of
new recruits training with municipal police officers at an accredited police academy, and
this may result in the rather large number of hours of new officer training reported by
institutions. There also seems to be a rather wide range of training time for the schools
represented in the present research, as indicated by the relatively large standard deviations
for both classroom and field training (248.5 and 311.8 hours respectively).

Seventy-five percent of the institutions included in the present study indicated that

102

their officers retained either state or local arrest powers, while the remaining quarter held
only the arrest powers of the average citizen. This figure could possibly correspond to the
number of institutions which outsource their protective services to contract security
versus those which do not. As Table 6 demonstrates, 75 percent of the colleges and
universities supply their own campus security officers, while the other quarter contract
out for such services. Of those 25 percent that contract out for services, an average of
31.7 percent of campus protective services is provided by outside vendors at these
institutions. Private security firms are the most popular type of outsourcing used by the
schools contracting out (70 % of those outsourcing use such firms), while sworn officers
fiom local municipalities surrounding the campus, from nearby state police barracks, or
from the county sheriffs department supply 28 percent of the outsourced services. The
remaining two percent is supplied by other firms or agencies.

The final descriptive variable in Table 6 concerns the mean number of persons
employed either full- or part-time by campus public safety departments, in both a sworn
and non-swom capacity. Ostensibly, these averages are Significantly correlated with the
size of the institution served. In the aggregate, the institutions in this data set employed an
average of about 16 sworn officers on a full-time basis and about 1 sworn officer on a
part-time basis. These schools also employed an average of 14 civilian employees full-
time and about 12 civilians part-time. Again, because this data set represents the first
administration of the LEMAS questionnaire to campus safety departments, there is no
earlier information on department size with which to compare these figures. However, a

working hypothesis could be, given prior research detailing the more serious nature of

103

campus criminality, and the stipulations placed on institutions by the Campus Crime and
Security Act of 1990 to report gross numbers of crimes and prevention strategies, that
campus safety departments are growing in terms of personnel rather than declining.

In addition, because most campus safety departments have the burden of
protecting students, maintaining order, and fighting crime placed squarely upon them,
additional personnel may be needed to meet these growing demands. The possibility of
civil litigation for inappropriate responses to incidents on campus also forces campus
departments to become aware of accountability issues which do not concern their
municipal counterparts. Thus, the conclusion that campus police departments are subject
to a large variety of pressures when performing routine services seems valid.

The accountability issue is but one of many concerns of the campus law
enforcement director. As evidenced from Table 7 below, the public safety departments at
the larger campuses included in the present study perform a variety of duties, many of
which are very similar to those undertaken by municipal departments from surrounding
communities. These varied duties are performed for a rather distinct constituent base,
which very likely has conflicting perspectives regarding how laws, rules, and regulations
should be enforced in the campus environment. Looking at the specific duties listed on
the LEMAS survey, and presented below, it is clear that the evolution of campus public
safety departments has progressed as described by Sloan (1992b), such that these
departments have primary or exclusive responsibility for maintaining order, answering
calls for service, and enforcing legal statutes on school property. Whereas a few decades

ago, institutions reported emergencies or crimes to the police departments in the

104

municipality surrounding the institution, today campus public safety is the domain of the
institution itself, and therefore this sovereignty is reflected in the percentage of schools
which report primary responsibility for the duties which follow in Table 7. Primary
responsibility for reporting crime data and information on crime prevention programs in
place at the institutions rests with 95 percent of the schools included in the present
research. Meaning that a vast majority of campus safety administrators are cognizant of
the tenants of the Student Right-to-Know and Campus Security Act of 1990 as well as the
obligation placed upon them by this legislation. This finding may be interpreted as a
recognition by the institutions of the problems and possible ramifications of an
insufficient crime prevention strategy or an ill-trained campus safety department should a
serious crime (or crimes) occur which, in the eyes of the court, were preventable or whose
effects could have been ameliorated.

The broad range of duties listed in the table above also demonstrate the range of
responsibilities of campus public safety departments. These agencies are given primary
handling tasks ranging from building lock-up to the security of nuclear facilities, all the
while knowing that any breech or shortcomings in performing these duties leaves the
institution liable for monetary damages. This wide range of duties, coupled with the ever-
present possibility of litigation, places tremendous pressure on colleges and universities
to be prepared for numerous situations. The argument could be made that given the
context in which campus safety officers must perform their duties, and the specialized
populations they must oversee, makes policing the collegiate campus one of the most

challenging and diverse models of law enforcement in the United States.

105

Table 7.- Duties and Responsibilities of Campus Public Safety Departments

 

 

Duty/Responsibilities

 

Duties for Which Campus Public Safety Department has
Primary Responsibility (N =581)

Traffic law enforcement

Central alarm monitoring
Accident investigation

Training academy operation
Emergency fire services
Environmental health and safety
Fingerprint processing
Dispatching calls for service
Security for nuclear facility
Adherence to the Campus Crime and Security Act of 1990
Search & rescue

Key control

Stadium event security

Fire inspection

Parking enforcement

Building lockup/unlocking

Vice enforcement

Animal control

Parking administration

Campus switchboard operation
Arena/coliseum event security
Personal safety escorts
Hospital/medical center security
Fire prevention education
Traffic direction and control
Campus transportation system
Emergency medical services

Primary Enforcement of Drug Laws (N=5 80)

Yes
No

% Yes

84
80
88

19
26
50
90

95
29
38
70
31
85
85
56
35
72
22
67 '
86
12
31
89
25
36

%

84
16

106

 

Table 7.—(cont’d)

Participation in a Multi-Agency Drug Task Force (N=5 79) %
Yes 15
No 85

Departments Seizing, Eradicating, or Destroying Types of % Yes

Narcotics in 1994 (N=572)

Amphetamines l2
Barbituates . 6
Cocaine (Crack) 24
Cocaine (Powder) 23
Hashish 10
Heroin 4
LSD 22
Marijuana 83
Methamphetamine 10
Methaqualone l
Morphine 1
Opium 0
PCP 2
Synthetic/designer drugs 3
Other narcotics 5

 

 

107

The proliferation of narcotics in most every aspect of American society is also
reflected in Table 7. While only (fifteen percent of campus agencies participate in multi-
agency drug task forces, 84 percent are given primary responsibility for enforcing
narcotics laws. The final variable in the table shows that college campuses are not
immune fiom the darker aspects of social life, as 83 percent of all departments had seized,
eradicated or destroyed some type of is illicit narcotic during 1994. Not surprisingly, the
most common form of illegal drug on campus marijuana, however other so-called
“harder” drugs are finding their way on campus with relative frequency. Almost one-
quarter of the participating departments report the seizure or eradication of cocaine
(powder or crack) in 1994, and 22 percent report contact with the drug LSD. Other
narcotics are less prevalent, however few known illegal substances are not represented, in

any form, on campus police blotters.

Multivariate Analyses

Corresponding with the research questions outlined in Chapter 2, a number of
multivariate analyses were performed in an efi‘ort to gauge which factors were
significantly associated with levels of violent and prOperty crime, as well as with serious
criminality on campus overall. Because the independent variables utilized in this
dissertation may be reclassified under three subheadings, separate analyses were
performed for each of the three sub-groups, followed by a complete model in which all
subgroups of variables were regressed upon gross crime figures for the responding

campuses.

108

Before entering into a discussion of the modeling of independent variables, it is
necessary to define the nature of the outcome, or dependent, variables that will be used in
the following analyses. As most are aware, gross numbers of crimes occurring in a given
jurisdiction are highly dependent upon the size of that areas population. Therefore,
analysis of the factors contributing to crime or its suppression are virtually meaningless
without a standardization of these figures based on population size.

As discussed in the preceding chapter, there exist some disagreement as to which
indicator of overall population is most appropriate in Standardizing rates of campus
crime. Some researchers prefer using census-type figures such as the number of residents
on campus, while others argue that many different constituents, not simply residents of
university housing, converge in time and Space daily on campus, therefore all constituents
(namely all students and employees) should be the standardization base. In several
analyses to follow, the dependent variable of “crime rate” will be operationalized using
both denominators, thus alleviating the disagreement and allowing a comparison between
the two outcomes attributable to the same independent variables. In later analyses,
however, numbers of crimes remain unstandardized, and total number of constituents and
resident population are introduced as independent variables, thus alleviating the need to

standardize crime statistics by population size.

Factor Analysis and the Construction of Scales

The LEMAS questionnaire contains a number of variables which tap a similar

social construct. For example, the section of the questionnaire which asks schools to

109

indicate which selection criteria they use in hiring new public safety officers are all
predictors of the construct “selectivity in hiring”, and are thus highly correlated with one
another. To execute valid regression analyses and avert problems associated with
multicollinearity of independent variables, factor analysis was performed and scales were
constructed consisting of survey items tapping a similar theoretical construct. For two of
the three subgroupings of independent variables, items which were thought to logically be
indicative of the same prevailing idea, factor analyses to determine empirically the
structuring of concepts were conducted. Results of these factor analyses are presented

below in Tables 8 and 9.

Crime Prevention Strategies

A major part of the LEMAS questionnaire distributed to campus public safety
departments was the section outlining crime prevention strategies in place at the
institutions. Such information is of importance because the Campus Crime and Security
Act of 1990 mandates a description of the overall crime prevention strategy in place on
campus. Therefore the individual items fiom the questionnaire are all indicative of the
relative sophistication of crime prevention tactics at the institutions. A factor analysis was
performed on items thought to tap a similar dimension of crime prevention on a college
and university campus. As Table 8 indicates, the items included in the analysis loaded on
five separate factors, in a rather logical and consistent manner.

The four questionnaire items reflecting the use of communications equipment by

campus public safety departments all loaded highly on component 2 of the factor

110

analysis, therefore these items were combined into a composite scale (COMMX, alpha=
.49). As a rule of thumb, a factor loading of greater .50 defines an item as loading
sufficiently high enough on a factor for inclusion in a composite measure. All of the
items in the COMMX scale fit this criteria but one (Portable radios= .49) but because the
loading was borderline, and there was a logical basis for the item to remain in the
analysis, the decision was made to include the item in the scale. As a result, the overall
reliability of the scale as measured by Cronbach’s alpha is lowered somewhat by
incorporating the item.

A second composite scale was constructed using many of the remaining items.
This general crime prevention program scale (PREVENTX, alpha= .76) included all but
two of the items which loaded higher than .50 on factor number one. Those two items,
alcohol abuse education and drug abuse education also loaded fairly high on a separate
factor (number 3), and because they are indicative of a unique type of prevention
education (that of substance abuse), the decision was made to combine these two items
into a separate measure (SUBABUSX, alpha= .89). Of the remaining two items fiom the
factor analysis, whether or not the institution utilized a student security, or “mule”, patrol
loaded separately on factor 5. This item was included separately in subsequent analyses
because it is indicative of the volunteer-driven strategies used by institutions to prevent
crime, described in the theoretical fiamework. The item ‘Other prevention prograrn’ was
omitted from subsequent analyses due to the low number of institutions reporting other
programs (12%, see Table 5) and because the exact nature of these programs is not

discernible fi'om the LEMAS data set.

