what’s driving fatal law enforcement...
TRANSCRIPT
WHAT’S DRIVING FATAL LAW ENFORCEMENT COLLISIONS?
A STATE-LEVEL ANALYSIS
by
Bryon G. Gustafson
B.S., Excelsior College, 2003
M.P.A., University of Southern California, 2006
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Affairs
2012
ii
© 2012
BRYON G. GUSTAFSON
ALL RIGHTS RESERVED
iii
This thesis for the Doctor of Philosophy degree by
Bryon G. Gustafson
has been approved for the
Ph.D. in Public Affairs
by
Mary Dodge, Chair and Advisor
Paul Teske
Gerald L. Williams
Michael Hooper
Chester A. Newland
November 5, 2012
iv
Gustafson, Bryon G. Ph.D., Public Affairs
What’s Driving Fatal Law Enforcement Collisions? A State-Level Analysis
Thesis directed by Professor Mary Dodge
ABSTRACT
Traffic collisions are the leading cause of death for law enforcement officers
(LEOs) in the United States. As automotive and highway technologies have improved,
traffic fatalities in the general public have steadily decreased since reaching an all time
high in 1972. This has not been the trend among LEOs as they have generally
experienced increased or steadily persistent incidence of fatal traffic collisions since the
1970s. Even so, significant variability exists among states in terms of LEO traffic fatality
rates.
This research investigates this variability among states employing a mixed-
methods, state-level analysis to identify policy-relevant areas of opportunity to address
these LEO traffic fatalities. The research primarily covers the period 1995 to 2009 and
includes the 50 states and the District of Columbia. This study utilizes data from the
Federal Bureau of Investigation Crime in the United States and Law Enforcement
Officers Killed and Assaulted reports, the National Highway Transportation Safety
Administration Fatality Analysis Reporting System, the International Association of
Directors of Law Enforcement Standards and Training Sourcebook, and the California
Commission on Peace Officer Standards and Training State-Level Differences in Law
Enforcement Officer Traffic Fatalities survey, among others. Differences in state
highway spending, regulatory policy, law enforcement training, LEO and general public
traffic fatalities, and other state-level variables are explored through cross-sectional
v
regression analysis and qualitative content analysis. A state-level description of LEO
traffic fatalities and a complete dataset are also provided.
Results indicate a number of significant state-level variables and trends that
contribute to differences in LEO traffic fatality rates. Maximum highway speed limits
and general population traffic fatalities are found to be statistically significant predictors
of LEO traffic fatalities. Themes relating to (a) LEO Exceptionalism, (b) Agency
Sovereignty, (c) Training, and (d) External Control Loci, emerged from survey responses.
Policy implications are discussed and an implementation model is proposed for
enacting policy changes to reduce LEO traffic fatalities. Overall, many areas of
opportunity exist for increasing safety and reducing loss of life. However, political will
and buy-in are requisite components for state-level implementation and policy change.
An agenda for future research also is proposed.
The form and content of this abstract are approved. I recommend its publication.
Approved: Mary Dodge
vi
DEDICATION
This dissertation is dedicated to the memory of the 840 law enforcement officers
who died between 1995 and 2009 in the traffic collisions investigated in this study.
These women and men gave their lives in the line-of-duty in service to communities
throughout the nation. In dedicating this work to their memory, my hope is that lessons
learned from their sacrifices will prevent future traffic fatalities.
vii
ACKNOWLEDGMENTS
I would not have completed this research if it were not for the support of family,
friends, and colleagues. This list will undoubtedly be incomplete—I apologize in
advance for any oversights. Among the many who helped me get here, I would like to
specifically recognize the following:
§ Above all, I thank my wife, Dr. Sara E. McClellan. She set an example of excellence,
discipline, and perseverance. Through years of long nights and lost weekends, she
was a constant and understanding support. But for Sara, I would not have finished.
§ My parents, Ken and Nelda Gustafson, and my in-laws, Art and Diana McClellan, as
well as other family and friends, provided ongoing encouragement that was much
appreciated and needed.
§ My friend and colleague Dr. Kevin Wehr of California State University, Sacramento
read multiple drafts and provided thoughtful feedback as well as personal coaching
and encouragement.