111

Table 8. - Factor Loadings for Crime Prevention Items from the LEMAS

Questionnaire

 

 

 

Components & Loading

Questionnaire Item 1 2 3 4 5
Portable radios .02 .49 .38 .20 -.40
Mobile vehicle radios .31 .67 .19 -.1 1 .17
Cellular phones .18 .53 -.30 -.07 .08
Base Station Radios .19 .64 .13 -.22 -.21
Victim assistance program ~.60 -.19 -.28 .05 -.33
General crime prevention program .54 .09 -.23 .15 -.05
Hate crimes prevention program .55 -.07 -.29 -.O3 -.17
Student security patrol .13 .29 -.32 .32 .69
Date rape prevention program .79 -.11 -.06 .16 -.05
Stranger rape prevention .78 -.06 -.06 .13 -. 12
Self-defense training .52 -.09 -.11 .13 -.04
Alcohol abuse education .71 -.21 .46 -.18 .25
Drug abuse education .71 -.20 .46 -.21 .26
Other prevention program -. 12 -.02 .38 .81 .01

 

 

112

Table 9. — Factor Loadings for New Officer Selection Techniques from the LEMAS
Questionnaire

 

 

 

Components & Loading

Questionnaire Item 1 2 3

Written test .54 -.21 -.09
Oral interview .05 .03 .82
Board interview .54 .06 -.47
Background investigation .41 .61 -.03
Criminal records check .44 .61 -.03
Physical agility test .65 -.27 .15
Psychological screening .74 -.13 .01

Medical exam .71 -.02 -.07
Polygraph exam .36 -. 19 .26
Drug screening .63 -. 17 .06

Other selection criteria .03 .53 .32

 

 

113

New Officer Selection Techniques

Within this subgrouping of independent variables, a number of items fi'om the
questionnaire are indicators of the different selection techniques used by institutions in
the hiring of new officers. Thus, a factor analysis was performed to determine which
items could be combined into scales for inclusion in later analyses. A majority of the
items included in the analysis loaded highly (above .50) on factor 1, and were
subsequently combined into a 6-item composite measure indicative of selection
techniques (SELECTX, alpha= .74). Of the remaining four items, two loaded highly on
factor 2. These items, “Background investigation’ and ‘Criminal records check’
(BCKGRNDX, alpha= .49) were combined into an item indicative of the institution’s
efforts policies of inquiring about the personal histories of applicants. The item
‘Polygraph examination’ failed to load significantly on any factor and thus was
eliminated, and the item ‘Other selection criteria’ was omitted for the same reasons given

above in rejecting ‘Other prevention program’.

Correlations Among Factors Influencing the Numbers of Crimes Occurring at Post-

Secon I 'tutions

A fairly common problem in linear regression analyses is that of multicollinearity
among the factors influencing the outcome. In other words, two or more variables are
essentially measuring the same theoretical construct. One method of alleviating this
problem is through the creation of composite variables, or scales, as was done in the

preceding section. Afier creating scales, correlation coefficients (Pearson’s r) are

114

computed for the resultant composite scales, as well as for the remaining variables to be
used in linear regression analyses. A table of correlation coefficients will indicate
variables that are strongly associated and thus indicative of the same theoretical construct.
Should such an instance occur, and the variables remained in the multiple regression
analyses, invalid results would be obtained as the inclusion of all variables which are very
highly correlated is equivalent to including a variable more than once. The correlation
coefficients for the regression analyses to follow are presented below in Table 10.
Clearly, some of the variables used in this dissertation are correlated to some
degree, and this is to be expected as, conceptually, all of the factors presented above are
thought to be associated with campus crime as specified within the theoretical perspective
of Situational crime prevention. However, none of the 23 independent variables are
correlated highly enough to be indicative of the same construct (see Morash and Haarr,
1995). Thus, given that only weak to moderate correlations exist between the factors
thought to influence campus crime, multicollinearity is not thought to be a confounded in

the ensuing multivariate analyses.

OLS Regession Analyses - Partial Models

Ordinary Least Squares regression was used to determine the influence of
demographic, crime prevention, and public safety department operations upon the number
of crimes occurring on the campuses surveyed by the Bureau of Justice Statistics.
Initially, analyses were performed for four subgroupings of factors thought to influence

campus criminality. The last analyses incorporate all of the factors into a full model

115

Table 10.- Correlations Among Factors Influencing the Number of Campus Crimes

 

 

 

 

 

Item 1 2 3 4 5 6 7 8 9 10

1.CJPROG .29" .12* .12" .06 .04 .04 -.01 .17** .17**

2. CONSTITU .68" .24" .59M .38" .44" -.12* .24" .38"

3. RESPOP -.01 .66M .54** .46" -.16** .22“ .32"

4. ENVIRONM .00 -.14** -.02 .03 -.03 .05

5. BUILDING .61” .45** -.04 .17" .24"

6. ACRES .59" -.07 .18” .19"

7. ROADS -.15** .17" .27"

8. EDUCATON -.13** -.ll*

9. CTRAIN .33"
ll 12 13 14 15 16 17 18 19 20

1.CJPROG .19" .07 .20** .04 .08 .06 .09“ .06 .17" -.05

2. CONSTITU .42" .14" .35" .16" .31" .33" .16" .08 .26" .03

3.RESPOP .23M .11“ .24" .20“ .22" .30" .19“ .18** .22M -.03

4. ENVIRONM .02 -.02 .04 -.05 .12" -.02 -.01 -.19** .03 .19"

5.BUILDING .19" .09* .23" .19“ .15" .26" .12" .06 .18" -.03

6.ACRES .23" .08 .21" .15“ .06 .22" .10* .14“ .14" -.07

7. ROADS .25" .06 .22" .07 .13" .22" .12* .15M .17" -.05

8. EDUCATON -.20** -.11* -.05 -.05 -.06 -.16** -.08* .00 -.05 .09*

9. CTRAIN .54" .24**.42** .12“ .15" .32" .11“ .16" .24" -.11*

21 22 23

l.CJPROG .18” .12" .07

2. CONSTTTU .30" .19M .16"

3. RESPOP .22“ .17" .27"

4. ENVIRONM -.01 -.03 -.11*

5. BUILDING .20" .15" .33"

6.ACRES .14“ .15" .29"

7.ROADS .17" .18" .18"

8. EDUCATON -.09* -.09* -.04

9. CTRAIN .50" .36" .10*

Table 10. - (cont’d)

116

 

 

 

 

ll 12 l3 14 15 16 17 18 19 20
10.FTRAIN .31** .13** .29** .16** .15** .27“ .14** .15** .20** .00
11.1NTERVWX .32** .42** .15** .21** .33" .14** .12** .27** -.03
12. BCKGRNDX .26“ .14" .10* .18** .05 .15** .l2** -.07
13. ARMED .12** .07 .30** .06 .09* .10* -.l7**
14.911SYS .08 .13* .O9* .10* .14** -.10*
15. BLUESYS .21** .08* .12 .21** -.01
16. COMMX .16" .10* .l9** -.l6**
17. MULES .01 .10* -.10*
18. SUBABUSX .50** -.01
19. PREVENTX .01

11 12 13 14 15 16 17 18 19
20. OUTSOURC
21. ARREST .51** .34" .67“ .11“ .10* .32” .10* .16** .15*
22. DRUGLAWS .32“ .29” .47" .14" .11** .21 ** .09* .14" .12M
23. DRUGTASK .19** .08 .19” .09* .03 .16" .09* .18“ .13**
21 22 23

20. OUTSOURC .15* .12" .13”
21. ARREST .55” .18"
22. DRUGLAWS .14”

23. DRUGTASK

 

 

117

regressed upon gross crime figures for the participating institutions.

As previously mentioned, index crime statistics were standardized by both the
number of total constituents and by the resident population of the campuses
(INDEXCON, INDEmES). Also, index crimes were analyzed in their totality, by
violent crimes only, and by property crimes only (V IOCON, VIORES; PROPCON,
PROPRES). Table 11 presents results for the dependent variable standardized by total
number of constituents, Table 12 presents results for the dependent variable standardized
by resident population.

Diagnostics were obtained for each regression analysis to follow, and the Durbin-
Watson Statistic is presented in the tables for each run. This statistic is useful for
evaluating the presence of autocorrelation (StatSofi, 1995:1675). A generally accepted
cutoff point for the Durbin-Watson statistic is 1.8; a figure below this indicates unstable
regression coefficients (StatSoft, 1995). In the following analyses, a handful of regression
analyses produce Durbin-Watson statistics lower than 1.8, however other diagnostics
such as variance inflation factors (VIF) as well as the correlation matrix of the
independent variables are not indicative of multicollinearity. The matrix indicates only
weak to moderate correlations among the independent variables and the VIP (not Shown)
are well below the generally accepted cutoff of 5.0 (range of VIP for subsequent analyses
= 1.0 - 2.8) (for a discussion of variance inflation factors, see Hair, Anderson, Tatharn,
and Black, 1995). Therefore, the decision was made to include the runs producing low
Durbin-Watson statistics, but the possibility of autocorrelation is present, thus the results

from these analyses must be interpreted with caution.

118

Campus Public Safegy Department Selection Techniques for New Officers

 

The first step in establishing, or implementing a renewed sense of,
professionalism is to hire quality individuals as new officers. The following tables present
the results of a multivariate regression analysis of departmental selection
techniques/criteria upon standardized crime figures for the campuses responding to the
LEMAS questionnaire administered by BJ S.

Table 11 indicates a significant negative relationship between the educational
requirements of new officers and total crime rates at the surveyed institutions, meaning
that the more highly educated the new officers, the lower the total index crime rate and
vice versa. This relationship holds true for property crimes as well, thus lending support
for the idea that more highly educated campus public safety ofiicers may have a positive
effect on lowering crime rates. Conversely, Table 11 also indicates that the more
selection criteria utilized by the schools to choose recruits out of the new officer applicant
pool, the higher the crime rate. That is, the more selective the institution in hiring public
safety officers, the more crime occurring on campus. It is very possible that, temporally,

high crime rates were present on campus before selectivity in hiring new officers

 

increased at these institutions, but such information cannot be gleaned from a cross-
sectional data set such as the one used for this dissertation. Furthermore, because this data
set is the first of its kind, it is not feasible to examine existing data to determine when
selectivity in hiring increased or what crime rates at colleges and universities were prior

to 1994.

119

Table ll.— OLS Regression Analysis of Campus Public Safety Department Selection
Techniques upon Crime Figures Standardized by Total Number of Constituents.

 

 

 

INDEXCON VIOCON PROPCON
beta t beta t beta t
EDUCATON -.106 -1.80** .049 .86 -.108 -1.86*
CTRAIN .035 .52 .030 .45 .035 .51
FTRAIN -.006 -.094 -.048 -.82 -.001 -.02
INTERVWX .141 2.00** .083 1.23 .143 2.04”
BCKGRNDX -.013 -.21 -.06 -1.00 -.006 -. 10
F 2.71 ** .78 2.86**
DW 1.90 1.96 1.91
R .21 .11 .21
Adj. R—squared .03 -.01 .03

 

***p<.01 **p<.05 *p<.10

 

 

120

None of the other selectivity-related factors were found to influence campus crime
rates standardized by number of constituents, and neither of the significant relationships
discovered for total index and property crimes were significant for violent campus crime,
indicating that other factors are more closely associated with violent campus criminality
than new officer selection criteria.

Interestingly, the significant relationships in Table 11 between educational
requirements for new officers and serious criminality found when crime rates were
standardized by number of campus constituents, were not found when crime figures were
standardized by the resident population of the campus. As Table 12 shows, the positive
relationship between the selectivity index and crime rates remains (indeed, appears
stronger for index crimes overall; beta = .220, sig.=.00) and is found for violent crime and
property crime, as well as all index crimes combined. In addition, a positive relationship
is indicated between the number of classroom training hours and violent crime rate,
perhaps reflecting the increased training new recruits must complete before assignment to
a campus with a relatively large amount of violent crime.

Another indicator of selectivity, background checks of new recruits, was found to
have a significant, negative relationship with crime rates standardized by resident
population. That is, those institutions who perform both a background investigation in
addition to a criminal records check of recruits tend to have lower violent crime rates
when compared with institutions which do not perform such checks. Again, it is possible

that by only selecting new recruits with clean backgrounds, the philosophy of

121

Table 12.- OLS Regression Analysis of Campus Public Safety Department Selection
Techniques upon Crime Figures Standardized by Resident Population.

 

 

 

INDEXRES VIORES PROPRES
beta t beta t beta I
EDUCATON .005 .08 .026 .41 .088 .09
CTRAIN .103 1.34 .132 1.81* .100 1.29
F TRAIN .025 .37 -.046 -.70 .029 .41
INTERVWX .220 2.71*** .149 1.95* .223 2.74***
BCKGRNDX -.067 -.99 -. 121 -1.91* -.064 -.94
F 4.00*** 2.90" 4.07***
DW 1.57 1.91 1.57
R .29 .23 .29
Adj. R-Squared .06 .03 .06

 

I“!!! P < _01

#*p<.05 *P<

.10

 

 

122

professionalism within campus public safety is being upheld, and thus is reflected in

overall crime rates.

Campus Public Safeg Department Structure and Function

Another subgrouping of factors thought to influence campus crime rates deals
with the structure and function of the departments themselves. An educated, well-trained
public safety agency will be most successful when operating within a stable, well-
organized departmental scheme. The four items included in this subgroup were chosen as
indicators of campus public safety departments that are highly professional and mirror
municipal and state law enforcement agencies in their performed duties. The item
‘Officer arrest powers’ is a proxy for sworn vs. non-swom personnel. That is, campus
public safety officers who have gone through a municipal or state training academy are
distinguished from non-sworn personnel or contract security. The items pertaining to
drug law enforcement are further indicators of a developed, professional department.

Tables 13 and 14 present the results of multivariate regression analyses of
variables pertaining to departmental structure and function upon standardized crime
figures. Again, the dependent variable is standardized in two forms, and serious
criminality is categorized into total index crimes, violent index crimes, and property
index crimes. Mixed results were obtained when examining these regression analyses for
the different standardizations of the dependent variable crime rate. When looking at
crimes by the number of constituents (Table 13), those campus public safety departments

which are responsible for drug enforcement and are part of multi-agency drug task forces,

123

Table 13.-- OLS Regression Analysis of Campus Public Safety Department
Structure and Function upon Crime Figures Standardized by Total Number of

Constituents.

 

 

 

INDEXCON VIOCON PROPCON

beta 1 beta 1 beta t
OUTSOURC .001 .02 .190 3.91 *** -.013 -.27
ARREST -.062 -1.02 .046 .79 -.065 -1.06
DRUGLAWS .208 3 .43* * * .082 1.44 .206 3 .39* * *
DRUGTASK .113 2.21** -.037 -.76 .119 2.33“
F 4.66*** 4.80*** 4.79***
DW 1.84 1.97 1.81
R .22 .21 .22
Adj. R-squared .04 .04 .05

 

***p<.01 **p<.05 *p<.10

 

 

124

Table l4.- OLS Regression Analysis of Campus Public Safety Department
Structure and Function upon Crime Figures Standardized by Resident Population.

 

 

 

INDEXRES VIORES PROPRES
beta t beta t beta t

OUTSOURC -.016 -.28 .145 2.66*** —.023 -.39

ARREST .151 2.05** .135 1.98** .153 2.07**

DRUGLAWS .036 .50 .016 .24 .034 .47

DRUGTASK -.096 -l.62 -.062 -1.14 - .094 -l.6O

F 2.57** 3.20** 2.60**

DW 1.41 1.86 1.41

R .19 .19 .19

Adj. R-squared .02 .03 .04

 

***p<.01 **p<.05 *p<.10

 

 

125

tend to patrol campuses with higher total index and property crime rates. The only
significant finding for rates of violent crime is that those campuses which contract out for
protective services tend to have greater rates of violent crime. such a finding would
certainly support the argument that a proactive public safety department, housed on
campus, could be effective in lessen or reducing the numbers of violent index crimes.
Table 14 presents somewhat different results. When crime rates are standardized
by resident population, the variable ‘Officer arrest powers’ is significant for all three
categorizations of the dependent variable, thus indicating that those campus safety
departments who utilize sworn law enforcement officers also tend to have higher crime
rates. Again, this finding could be the result of schools responding to increasing rates of
all index criminality by instituting sworn law enforcement officers as agents of public
safety on campus. The relationship between the two items ‘Drug law enforcement’ and
‘Participation in a multi-agency drug task force’ and crime rates, significant in Table 13,
does not hold true for the other standardization of the dependent variables. However, as
mentioned above, because of the low Durbin-Watson statistics in Table 14, this result is

interpreted with caution.

Crime Prevention Proms and Strategies

A major component of this dissertation is to determine the relationship between
crime prevention programming at larger colleges and universities in the United States and
crime rates at these institutions. Following the rationale of Situational crime prevention,

those institutions with a relatively well-deve10ped strategy for deterring potential

126

criminal acts on campus should have, proportionately, less crime than those institutions
without such proactive policies. Tables 15 and 16 present the results of the regression
analyses evaluating this assertion.

Most notable in Table 15 is the differing significance of several independent
variables across the different configurations of the dependent variable. For example, there
is a positive relationship between the utilization of communications technologies by
campus public safety departments (COMMX) and total index crime rates which is
significant at the p < .01 level. However, this relationship disappears for violent index
crime, then reappears for property index crime. A similar relationship exists between
blue-light emergency telephone systems (BLUESYS) and crime rates. These findings
suggest, again, that given the wording of the Campus Crime and Security Act of 1990,
and the potential of civil litigation against institutions seen as negligent in providing
student safety, that schools are responding to higher rates of crime by instating various
proactive crime prevention schemes, such as utilizing better communications.

Looking at violent index crime only in Table 15, the connection between campus
public safety officers carrying firearms and violent crime rates is also positive, meaning
that officers are more likely to be armed on campuses with higher rates of violent crime.
This also suggests a patterned response by institutions to implement an effectual crime
prevention, or crime fighting strategy, to help ensure constituent safety. Furthermore, and
contrary to what would be expected given the framework of the Situational perspective,
overall crime prevention strategies do not seem to have the effect of ameliorating

criminal activity on campus. Schools with more involved crime prevention programming

127

Table 15.— OLS Regression Analysis of Campus Crime Prevention Programs and
Strategies upon Crime Figures Standardized by Total Number of Constituents.

 

 

 

 

INDEXCON VIOCON PROPCON
beta 1 beta 1 beta t

ARMED -.058 -1.12 .105 2.05" -.065 -1.26
9118YS -.008 -.16 -.110 -2.20** .001 .02
BLUESYS .137 2.60** -.019 -.37 .143 2.72***
COMMX .172 3.19*** -.035 -.67 .179 3.34***
MULES .021 .41 -.053 -1.06 .025 .50
SUBABUSX -.101 -1.72* -.071 -1.23 -.100 -1.71*
PREVENTX .102 1.68* .077 1.31 .102 1.67*
F 432*“ 1.78“ 4.69***
DW 1.95 2.04 1.95
R .28 .17 .29
Adj. R-squared .06 .01 .06
***p<.01 **p<.05 *p<.10

 

 

128

Table l6.— OLS Regression Analysis of Campus Crime Prevention Programs and
Strategies upon Crime Figures Standardized by Resident Population.

 

 

 

INDEXRES VIORES PROPRES
beta t beta 1 beta t

ARMED .159 2.65*** .198 3.51*** .156 2.61**
911SYS -.082 -1.42 -.094 -1.70* -.078 -1.35
BLUESYS .069 1.14 -.005 -.082 .069 1.15
COMMX .045 .70 -.055 -.94 .050 .79
MULES .067 1.15 -.006 -.11 .068 1.17
SUBABUSX -.228 -3.36*** -.112 -1.74* -.230 -3.40***
PREVENTX .079 1.11 .077 1.16 .081 1.14
F 4.16*** 2.59" 4.20***
DW 1.54 1.94 1.55
R .31 .23 .31
Adj. R-squared .07 .03 .07

 

***p<.01 **p<.05 *p<.10

 

 

129

(PREVENTX) tend to have correspondingly high rates of total index and property index
crime.

One finding fi'om Table 15 which is perhaps indicative of the Situational
perspectives’ effectiveness is the significant negative relationship between substance
abuse programming (SUBABUSX) and rates of total index and property index crime.
These findings suggest that schools which embark on an earnest campaign to curb drug
and alcohol abuse experience less index and less property crime than those schools not
offering substance abuse prevention. This idea is consistent with the ideas of campus
crime scholars who posit a direct link between substance (specifically alcohol) abuse and
acts of criminality on campus (e.g., Palmer, 1993:57). Because many schools wrestle
with balancing students’ want of a social atmosphere on campus with the problem of
binge drinking, if such a finding were replicated in other empirical studies, there would
perhaps be an in-depth call for re-examination of an institution’s present approach to
substance abuse education.

The results presented in Table 16 further support the above idea that substance
abuse prevention is associated with lower rates of criminality. In addition, the
relationship holds true for violent index crime as well, when crime figures are
standardized by the resident population of the campus. Also of note is the significant
positive relationship between armed campus public safety officers and crime rates. For all
three categorizations of the dependent variable, those schools with armed officers
patrolling their campus are also likely to have higher crime rates. The reasoning behind

this finding is comparable to that given for the Significant positive result in Table 15

130

between armed officers and high rates of violent index crime; schools with a real or
perceived crime problem may issue firearms to officers as a preventative or safety issue.
Contradicting the logic of the Situational perspective, there exists no relationship
whatsoever between crime prevention programming and index crime when the latter is
standardized by resident population. Given that the nature of these preventative measures
are educational in nature, it stands to reason that, as for all crime prevention activities in
general, a sufficient time lag must exist between the implementation of the program and
when results become manifest empirically. Thus, it would certainly be premature to say
that crime prevention and suppression efforts, as outlined by the Situational perspective
and operationalized within this data set, do not reduce the amount of serious criminality

on college and university campuses.

Characteristics of the Institutions

To follow the reasoning of the environmental criminology element of the
Situational perspective, it is necessary to examine the impact of campus demographics on
campus crime. AS most realize, the gross numbers of crimes occurring in any jurisdiction
are highly associated with the number of persons in that jurisdiction at any particular
point in time. Thus, for campus criminality, the gross number of crimes for any school is
at least partially a function of the student enrollment, number of employees, and number
of persons residing on campus. For the following partial analysis, gross crime figures for
the 680 institutions included in the BJ S study are lefi unstandardized, and the number of

campus constituents and resident population of the campus are included in the partial

131

model as independent variables. The remaining demographic variables are then also
included to ascertain whether they are significantly associated with campus crime when
size of the institution is held constant.

Because gross crime figures are left in their raw form, only three (as opposed to
six for the other three partial models) regression analyses were performed on total index
crimes, violent index crimes, and property index crimes. Table 17 illustrates the findings
from the regression analyses and, as expected, significant positive associations with crime
figures are found between both the number of constituents and size of the resident
population on campus. That is, the larger the school in terms of enrollment, number of
employees, and on-campus residents, the more crimes occurring on that campus.

More interesting results are found when significant results for the other
demographic variables are examined. In keeping with the rationale of Situational crime
prevention, the item ‘Miles of campus roadways’ is positively associated with both total
index crimes and property index crimes. Without doubt, a large network of campus
throughways provides numerous situations for crimes of opportunity to occur, by both
making targets more attractive and unguarded. Multiple roadways also provide easy and
quick escape routes for those who commit criminal activities on camms. The variable
‘Nature of criminal justice programming’ was found to be negatively associated with
criminality on campus, indicating that those institutions with more highly developed
criminal justice curriculums had fewer serious crimes committed on campus. The initial
logic behind including this item in the analyses was that an active criminal justice

department could supply consultants and student labor/volunteers for campus public

132

safety departments. Further, the faculty fi'om an active criminal justice department could
help the public safety department develop a crime prevention strategy befitting the
campus and its’ particular crime problems. This is but one possible explanation for this
significant finding, and because the inclusion of this variable in the partial model was
done for exploratory purposes, it is not appropriate to make conclusions without any
theoretical or logical backing.

From the results of the analysis above, it may be apropos to look at which
findings are n_ot significant, as well as those that are. Table 17 shows that two items
indicative of the tenets of environmental criminology are not Significantly related to
crime figures on campus. However, it would be erroneous to conclude that the number of
buildings and land area of colleges and universities could not be linked with serious
criminality. The idea of constructing facilities with crime prevention in mind goes back at
least as far as Newman’s (1972) book on “defensible space.” Post-secondary institutions
in the United States are constantly undertaking new construction projects to upgrade and
make additions to existing facilities. It seems plausible that crime prevention through
environmental design (CPTED) is at least a small part of the planning stages of any new
construction project.

Therefore, more attention being paid to prevention through design, coupled with
crime prevention programming for students and awareness of potentially dangerous
situations by campus public safety, may explain the lack of significant results between
these environmental items and serious criminality. Thus, crime prevention strategies may

already be paying dividends for those institutions who have heeded the advice of

133

Table 17.- OLS Regression Analyses of Campus Demographic Characteristics
upon Gross Campus Crime Figures.

 

 

 

 

TINDEX VINDEX PINDEX

beta t beta t beta t
CJPROG -.060 -2.01** .076 1.35 -.064 -2.13**
CONSTTTU .687 15.13*** .306 3.59*** .696 15.26***
RESPOP .273 6.64*** .433 5.60*** .261 6.31***
ENVIRONM .039 1.28 .036 .64 .038 1.26
BUILDING .003 .079 -.066 -.87 .006 .15
ACRES -.022 -.61 -.028 -.42 -.023 -.63
ROADS .071 2.16M .034 .55 .070 2.12**
F 161.69*** 2345*" 160.11***
DW 2.33 1.98 2.34
R .92 .92
Adj. R-squared .85 .84
***p<.01 **p<.05 *p<.10

 

 

134

environmental criminologists. Again, because of the cross-sectional nature of the data set,
such effects of time are not discernible, and conclusions are Speculative without
replication.

Finally, contrary to popular belief, no association was found between serious
criminality and nature of the campus environment. That is, urban campuses to not appear
to be more ‘crime-plagued’ than their suburban and rural counterparts. It must be
repeated that the LEMAS data used in the present study measures only crimes occurring
on or within campus property. Thus, it is possible for a campus to be a relatively safe
environment for study, even if the smrounding community has a high crime rate. It would
seem that colleges and universities could benefit from a public relations standpoint if this
finding were replicated, however it does little to diminish potential students’ fear of
victimization off campus. However, the notion that a community’s crime problem
necessarily permeates the college or university campus does not seem to hold true given
Bromley’s (1995) results, nor does it appear fi'om the above finding that urban campuses

are necessarily more dangerous than those in less populous environs.

OLS Re ession Anal ses - F Model
In the section above, Specific subgroupings of independent variables were
examined as partial models of a larger test of the Situational perspective’s predictive
power concerning crime figures and rates at larger colleges and universities in the United
States. The final analyses for this dissertation are full model regressions incorporating all

four of the above partial models upon gross crime figures. Such a procedure will

135

highlight spurious, or false, significant relationships within the partial models as well as
indicate those partial model relationships which hold true, and are thus more robust and
valid. Full model results are presented below in Table 18. Not unexpectedly, several of
the significant associations from the partial models are not significant in the full model,
however some relationships remain, and thus are the focus of the discussion to follow. As
anticipated, the association between size of the institutions, in terms of number of
constants and number of on-campus residents, remains intactnthe larger the school the
more serious criminality. However, even with size of the institution taken into account,
other factors are also found to be significantly associated with serious campus
criminality. Most surprising perhaps is the fairly strong relationship between the nature of
criminal justice programming and serious crime. The association from the partial model
holds true above, as schools with undergraduate and graduate programs in criminal
justice appear to have less overall on-campus criminality. Again, because of the lack of
previous empirical work surrounding this issue, conclusions drawn from this result
should be seen only as speculative in nature, however this result does support the idea
that having a group of criminal justice scholars and students of the discipline in residence
may have an ameliorating influence on serious campus criminality.

Another Significant finding concerns the selection and recruiting of new campus
public safety officers. Institutions that utilize a number of criteria to aid in choosing
recruits from an applicant pool are more likely to have more total index and property
index crimes. This result may also reflect the notion that schools with rising crime rates

and/or more qualitatively serious crimes occurring in the recent past may implement new,

136

Table 18.- OLS Regression Analyses of the Full Model upon Gross Crime Figures.

 

 

 

 

TINDEX VINDEX PINDEX
beta t beta t beta t
Charact. of Institutions
CJPROG -.101 -.246** .062 .79 -.104 -2.56**
CONSTTTU .674 11.03*** .277 2.36“ .681 11.17***
RESPOP .265 4.05*** .369 2.97*** .251 3.86***
ENVIRONM .052 1.20 .035 .42 .050 1.16
BUILDING -.023 -.39 -.l44 -1.27 -.016 -.27
ACRES -.060 -1.28 -.028 -.32 -.061 -1.30
ROADS .059 1.35 .010 .11 .059 1.34
Select. and Train. Criteria
EDUCATON -.018 -.48 .029 .40 -.021 -.56
CTRAIN .002 .04 .106 1.11 .000 -.01
FTRAIN -.017 -.38 .017 .21 -.016 -.36
INTERVWX .091 1.79* .030 .31 .093 1.82*
BCKGRNDX .012 .31 -.077 -1.07 .014 .36
Structure and Functions
OUTSOURC .040 1.09 .125 1.74* .035 .95
ARREST -.090 -1.60 -.013 -.12 -.090 -l.60
DRUGLAWS .027 .64 -.032 -.39 .024 .54
DRUGTASK .054 1 .42 .032 .44 .053 1 .41

137

Table 18.- (cont’d.)

 

Crime Prevention

ARMED .035 .72 .086 .91 .033 .67
911$YS -.012 -.32 -.061 -.82 -.013 -.33
BLUESYS -.008 -.20 -.081 -1.07 -.004 -.10
COMMX .056 1.32 .031 .38 .056 1.30
MULES .031 .84 -.042 -.59 .034 .93
SUBABUSX -.007 -.15 -.102 -1.22 -.005 -.1 1
PREVENTX -.022 -.49 .144 1.69* -.028 -.63
F 30.73*** 3.63*** 30.95***
DW 2.38 2.16 2.38
R .92 .61 .92 .
Adj. R-squared .81 .27 .8 l

 

***p<.01 **p<.05 *p<.10

 

 

138

innovative strategies to combat this problem. Professionalizing the campus public safety
department through rigorous hiring standards is another method many institutions have
adopted.

Proponents of professionalism will further be supported by the result which
indicates that high numbers of violent index crimes are significantly associated with
campuses which outsource for their protective services. In the full model presented
above, the finding that schools which contract out to law enforcement and private security
agencies are more likely to see higher rates of violent crime bolsters the position that only
a professional, in-house public safety agency is familiar with the problems specific to the
campus being served.

In order for the public safety department to be truly proactive and preventative in
nature, they must be a part of the campus network in much the same way neighborhood
community police officers are in tune with the goings—on in their beat. Although it is
perhaps a generalization, outside contractors most likely do not have the personal stake in
the well-being of the campus in a way that an in-house department would have, therefore,
while both types of services could respond to calls for service, only the in-house
department has the capacity to be proactive. Meaning that crime prevention and the
implementation of strategies aimed at lessening the probability of victimization require
significant expenditures of time and resources that contract providers may be unwilling or

unable to supply.

CHAPTER V
DISCUSSION AND CONCLUSIONS
Summg of the Findings

Critics of Situational crime prevention are quick to point out the theoretical as
well as empirical shortcomings of the perspective, and the results of this dissertation
would seem to add fuel to the fire. The formulation of the perspective is denounced for
being a reactionary, ex post facto framework regarding criminal behavior. Naysayers
point out what they believe to be a gross indiflerence to the conditions and factors within
the social framework that breed individuals who, for whatever reason, are inclined to
commit acts of serious criminality. Further, it is said that if society continues to focus
upon protecting the law-abiding from the law-breakers, those who do not perpetrate
criminal acts are forced to cloister themselves behind protective barriers of whatever sort
(either physical or psychological), worsening the social structmal and environmental
factors that are thought to influence serious criminality. In a sense, society would retreat
behinds its crime prevention barriers leaving the “barbarians at the gate.”

While, as was argued in Chapter 2, it is certainly true that the Situational
perspective does little to explain or influence the genesis of criminality among groups or
individuals, it is as necessary to have reactionary theoretical perspectives as it is

necessary to have a criminal justice system designed to react to instances of wrong-doing.

139

140

This dissertation has described the needs and responses of one particular social
institution, the college and university, in protecting its specific constituency from
criminal acts committed by both constituents themselves and outsiders on campus
property. To be sure, persons affiliated with post-secondary institutions would most
certainly relish a revolutionary idea or ideas that would make prolific steps forward in
eradicating unlawful behavior in American society, however, these institutions certainly
do not have the luxury of waiting for such a remedy to be found. Furthermore, the
Student Right to Know and Campus Security Act of 1990 has forced the hand of
administrators at post-secondary institutions to look at the crime problems at their schools
with prevention as the best alternative. In doing so, the Situational perspective is perhaps
the most useful tool available to college and universities in terms of criminological
theory.

The results presented in the previous chapter were not what was to be expected
given the theoretical fiamework and the prior empirical work concerning campus
criminality. Both the partial model and firll model analyses indicate that the most
predictive factor concerning crime rates is the physical size of the institution, and
systemic responses, derived fiom the Situational perspective and other crime prevention
literature, to these crimes seem to have little tangible effect. This is especially true for
serious violent campus criminality, where even the explanatory value of institution size is
somewhat limited, thus indicating that other factors not included in the above models are
influencing these acts.

However, it would be erroneous, given the exploratory nature of the data, to make

141

overarching generalizations about campus crime or the success of responses to campus
criminality at this point. The results from Chapter 4 may be more a function of gathering
base line data (or, a starting point for longitudinal analyses), than any valid empirical
evidence against the Situational perspective. To reiterate, given the changes campus
safety departments have undergone due to the Campus Crime and Security Act of 1990, it
is wholly possible, and perhaps probable, that the crime prevention programs presented in
the above analyses are responses to crime rates, and have thus been implemented after
rates of criminality had already peaked. Therefore, in order to ascertain the effectiveness
of these programs at lowering crime rates, a sufficient amount of time must elapse
between the point at which the programs were implemented, and a point at which the
effectiveness of theses strategies can be measured. Unfortunately, the time lag for the data
set used in this dissertation is probably not sufficient enough to see an appreciable effect.

Because the administration of the LEMAS questionnaire to campus public safety
departments in 1994 represents the first systematic, nationwide empirical assessment of
campus crime figures and institutional responses to criminality, there is of course no way
of gauging the extent of criminality prior to 1994. Because of this, it is possible that
crime prevention strategies and campus public safety policies aimed at lessening the
probability of constituent victimization may have indeed reduced crime, both
qualitatively and quantitatively, thus supporting the tenets of the Situational perspective.
Of course, the only way to gauge this effect is through detailed replication and the use of
alternative methodologies to examine the phenomenon of on-campus crime.

There does exist a small number of empirical studies that have focused on

142

criminality at individual institutions (e. g. Sloan, Fisher, and Wilkins, 1993a), however
without any comparisons and with a shift in reporting practices and preventative
responses due to the Campus Crime and Security Act of 1990, the face of the problem is
dramatically different presently, five to six years after implementation of the legislation.
Thus, whatever results are gleaned from this initial data set serve as a comparison for
future replications, and are not to be construed as a conclusive discourse on criminality at

American post-secondary institutions.

Limitations of the Study

The problem that every exploratory study encounters concerning a lack of existing
empirical research to use as a comparison has already been addressed, and is perhaps the
biggest impediment to making generalizations from the results obtained in this
dissertation. Also, as discussed in Chapter 3, there are a host of possible dilemmas
associated with the use of existing data, most of which were avoided in this dissertation.
There are some other limitations to the study which deserve attention at this point.
Although most are common, unavoidable limitations inherently associated with the
choice of research methodology, certain precautions and modifications can be made up
front, prior to actually undertaking the measurement phase of the project. Below is a
discussion of some of these problems, along with recommendations for alleviating these

entanglements in future replications of this study.

143

Utilization of the Survey Research Method

A rather complicated impediment to overcome in any survey research is the
inherent problem of artificiality, especially that survey research which produces large
data sets with a multitude of variables which are to be analyzed quantitatively. The
variables selected for incorporation into the present study are indicators which serve as a
“pencil and paper” test, and lose a good deal of meaning when abstracted from their real
life context. As any undergraduate student who has completed a research methods course
would know, a check-marked answer on a questionnaire says nothing about the depth of
meaning inherent in that answer; in plain terms, a response becomes black and white,
when, in reality, campus crime and violence is an extremely complex issue which is

contextual in nature.

Specific to this dissertation is the problem of gauging implementation of crime
prevention programs on college and university campuses. As any scholar investigating
law enforcement practices can attest, there are frequently glaring differences between
policies set forth on paper and how those initiatives subsequently materialize as a part of
everyday procedure. No matter the quality of any data set, an analysis of implementation
is Simply not possible without some further investigation, optimally on sight, fiom a
qualitative approach. Field observation of this sort would invariably provide insight as to
the validity (or, more likely invalidity) of results generated quantitatively. Most policy
analysts would say that the most arduous task of evaluation research is not
methodological in nature, rather it is the dilemma of translating plans which hold a great

deal of potential into successful operations. Try as one might, this dilemma cannot be

144

evaluated from afar with quantitative data, thus a future analysis of this issue utilizing a
different methodology is indeed warranted in order to ascertain the validity of generated

results.

Conceptualization and Operationalization

Tied to the problems associated with the survey method is the disagreement
conceming how to best define concepts and operationalize variables as indicators of the
concepts within a questionnaire. In larger data sets, many social scientists prefer to keep
variables and their responses as simple as possible in order to better facilitate later coding
of the responses into a computerized data set. By restricting the range of possible
responses in this manner, one risks losing a good deal of content validity in that the
check-marked response may not be fully indicative of the respondent’s desired answer to
a given item. Therefore, results gleaned fiom such restrictive questionnaire items may not
be truthful representations of the concept as it exists in social reality.

Such problems with large data sets are difiicult to overcome, and are perhaps best
resolved through smaller scale replications of the original study with a smaller subset of
the original sample. By lowering the sample Size (keeping constant the use of probability
sampling techniques), the social scientist is able to utilize survey items which focus on
the depth of understanding into a particular set of concepts, thereby assuring a
proportionate increase in validity by allowing more detailed replies to questionnaire items
by respondents. Of come, more detailed responses are often difficult to code into

quantitative data for computerized analyses, however the benefit of an increase in

145

information from respondents (in the present case, campus law enforcement
administrators) could provide insight which would lead to modifications of future survey
instruments such that they would be more indicative of the research questions under
investigation.

As the results from this dissertation indicate, not all of the factors most closely
associated with campus crime are included in the LEMAS questionnaire, thus a
modification of that instrument is perhaps in order so that the issue of campus criminality
and institutional response may be better gauged. As previously discussed, the LEMAS
questionnaire is a tool used to systematically evaluate the characterization, structure, and
function of municipal police departments and their responses to crime and deviance
within their specific jurisdictions. Because the context of crime at post-secondary
institutions and constituent make-up on campus differs qualitatively from the municipal
department, it is perhaps unreasonable to expect truly valid results to be obtained from a
survey instrument not entirely fitted for the context under examination. Thus, it may be
appropriate to modify the LEMAS questionnaire specific to campus crime and public

safety departments before it is used in any replication of the 1994 results.

Limitatipps of the Data Analysis

For the quantitative data analysis in this dissertation, the most common regression
technique (ordinary least squares) was chosen, primarily because of the exploratory
nature of the study and because working hypotheses/research questions were derived

from a Specific theoretical fiamework and not from the results of prior empirical work. As

146

such, factors thought to influence campus criminality are not weighted in terms of their
explanatory power relative to crime figures. Thus, all independent variables are
introduced into the full and partial models simultaneously, without regard for which
factors may be more predictive. Entering variables into the models starting with the
independent variable thought to have the most explanatory power is indicative of
stepwise regression, but such a procedure may produce spurious results if independent
variables are entered improperly (see Hair et al., 1995). In addition, some methodologists
argue that stepwise procedures produce only marginally better results than ordinary-least
squares models, for the latter, by definition, identify the factors with the most explanatory
strength.

Another potential problem regarding the quantitative analysis has to do with the
reliability and validity of the variables used in the models. A discussion regarding the
validity of data from survey research has already been put forth as has the potential
problems resultant from regression analyses with low Durbin-Watson statistics, however
a more pressing concern deals with the reliability of the composite scales created with
items tapping a similar construct. As a rule of thumb, a composite variable should have a
reliability as measured by Cronbach’s alpha of at least .70.

In the above analyses, two composite variables are used that have reliability’s less
than .70. The decision was made to include these composite items with low reliability’s
for several reasons. First, the factor analyses which identified the items tapping a similar
dimension were consistent with the conceptual logic that placed these items together in

sections on the LEMAS questionnaire. That is, the statistics generated by the factor

147

analysis confirmed the theoretical underpinnings of the social scientists who constructed
the survey instrument. Since the individual items were indicative of a larger abstraction,
both conceptually and statistically, they were included in the analyses despite their low
overall reliability. Secondly, Cronbach’s alpha often gives a more conservative (or lower)
estimation of reliability when few items are included in the composite scale. Because the
composite variables with low alphas in this case have two and four items, it may be that
the estimate of reliability is correspondingly low. Third, if the individual items were not
merged into composite variables, there would still be sound theoretical reason to include
them in the regression analyses. However, those items would also correlate highly (in
most cases above .70), thus multicollinearity could become a problem unless some items
were not included in the analyses. For these reasons, the composite variables are included
in the analyses, however any significant findings concerning the two scales with lower
alphas should be viewed with caution.
Policy Implications

It is not sound public administration to base policy decisions on the basis of one
exploratory study that has yet to be replicated and supported, however the findings from
this dissertation suggest that certain actions could presently be taken by post-secondary
institutions to help lessen the extent of serious criminality on campus property. The
provisions derived from Situational crime prevention remain sound despite a lack of
empirical support provided by this dissertation. The possible reasons for non-significant
findings are presented above, and there is no empirical evidence contradicting the

effectiveness of the programs and strategies, thus it is very possible that a number of such

148

prevention policies may prove successful on some campuses. Thus, one recommendation
put forth would be for institutions of higher education to continue and expand their crime
prevention initiatives for constituent protection and future empirical analysis.

Secondly, there seems to be evidence supporting more stringent hiring practices of
new campus public safety officers. This attention paid to professionalism has benefited
municipal police departments in terms of constituent service and has aided in the
implementation of community- or problem-oriented policing, and would likely do the
same for larger colleges and universities. A more professional public safety department
would conceivably employ officers who are knowledgeable of crime prevention
strategies, of dealing with the victims of serious crime, and able to solve non-criminal
disputes before such conflicts become police matters. These officers would also be more
inclined to be abreast of and support new strategies for dealing with or preventing
criminality and be able to customize these tactics specifically for their campus and
constituency. Indeed, the positive objectives of the Situational perspective are dependent
upon a public safety department with the willingness and capability to put these ideals
into practice, and subsequently evaluate the results respective to their jurisdiction.

Third, there does seem to be some, albeit tenuous, empirical evidence against the
outsourcing of campus protective services, especially concerning levels of violence on
campus. Following the logic of the Situational perspective, the implementation of its’
ideas rests solely with the commitment the institution and public safety department has to
preventing criminal activity on campus property. It stands to reason that schools which

contract out their protective services to outside agencies or corporations will not receive

149

the same commitment to prevention as do institutions with public safety departments in-
house. The problem-oriented model of police work posits that a thorough knowledge of
specific problem areas and conflicts is necessary so that preventative policies can be most
expeditiously dispensed. Such knowledge is accumulated over time, and by officers with
a stake in the well-being of the jurisdiction being served. Because most outside suppliers
of protective services do not view the campus as their sole province of responsibility (this
is certainly the case for contracts with law enforcement agencies), in-depth involvement
with and knowledge of the potential problem areas on campus is more than likely lesser
than their in-house counterparts. This discussion is not meant to imply that outside
agencies necessarily provide inadequate protective services, rather is highlighting the
depth of involvement necessary for Situational tenants to be successful. Certainly it is
possible for campuses which outsource to have proactive, involved protective services,
but the empirical evidence presented in this dissertation indicates that, for the prevention
of serious violent criminality, in-house public safety is more effective.

Fourth, because the possibility of crime displacement is always possible, there
should be cooperative efforts between campus and municipal law enforcement agencies
focusing on crime prevention, if such strategies presently do not exist. It is very common
for students to live off-campus in rental housing throughout communities adjacent to the
institution. Frequently, these temporary student residents have adversarial relationships
with permanent residents and authorities alike in these communities. Cooperation in
crime prevention, optimally utilizing community policing strategies, could have the

added benefit of improving “town-gown” relations by bringing students and permanent

150

residents together for the common cause of preventing criminal behavior.

Fifth, on campus, administrators and campus public safety directors must assess
their unique crime prevention needs. Not every campus needs domestic violence
prevention, for example, and some campuses need to focus more on violent crime than
others. A common mistake in the administration of criminal justice policy is to
implement programs or strategies simply because they have been found somewhat
successful in other locales. There are numerous instances of a policy being labeled a
“failure” because it has not produced positive results found elsewhere. Most often, these
poor outcomes are the result of an inappropriate implementation and not the failure of the
strategy itself. A college or university can easily avoid this frustration and waste of
resources by performing its own needs assessment or by having a professional security
survey performed by a reputable security contractor.

Sixth, in keeping with Clarke’s (1992) vision of the Situational perspective, the
results of this and future analyses like it should be disseminated in a forum where all
interested parties would have easy access. It may be inferred that Clarke (1992) believed
that the academic press was a limiting option, and that a forum such as a trade publication
or internet website would be better suited for publication of such results. Presently, there
are several websites devoted specifically to campus public safety and law enforcement
which may perhaps be interested in posting the results online, or a personal website
would also be sufficient.

Aside from these specific policy recommendations, it is important for post-

secondary institutions to not lose sight of the fact that even if incidents of criminality are

151

increasing and becoming qualitatively more serious, that the college campus remains a
rather protected environment if specific precautions are followed to insure individual
safety. As is evidenced from the results presented above, most serious criminality
occurring at post-secondary institutions is a violation of property, and not interpersonal
violence. Indeed, results from the analyses of total index crimes and property index
crimes are very Similar because a majority of serious (Part I) offending at colleges and
universities is property-oriented. Thus, while the prevention and control of violent crime
is paramount for many large institutions, because of constituent fear and the potential for
civil liability, the empirical data supports the assertion that property offenses are the most
common problem on-campus, and thus should not be forgotten in efforts to prevent
violence. While interpersonal violent offenses will most likely always be the main
concern of administrators and constituents alike, policies aimed at stemming the rise of
serious property offenses could serve a dual purpose. First, such strategies would attack
what is presently the most pressing crime problem, quantitatively, and second, a focus
upon preventing serious property crimes could help stymie the climate of lawlessness that
could potentially breed violence. From the rational choice perspective, a campus or
jurisdiction seen as unyielding when it comes to property crime does not serve as an
attractive environment for an offender to commit violent crimes. Of course, it is quite a
stretch to say that violent offenders operate fiom a rational choice perspective, but it not a
stretch to say that a crackdown on property offenses and even nuisance offenses could not
help in lessen the amount of violent crime as well (e. g. Wilson and Kelling, 1982; Wilson

and Herrnstein, 1985).

152

Recommendations for Future Research

Several recommendations for future empirical studies into the t0pic of campus
crime prevention have already been alluded to in former sections of this dissertation,
however it is appropriate to restate some of these recommendations more fully. First and
most obviously, because the 1994 data collected by BJ S and used here represents the first
systemic analysis of criminality and prevention programs at a nationwide sample of larger
colleges and universities after the implementation of the Student Right-to-Know and
Campus Security Act of 1990, the next logical step is a full replication of the study at a
later date to compare with these results. The Bureau of Justice statistics had planned a
replication using the LEMAS questionnaire for 1997 crime figures, however, three years
may not be a sufficient enough time lag for results of prevention strategies to become
apparent. Most likely, a five or ten year interval would be fitting, and the appropriate
modifications to the LEMAS instrument would perhaps improve the overall validity of
the project. Such modifications may include more detailed response sets for items (above
simple dichotomous ‘yes/no’ matrices), as well as information concerning the overall
crime prevention plan at the institution and how long the strategy has been in place.

The extent of implementation of existing prevention systems could be measured
using a traditional security survey (Richards, 1996), however implementation would
probably best be evaluated by on-campus observation of the campus public safety
department. It cannot be stressed enough that implementation is an oft-ignored missing
link in crime prevention that is overlooked time and again, even by trained social

scientists. Written policies concerning innovative policies for preventing on-campus

153

criminality are not necessarily indicative of a successful strategy. A thorough knowledge
of specific problems and issues at the institution as well as a commitment by the campus
administrators, director of public safety, and officers within the department are necessary
for Situational crime prevention to even have a chance at success. Therefore, as much
detailed knowledge as possible is required to evaluate the implementation of crime
prevention strategies, and this is best done through on-site observation over an extended
period of time.

Conceptualizing field observation, if permission were granted by the institutions
themselves, would not be excessively difficult, given that a subsample of the original
participating institutions could be developed since each institution is identified in the data
set. The subsample could be drawn either by selecting individual institutions at random,
or perhaps by selecting a particular state at random and then performing observations at
all schools in that particular state. The logistics and feasibility of such a study in terms of
personnel and resources would be the most problematic element. A great deal of time and
money could be required for each sight visit, and thus such a project could probably not
be undertaken without substantial outside frmding. On a positive note, this type of
research would probably be looked favorably upon by a number of funding institutions
such as the National Institute of Justice, the Bureau of Justice Statistics, the Police
Executive Research Forum, or even the institutions themselves, Since they are have the
greatest stake in preventing campus crime.

Future research also must elaborate upon those factors most associated with

serious interpersonal violent crime on campus. As the results from this dissertation

154

indicate, those factors which influence violent crime on campus were not contained in the
LEMAS questionnaire, thus an alternate approach to conceptualizing campus violence is
needed. Much has been written in terms of the factors contributing to specific instances of
violence, however campus public safety responses to violence overall must be better
operationalized and evaluated so that campuses remain safe for study well into the

twenty-first century.

APPENDIX A

Definitions and Coding of Variables Used in OLS Regression Analyses

155

APPENDIX A

Definitions and Coding of Variables Used in OLS Regression Analyses
Demndent Variables
TINDEX = Total # of Index Crimes
VINDEX = Total # of Violent Index Crimes
PINDEX = Total # of Pr0perty Index Crimes

INDEXCON = Total # of Index Crimes / Total Number of Constituents (Enrolled
Students and Employees) Affiliated with the Institution

VIOCON = Total # of Violent Index Crimes / Total Number of Constituents (Enrolled
Students and Employees) Affiliated with the Institution

PROPCON = Total # of Property Index Crimes / Total Number of Constituents (Enrolled
Students and Employees) Affiliated with the Institution

INDEXRES = Total # of Index Crimes / Total # of Individuals Residing in Campus-
Owned Housing

VIORES = Total # of Violent Index Crimes / Total # of Individuals Residing in Campus-
Owned Housing

PROPRES = Total # of Property Index Crimes / Total # of Individuals Residing in
Campus- Owned Housing

Indemndent Variables
Characteristics of the Institutions
CONSTTTU = # of Constituents Affiliated with the Institution

RESPOP = # of Residents Living in On-Campus Housing

156

BUILDING = # of Buildings on Campus
ACRES = Land Area of the Institution in Acres
ROADS = # of Miles of Campus Roadways
ENVIRONM = Nature of the Campus Environment/Location
1= Rural
2= Suburban
3= Urban
CJPROG = Nature of Criminal Justice Programming at the Institution
1= None
2= BS/BA in Criminal Justice Only
3= MA/MS in Other Discipline w/CJ concentration
4= MA/MS in Criminal Justice
5= PhD in Other Discipline w/CJ concentration
6= PhD in Criminal Justice
Crime Prevention Proggams and Strategies
ARMED = Campus Public Safety Officers Armed While on Duty

1= No
2= Yes

91 ISYS = 911 Emergency Telephone System

1= No
2= Yes

BLUESYS = Emergency Blue Light Telephone System

1= No
2= Yes

MULES = Student Security Patrol

1= No
2= Yes

COMMX = Composite Scale Indicating Utilization of Communications Technology

157

SUBABUSX = Composite Scale Indicating Scope of Substance Abuse Education
Programming

PREVENTX = Composite Scale Indicating Scope of General Crime Prevention
Programming
Campus Public Safeg Department Selection and Training Criteria
EDUCATON = Educational Requirement for New Officers
1= No Educational Requirement
2= High School Diploma or Equivalent
3= Some College Required
4= 2-Year Degree Required
5= 4-Year Degree Required

SELECTX = Composite Scale Indicating Scope of Selection Techniques Utilized by the
Institution for the Hiring of New Officers

BCKGRNDX = Composite Scale Indicating Scope of Background Investigation Utilized
by the Institution for the Hiring of New Ofiicers

CTRAIN = # of Classroom Training Hours Required

F TRAIN = # of Field Training Hours Required

Campus Public Safefl Department Structure and Functions
OUTSOURC = Department Outsourcing of Protective Services

l= No
2= Yes

ARREST = Public Safety Officers Patrolling Campus Granted Arrest Powers

l= No
2= Yes

DRUGLAWS = Public Safety Departments Responsible for Enforcing Narcotics Laws

1= No
2= Yes

158

DRUGTASK = Public Safety Departments Participating in Multi-Agency Drug Task
Force

1= No
2= Yes

APPENDIX B

Items Comprising Scales Used in OLS Regression Analyses

159

APPENDIX B

Items Comprising Scales Used in OLS Regression Analyses

Crime Prevention

Communications Technology Utilization (CO_Ml\_/IX) (alpha= .46)

1. Usage of Portable Radios All Items Coded 1= No 2= Yes
2. Usage of Motor Vehicle Radios

3. Usage of Cellular Phones

4. Usage of Base Station Radios

Substance Abuse Prevention Proms (SUBABUSX) (alpha= .89)

1. Special Unit- Alcohol Education Both Items Coded 1= No 2= Yes
2. Special Unit- Drug Education

General Crime Prevention Education Prog;a_m§ (PREVEN 11X) (alpha= .76)

1. Special Unit- Victim Assistance All Items Coded 1= No 2= Yes
2. Special Unit— Crime Prevention

3. Special Unit- Hate Crimes Prevention

4. Special Unit- Date Rape Prevention

5. Special Unit— Stranger Rape Prevention

6. Special Unit- Self-Defense Training

 

Campus Public Safety Dgpartment Selection and Training Criteria

Selection Criteria for New Officers (SELEC IX) (alpha= .74)

1. Selection Technique- Written Test All Items Coded 1= No 2= Yes
2. Selection Technique-- Board Interview

3. Selection Technique- Physical Agility Test

4. Selection Technique— Psychological Screening

5. Selection Technique- Medical Examination

160

6. Selection Technique-- Drug Screening
Background Investigation of New Officers (BCKGRNDX) (alpha= .49)

1. Selection Technique-- Background Investigation Both Items Coded 1= No 2= Yes
2. Selection Technique-- Criminal Record Check

LIST OF REFERENCES

161

LIST OF REFERENCES

Adler, Freda, Mueller, Gerhard O.W., and Laufer, William S. (1995). Criminology : The
Shorter Version (2nd Edition). New York, NY : McGraw-Hill.

Akers, Ronald L. (1990). Rational Choice, Deterrence, and Social Learning Theory in

Criminology : The Path Not Taken. The Journal of Criminal Law & Criminology.
81(3):653-676.

. (1991). Self-Control as a General Theory of Crime. Journal of Quantitative
Criminology.7 (3)2201-211.

Babbie, Earl (1992). The Practice of Social Research 6th Edition. Belmont, CA:
Wadsworth.

 

Beeler, Karl J ., Bellandese, Susan Dunton, and Wiggins, Carol A. (1991). Campus

Safeg: A Survey of Administrative Perceptions and Strategies. Washington,
DC: National Association of Student Personnel Administrators, Inc.

Bennett, Richard R. (1991) Routine Activities : A Cross-National Assessment of a
Criminological Perspective. Social Forces. 70(1): 147-163.

Benson, Bruce L. (1993). Community Policing Works at Michigan State University.
Journal of Securig Administration. 16(1):43-52.

Bordner, Diane C. and Petersen, David M. (1983) Campus Policing : The Nau_1re of
Universig Police Work. Lanham, MD : University Press of America.

Brantingham, Paul J. and Brantingham, Patricia L. (1984). Patterns in Crime. New York,
NY : Macmillan.

. (1991) Environmental Criminology (1991 Reissue). Prospect Heights, IL :
Waveland.

Bromley, Max L. (1992). Campus and Community Crime Rate Comparisons: A
Statewide Study. Journal of Securig Administration. 15(2):49-64.

162

. (1995a). Comparing Campus and City Crime Rates : A Descriptive Study.
American Journal of Police. 14(1):]31-148.

. (1995b). Analyzing Campus Crime and Police Resources : Implications for
Policy- Makers. Criminal Justice Policv Review. 7(2): 185-201.

. (1997). Motor Vehicle Thefts at an Urban University : An Exploratory Study.
Journal of Securigy Administration. 20(1):l-14.

Bromley, Max L. and Territo, Leonard (1990). College Crime Prevention and
Personal Safefl Awareness. Springfield, IL : Charles C. Thomas.

Bureau of Justice Statistics. Campus Law Enforcement Agencies, 1995 (NACJD Catalog
Number 6846). [Online] Available http://www.ojp.usdoj.gov/bj s/pub/ascii/clea95.
txt, December 11, 1996.

Chambliss, William J ., and Nagawasa, Richard H. (1969) On the Validity of Official
Statistics. Journal of Research in Crime and Delinquency. 6: 71-77.

Clarke, Ronald V.G. (1980). ‘Situational’ Crime Prevention : Theory and Practice.
British Journal of Criminology. 20(2): 136-147.

. (1988). Guest Editor’s Introduction to the Special Issue on Situational
Prevention. Journal of Securig Administration. 11(2):4-6.

. (ed.)(1992). Situational Crime Prevention: Successful Case Studies. New
York, NY: Harrow and Heston.

Clarke, Ronald V. and Cornish, Derek B. (eds.)(l983). Crime Contrql m’ Britain: A
Review of Policy and Research. Albany, NY: State University of New York
Press.

. (1985). Modeling Ofi‘enders’ Decisions: A Framework for Policy and
Research. In Crime and Justice Vol. 4. Michael Tonry and Norval Morris
(eds). Chicago, IL : University of Chicago Press.

 

Clarke, Ronald V.G. and Mayhew, Pat M. (1980). Desigmp' g Out Gm e. London:
HMSO.

Cohen, Lawrence E. (1981). Modeling Crime Trends : A Criminal Opportunity
Perspective. Journal of Research in Crime and Delinquency. l8(l):l38-164.

Cohen, Lawrence E. and Cantor, David (1981). Residential Burglary in the United States
: Life- Style and Demographic Factors Associated with the Probability of

Victimization. Journal of Research in Cm’ e and Delinquency. 18(1):113-127.

163

Cohen, Lawrence E. and Felson, Marcus (1979). Social Change and Crime Rate Trends :
A Routine Activity Approach. American Sociological Review. 44(August):588-
608.

Cole, George F. (1995). The American System of Criminal Justice (7th Edition).
Belmont, CA: Wadsworth.

College Entrance Examination Board (1997). The College Handbook. 1998, 35th Edition.
New York, NY: Guidance Publishing.

Cordner, Gary, Marenin, Otwin, and Murphy, John (1985). Police Responsiveness to
Community Norms: Guidance and Autonomy. American Journal of Police.
_5_(1):83-107.

Cornish, Derek B. and Clarke, Ronald V. (1986). The Reasoning Criminal : Rational
Choice Persppctives on Offending. New York, NY : Springer-Verlag.

. (1987). Understanding Crime Displacement : An Application of Rational
Choice Theory. Criminology. 25(4):933-47.

Crank, John P. (1990). The Influence of Environmental and Organizational Factors on
Police Style in Urban and Rural Environments. Journal of Research in Crime and

Delinquency. 27(2): 166-189.

Crime Awareness and Campus Security Act of 1990, 20 U.S.C. 1092 (1990).

Dodge, Susan (1991). With Campus Crimes Capturing Public Attention, Colleges Re-

Evaluate Security Measures and Stifien some Penalties. The Chronicle of Higher
Education. 37:A29-A30.

Egan, Timothy (1991). New Faces, and New Roles, for the Police. The New York Times.
April, 25:A1, B10.

Einstadter, Werner and Henry, Stuart (1995). CE 'polqgiqa; Theog : An Analysis of IQ
Underlyrp' g Assumptions. Fort Worth, TX: Harcourt Brace College Publishers.

Esposito, David and Storrner, David (1989). The Multiple Roles of Campus Law
Enforcement. Campus Law Enforcement Jqumal. 19(3):26-30.

Etheridge, Robert F. (1958). A Study of C_arppus Protective and Enforcement Agencies at
Selected Universities. Unpublished Ph.D. Dissertation. East Lansing, MI:

Michigan State University.

Felson, Marcus (1987). Routine Activities and Crime Prevention in the Developing
Metropolis. Criminology. 25(4):911-930.

164

. (1988). The Changing Ecology of Security. Journal of Securig
Administration. 11(2):8-11.

. (1994). Crime and Eve_rydav Life : InsightL and Implications for Society.
Thousand Oaks, CA : Pine Forge Press.

. (1997). A ‘Routine Activity’ Analysis of Recent Crime Reductions. The
Criminologist. 22(6): 1-3.

. (1998). Crime and Everyday Life (2nd Edition). Thousand Oaks, CA: Pine
Forge Press.

Felson, Marcus and Cohen, Lawrence E. (1980). Human Ecology and Crime: A Routine
Activity Approach. Hmnan Ecology. 8:3 89-406.

. (1981). Modeling Crime Rate Trends - A Criminal Opportunity Perspective.
Journalof Research in Crime and Delinquency. 18:138-164.

Fernandez, Adriana and Lizotte, Alan J. (1993). What’s Happening to Rates of Campus
Crime? The Public Persmctive (November/December):29-30.

. (1995). Relationship Between Campus Crime and Community Crime. I_r_1_

Campus Crime : Legal, Social, and Policy Persmctives. Bonnie S. Fisher and
John J. Sloan (eds.). Springfield, IL : Charles C. Thomas .

Fisher, Bonnie (1993). Guest Editor's Introduction to the Second Special Issue on Crime
and Fear of Crime on University and College Campuses and the Surrounding

Legal Issues. Journal of Securig Administration. 16(1):1-3.

.(1995). Crime and Fear on Campus. Annals of the American Academy of
Political and Social Science. 539:85-101.

Fisher, Bonnie and Nasar, Jack L. (1992). Students’ Fear of Crime and Its Relation to
Physical Features of the Campus. Jomnal of Securig Administration. 15(2): 65-
75.

Fisher, Bonnie S. and Sloan, John J. (1993). University Response to the Campus Security
Act of 1990: Evaluating Programs Designed to Reduce Campus Crime. Journal

of Securig Administration. 16(1):67-80.

Fisher, Bonnie S., Sloan, John J ., Cullen, Francis T., and Lu, Chien (1997). The On-
Campus Victimization Patterns of Students: Implications for Crime
Prevention by Students in Post-Secondary Institutions. In Crime Prevention at a
Crossroads. Steven P. Lab (ed.) Cincinnati, OH: Anderson.

165

Folven, Edwin L. (1995). Student Views of Campus Policing. Campus Law Enforcement
Journal. 25(1):21,25.

Fox, James Alan and Hellman, Daryl A. (1985). Location and Other Correlates of
Campus Crime. Journal of Criminal Justice. 13:429-444.

F ox, James W. (1977). Crime on Campus. Journg of College Student Personnel. 18(5):
345-351.

Fraenkel, Jack R. and Wallen, Norman E. (1993). How to Desigp and Evaluate Research
in Education (2nd Edition). New York, NY: McGraw-Hill.

Frost, Gregory A. (1993). Law Enforcement Responds to Campus Crime. Journal of
Securig Administration. l6(l):21-29.

Gardner, Erle Stanley (1970) Cops on Campus and Crime in the Streets. New York, NY:
William Morrow and Company, Inc.

Garofalo, James (1987). Reassessing the Lifestyle Model of Criminal Victimization. In

Positive Criminology : Essays in Honor of Michael J. Hindelang. Michael R.
Gottfredson and Travis Hirschi (eds.) Beverly Hills, CA : Sage.

Gearey, Robyn (1997). College Cop-Out: Closing the Books on Campus Crime. The New
Republic. 217(19):21-23.

Gelber, Seymour (1972). The Role of Campus Securig in the College Septqp' g.
Washington, DC : U.S. Government Printing Office.

Gove, Walter, Hughes, Michael, and Geerken, Michael (1985). Are Uniform Crime ‘
Reports a Valid Indicator of the Index Crimes?: An Afiirmative Answer with
Minor Qualifications. Criminology. 23:451-501.

Griffaton, Michael C. (1993). Forewarned is Forearmed: The Crime Awareness and
Campus Security Act. Case Western Law Review 43:525-590.

. (1995). State-Level Initiatives and Campus Crime. In Campus Crime : Legal,
Social, Qd Policy Pemctives. Bonnie S. Fisher and John J. Sloan (eds.).
Springfield IL : Charles C. Thomas.

Hair, Jr., Joseph R, Anderson, Rolph E., Tatham, Ronald L., Black, William C. (1995).
Multivariate Data Analysis (4th Edition). Upper Saddle River, NJ: Prentice Hall.

Haworth, Karla (1996). Education Department Finds a College Violated U.S. Crime-
Reporting Law. The Chronicle of Higher Education. 43(5):A44.

166

Hindelang, Michael J ., Gottfredson, Michael R., and Garofalo, James (1978). Victims of

Personal Crime : An Empirical Foundation for a Theog of Personal
Victimization. Cambridge, MA : Ballinger.

Huff, James A. (1990). Campus Security: The Mule Patrol. FBI Law Enforcement
Bulletin. 59(7): 14-17.

Hummer, Don and Biunphus, Vic W. (1996). Constituent of Satisfaction with Campus
Policing. Journal of Securifl Administration. 19(1): 1-15.

Hummer, Don, Austin, Thomas L., and Bumphus, Vic W. Forthcoming. Arming the
Campus Cops : A Descriptive and Multivariate Assessment of Support. Policing :

An International Jom'nal of Police Strategies and Mmement.

Jackson, Eric. (1992). Campus Police Embrace Community-Based Approach. The Police
Chief. 59(12):63-64.

Jackson, Patrick G. (1990). Sources of Data. In Measurement Issues in Criminology.
Kimberly L. Kempf (ed.). New York, NY: Springer-Verlag.

Jacobs, Jane (1961). The Death and Life of Great American Cities. New York, NY:

Random House.
Jeffery, C. Ray (1971).C rime Prevention Through Environmental Desigp. Beverly Hills,
CA: Sage.

Johnson, Robert P. (1994). Implementing Community Policing in a University
Environment. Campus Law Enforcement Journal. 24(3): 17-21,34-35.

Kessler, Rohn (1993). Preventing Violence and Crime on Campus. Journal of Securig
Administration. l6(l):53-63.

Kiecolt, K. Jill and Nathan, Laura E. (1985). Secondary Analysis of Survey Data. Sage
University Paper Series on Quantitative Applications in the Social Sciences, 07-
053. Beverly Hills, CA: Sage.

Kirtley, Jane (1997). Shedding Light on Campus Crime. American Journalism Review.
12(6): 50.

Klepper, Steven and Nagin, Daniel (1989). The Deterrent Effect of Perceived Certainty
and Severity of Punishment Revisited. Criminolqu. 27(4):721-746.

Klockars, Carl (1985). The Idea of Police. Beverly Hills, CA: Sage.

167

Lab, Steven P. (1992). Crime Prevention Approaches, Practices and Evaluations (2nd
Edition). Cincinnati, OH : Anderson.

. (ed.) (1997). Crime Prevention at a Crossroads. Cincinnati, OH: Anderson.

Lang, Paul (1993). Implementing Community-Based Policing on College Campuses.
Campus Law Enforcement Journal. 23(2):38-40.

. (1995). Organizational Structure: A Perspective for Campus Law
Enforcement. Campus Law Enforcement Journal. 24(6):29-32.

Lanier, Mark M. (1995). Community Policing on University Campuses: Tradition,

Practice, and Outlook. In Campus Crime : Legal, Social, and Policy
Persxctives. Bonnie S. Fisher and John J. Sloan (eds.). Springfield IL :

Charles C. Thomas.

Lederman, Douglas. (1993). Colleges Report 7,500 Violent Crimes on their Campuses in
First Annual Statements Required Under Federal Law. The Chronicle of Higher
Education. 39 (January):A32-A43.

. (1994). Weapons on Campus?. The Chronicle of Higher Education. 40
(March):A33- A34.

Lively, Kit (1997). More Than 10% of Colleges Fail to Publish Required Crime Reports,
Study Finds. The Chronicle of flgher Education. 43(26):A36.

Lizotte, Alan J. and Fernandez, Adriana (1993). Trends and Correlates of Campus Crime
: A General Rpmrt. Albany, NY : Consortium of Higher Education Campus

Crime Research.

Lunden, Walter A. (1962) The Theory of Crime Prevention. British Journal of
Criminology. 2:213-228.

Lunden, Walter A. (1983). A Decade of Crime on Campus. Thp Police Chief.
(September): 66-8.

Maxfield, Michael G. (1987). Lifestyle and Routine Activity Theories of Crime :
Empirical Studies of Victimization, Delinquency, and Offender Decision-Making.

Journal of Mtitative Cmm ' ology. 3(4):275-282.

Maxfield, Michael G. and Babbie, Earl (1998). Research Methods for Criminal Justice
and Cmp’ ' ology, 2nd Editiqn. Belmont, CA: West/Wadsworth.

168

Mayhew, Pat (1991). Crime in Public View : Surveillance and Crime Prevention. In
Environmental Criminology. Paul J. Brantingham and Patricia L. Brantingham
(eds.) Prospect Heights, IL : Waveland.

McEvoy, Sharlene A. (1992). Campus Insecurity: Duty, Forseeability, and Third Party
Liability. Journal of Law and Education. 21(2): 137-154.

McPheters, Lee R (1978). Econometric Analysis of Factors Influencing Crime on
Campus. Journal of Criminal Justice. 6(1): 47-52.

Middleton, Lorenzo (1981). Many Campuses Losing Ground in War Against Crime. The
Chronicle of Higher Education. 27(October ):5-6.

Miller, J. L. and Pan, Maureen J. (1987). Student Perceptions of Campus Police : The
Effects of Personal Characteristics and Police Contacts. American Journal of
Police. 6(2):27-44.

Miller, Roger C. (1990). Physical Environment Desigp and the Occurrence of Crime in
Campus Residence Halls. Unpublished Master’s Thesis. East Lansing, MI:

Michigan State University.

Monette, Duane R., Sullivan, Thomas J ., and DeJong, Cornell R (1994). Applied Social
Research, 3rd Edition. Fort Worth, TX: Harcourt Brace College Publishers.

Monkkonen, Eric H. (1992). History of Urban Police. In Modern Policing. Michael H.
Tonry and Norval Morris (eds) Chicago, IL: University of Chicago Press.

Morash, Merry and Haarr, Robin N. (1995). Gender, Workplace Problems, and Stress in
Policing. Justice Quarterly. 12(1):l 13-140.

Nagin, Daniel S. and Paternoster, Raymond (1993). Endming Individual Differences and
Rational Choice Theories of Crime. Law and Socim Review.27(3):467-496.

Newman, Jeffi'ey, A. (1996). Civil Liability and Campus Police. Campus Law
Enforcement Journal. 26(1):3-7.

Newman, Oscar (1972). Defensible Spy. New York, NY : Macmillan.

Nichols, W. David (1995). Violence on Campus - The Intruded Sanctuary. FBI Law
Enforcement Bulletin. 64(June): 1-5.

Nielsen, Swen C. (1970) General Ob ations of Or anizational and A ' 'Strative
Concepts for Universipy Police. Unpublished Master’s Degree Thesis. Provo, UT :
Brigham Young University.

169

Onishi, Norimitsu (1995). Face-Off at Small Colleges on Arming of Security. The New
York Times. 144(August 16):Bl l.

Pagon, Milan (1995). Policing the University Community : Police Discretion and Officer-
Student Relations. Campus Law Enforcement Journal. 25(1):3-8,35.

Palmer, Carolyn J. (1993). Violent Crimes and Other Forms of Victimization 1p
Residence Halls. Asheville, NC: College Administration Publications.

Paternoster, Raymond (1989). Absolute and Restrictive Deterrence in a Panel of Youth:
Explaining the Onset, Persistence/Desistence and Frequency of Delinquent
Offending. Social Problem. 36(3):289—309.

Peak, Kenneth J. (1988). Campus Law Enforcement: A National Survey of
Administration and Operation. Campus Law Enforcement JourJal. l9(5):33-35.

. (1989) Campus Law Enforcement in Flux: Changing Times and Future
Expectations. Campus Law Enforcement Journal. 19(6):21-25.

(1995). The Professionalization of Campus Law Enforcement : Comparing
Campus and Municipal Law Enforcement Agencies. In Campus Crime : Legal,
Social, and Policy Persmctives. Bonnie S. Fisher and John J. Sloan (eds.).
Springfield IL : Charles C. Thomas.

Peterson’s (1998). Graduate Prom in the Humanities, Arts and Social Sciences 1998
(Book 2), 32nd Edition. Princeton, NJ: Peterson’s.

Phillips, Peter W. (1993). Campus Law Enforcement as Community Policing and
Problem-Oriented Policing. Campus Law Enforcement Jam. 23(6): 19—22.

Pollard, Jeffrey W. and Whitaker, Leighton C. (1991). Cures for Campus Violence, If We

Want Them. In Campus Violence : Kinds, Cagsqs, and Cures. Leighton C.
Whitaker and Jeffrey W. Pollard (eds.). New York, NY: Haworth Press.

Powell, John W. (1994a). Campus Seem and Law Enforcement (2nd Edition).
Woburn, MA: Butterworth-Heinemann.

. (1994b). The Beginning-Yale campus Police Deparunent, 1894. Campus Law
Enforcement Journal. 24(4):2-5.

Powers, Andrew D. (1996). Community Policing - Reducing Fear of Crime on Campus.
Campus Law Enforcement Journal. 26(3):21-23.

170

Reaves, Brian A. and Goldberg, Andrew L. (1996). Campus Law Enforcement Agencies,
1995. Washington, DC: U.S. Department of Justice, Office of Justice Programs,
Bureau of Justice Statistics.

Rengert, George F. (1988). The Location of Facilities and Crime. Journal of Security
Administration. 1 1(2): 12-1 6.

Reppetto, Thomas A. (1976). Crime Prevention and the Displacement Phenomenon.
Crime and Delinquency. 22(2):]68-177.

Richards, George E. (1996). The Security Survey: Creating a Proactive Foundation for

Campus Crime Prevention. Journal of Contemmpqpy Criminal Justice. 12(1): 45-
53.

Riesling, Susan (1993). Problem Oriented Policing at the University of Wisconsin.
Journal of Securig Administration. 16(1):31-42.

Roark, Mary L. (1987). Preventing Violence on College Campuses. Journal of
Counseling and Development. 65(March):367-371.

. (1993). Conceptualizing Campus Violence : Definitions, Underlying Factors,

and Effects. In Campus Violence : Kinds, Causes, and Cures. Leighton C.
Whitaker and Jeffrey W. Pollard (eds.). New York, NY: Haworth Press.

Roncek, Dennis W. and Maier, Pamela A. (1991). Bars, Blocks, and Crimes Revisited :
Linking the Theory of Routine Activities to the Empiricism of ‘Hot Spots’.
Criminology 29(4): 725-753.

Rosenbaum, Dennis P. (1986). The Problem of Crime Control. In Community Crime
Prevention: Does it Work? Dennis P. Rosenbaum (ed.). Beverly Hills, CA : Sage.

Rubenstein, Harvey M., Murray, Charles A., Motoyama, Tetsuro and Rouse, Victor W.
(1980). The Link Between Crime and me Built Environment: The Current State of
Knowledge (yolume 1). (National Institute of Justice). Washington, DC: U.S.
Government Printing Office.

Scott, Eric J. (1976). College and Univgrpig Police Agencies. Bloomington, IN: Indiana
University Press.

Selke, William L. (1977). A Systems Model Approach to Evaluating Social Programs:
The Case of Yputh Service Bureaus. Unpublished Ph.D. Dissertation. East

Lansing, MI: Michigan State University.

171

Seng, Magnus J. (1995). The Crime awareness and Campus Security Act: Some
Observations, Critical Comments, and Suggestions. In Canhgus Crime : Lega_l_,_
Social, and Policy Persmctives. Bonnie S. Fisher and John J. Sloan (eds.).
Springfield IL : Charles C. Thomas.

 

Seng, Magnus J. and Koehler, Nancy S. (1993). The Crime Awareness and Campus
Security Act : A Critical Analysis. Journal of Crime and Justice. 16(1):97-110.

Sewell, James D. (1993). Campus Security in Today’s Institutions of Higher Education :
Asking the Right Questions. Journal of Securig Administration. 16(1):7-17.

Sherman, Lawrence W., Gartin, Patrick R, and Buerger, Michael E. (1989). Hot Spots of
Predatory Crime. Criminology. 27(1): 27-55.

Siegel, Dorothy G. and Raymond, Clarinda Haniss (1992). An Ecological Approach to
Violent Crime on Campus. Journal of Security Administration. 15(2):19-29.

Siegel, Larry J. (1989). Criminology, 3rd Edition. St. Paul, MN: West.

Sigler, David and Koehler, Nancy (1993). Victimization and Crime on Campus.
International Review of Victimology. 22331-343.

Sims, Olyn Suthem (1971). New Directions in Campus Law Enforcement. Athens, GA :
University of Georgia Center for Continuing Education.

Sims, Victor (1988). Small Town and Rural Police. Springfield, IL: Charles C. Thomas.

Skogan, Wesley G. (1974). The Validity of Official Crime Statistics: An Empirical
Investigation. Social Science Merly. 55:25-38.

Sloan, John J. (1992a). Campus Crime and Campus Communities : An Analysis of
Crimes Known to Campus Police and Security. Journal of Security
Administration. 15(2):31-47.

. (1992b). The Modern Campus Police : An Analysis of Their Evolution,
Structure, and Function. American Journal pf Police. 11(2):85-104.

. (1994). The Correlates of Campus Crime: An Analysis of Crimes Known to
Campus Police. Journal of Cr_1n_nn_al' ' Jpsflce. 22:51-62.

Sloan, John J. and Fisher, Bonnie S. (1995). Campus Crime : Legal, Social, and Policy

Contexts. In Campus Crime : Legal, Social, and Policy Pegpectives. Bonnie S.
Fisher and John J. Sloan (eds.). Springfield IL : Charles C. Thomas.

172

Sloan, John J ., Fisher, Bonnie S., and Wilkins, Deborah L. (1993a) Crime and Related

Issues on the UAB Campus: Initial rep_ort from a Two YeaLSmdv. Birmingham,
AL: Report to the Office of Vice President for Administration, University Of

Alabama-Birmingham.

. (1993b). Social Control on Campus : Attitudes Towag'd University Police bx
Members of an Urban Campus Community. Unpublished Paper presented at the
Annual Meetings of The American Society of Criminology, Phoenix, AZ
(October).

Sloan, John J ., Fisher, Bonnie S., and Cullen, Francis T. (1997). Assessing the Student
Right-to-Know and Campus Security Act of 1990 : An Analysis of the Victim
Reporting Practices of College and University Students. Crime and

Delinquency. 43(2): 148-168.

Smith, Michael Clay. (1988). Coping with Crime. New York, NY : Macmillan.

. (1989). Cops or Watchmen? Policy Considerations in the Role of Campus
Security. Campus Law Enforcement Journal. 19:38-39.

. (1992). Vexatious Victims of Campus Crime: Student Lawsuits as Impetus for
Risk Management. Journal of Security Amg' 'stration. 15(2):5-17.

Smith, Michael Clay and F ossey, Richard (1995). Crime on Campus : Legal Issues and
Campus Administration. Phoenix, AZ : American Council on Education/Oryx
Press.

StatSoft (1995). Statistica for Windows Volume 1: General Conventions and Statistics I.
Tulsa, OK: StatSoft, Inc.

Storrner, David E. (1989). Crime Statistics - Crime Rates. Campus Law Enforcement
Journal. 19(1): 4.

Storrner, David E. and Senarath, D.T.M. (1992). The Truth About Campus Crime : An
Analysis of Campus Crime Reports for Three Years in Pennsylvania. Campus
Law Enforcement Journal. 22(4):28-32.

Sullivan, Robert R. (1994). The Tragedy of Academic Criminal Justice. J opmaj of
Criminal Justice. 22(6):549-558.

Taylor, Ralph B. (1994). Research Methods in Criminal Justice. New York, NY:
McGraw-Hill.

173

Territo, Leonard, Halsted, James B., and Bromley, Max L. (1995). Crime & Justice in
America : A Human Perspective (4th Edition). Minneapolis/St. Paul, MN : West.

Tillinghast, David H. (1996). In the Line of Fire: The Case for Armed Campus Police.
Campus Law Enforcement Journal. 26:21-22.

Trojanowicz, Robert C., Benson, Bruce, and Trojanowicz, Susan (1988). Communitv

Policing : Universig Input into Campus Police Policy-Making. East Lansing, MI:
National Neighborhood Foot Patrol Center.

 

Tucker, Melvin L. and Stames, Bill (1993). Crime Prevention Through Environmental
Design : The Tallahassee Model. The Police Chief. (October): 130-133.

Utz, Thomas E. (1993). Gotcha! Campus Crime Prevention Program Earns High Marks.
FBI Law Enforcement Bulletin. 62(1 1):17.

Wallis, Allan and Ford, Daniel (1980). Crime Prevention Through Environmental
Desigp: An Omrational Handbook. (U .S. Department of Justice) Washington,

DC: U.S. Government Printing Office.

Weisheit, Ralph A., Wells, L. Edward, and Falcone, David N. (1994). Community
Policing in Small Town and Rural America. Crime and Delinquency. 40(4):549-
567.

Wilson, James Q. and Kelling, George L. (1982). The Police and Neighborhood Safety:
Broken Windows. Atlantic Monthly. 127: 29-38.

Wilson, James Q. and Herrnstein, Richard (1985). Crime and Human Nature: The
Definitive Study of the Causes of Crime. New York, NY: Simon and Schuster.

Yu, Jiang and Liska, Allen E. (1993). The Certainty of Punishment : A Reference Group
Effect and Its Functional Form. Criminology. 31(3):447-464.

Zahm, Diane and Perrin, Dana (1992). Safe for Study : Designing the Campus
Environment. Journal of Securig Administratiop, 15(2):77-89.

MICHIGQN smTE UNIV. LIBRRRIES
lllllllllllllll Ill lllllllllllllll”WWW
31293016880548