§ My friends and colleagues Dr. Aaron Conley, almost-Dr. Liz Tomsich (fingers
crossed for May 2013!), and Dan Benhammou helped me through the personal and
professional struggles of balancing life and finishing the dissertation.
§ My advisor Dr. Mary Dodge helped me through and kept the pressure on when I
needed it most. I greatly appreciated her patience and trust.
§ My committee member Dr. Paul Teske, Dean of the School of Public Affairs, made
time that he didn’t have to serve on my committee and support my research.
§ My committee member Dr. Jerry Williams came out of retirement to provide me with
much appreciated professional and academic counsel.
viii
§ My committee member Dr. Mike Hooper, Bureau Chief of the Center for Leadership
Development at CalPOST, took on this role in addition to many important CalPOST
responsibilities. I appreciated his careful reading and thoughtful feedback.
§ My committee member and mentor, Dr. Chester A. Newland, Duggan Distinguished
Professor Emeritus of the University of Southern California Price School of Public
Policy, has provided me with more than a decade of supportive mentorship. The best
things I know of public service I learned from Dr. Newland.
§ The California Commission on Peace Officer Standards and Training (CalPOST) and
Executive Director Paul Cappitelli made this undertaking possible.
§ CalPOST Assistant Executive Director Alan Deal was an amazing coach. He
checked on my progress regularly and read far more drafts and working papers over
the years than any executive should.
§ The CalPOST SAFE Driving Research Team (Dr. Geoff Alpert, Dr. Bryan Vila, and
others noted elsewhere here) were excellent sounding boards for ideas and
methodological approaches.
§ Many other CalPOST colleagues provided support: Assistant Executive Director
Robert Stresak, Bureau Chief Michael Gomez, Bureau Chief Ed Pecinovsky, and
Senior Law Enforcement Consultant Steve Craig.
§ Dr. Tom Rice of the University of California, Berkeley shared insights and provided
feedback on my statistical models. He also encouraged calm.
§ Last, but not least, my precious cats, Zoca and Novio, were always nearby and loving.
ix
TABLE OF CONTENTS
CHAPTER
I. THE ISSUE & ITS PROBLEMS FOR PUBLIC POLICY .......................................... 1
Distribution and Counts of LEO Deaths ................................................................. 2
Author Background—Expertise and Disclosure ..................................................... 6
Employment History and Issue Insight ............................................................. 6
Personal Experience with Driving and Luck .................................................... 7
Dying at Work ......................................................................................................... 8
Real People in Real Places .............................................................................. 10
Assessing the Problem .......................................................................................... 13
Argue and Avoid—It May Be a Problem, But Certainly Not Mine ............... 15
Weigh the Costs .............................................................................................. 16
Significance of the Problem .................................................................................. 17
Engaged Scholarship—Research for Everyone’s Sake ........................................ 18
The Moral Imperative ..................................................................................... 18
Scholarly and Practical Utility ........................................................................ 19
Goals, Objectives, and Contributions to Literature .............................................. 21
Policy Implications and Broad Impacts .......................................................... 21
Intended Audience and Target of This Research ............................................ 22
Outline for the Study ....................................................................................... 22
II. LITERATURE REVIEW ........................................................................................... 24
Historical Context ................................................................................................. 25
Perspectives on the Problem ................................................................................. 28
x
Previous Scholarship ............................................................................................. 29
Relevant Boundaries ....................................................................................... 30
Problems and Solutions—Prior Understandings and Responses .................... 32
Research Gap and Need ........................................................................................ 43
LEO Fatalities as a Public Policy Problem ..................................................... 44
Moving On to the Research .................................................................................. 45
III. RESEARCH QUESTIONS, DATA, & METHODS .................................................. 46
Research Question and Sub-Questions ................................................................. 46
Research Question 1 ....................................................................................... 46
Research Question 2 ....................................................................................... 46
Research Question 3a ...................................................................................... 46
Research Question 3b ..................................................................................... 46
Relationships Among Research Questions and Sub-Questions ...................... 47
Rationale for Applying a Regulatory and Economic Framework ........................ 47
Propositions .......................................................................................................... 49
Proposition 1 ................................................................................................... 50
Proposition 2 ................................................................................................... 50
Concepts and Definitions ...................................................................................... 51
Data Sources, Types, Description, and Issues ...................................................... 51
Sources ............................................................................................................ 52
Types and Description .................................................................................... 53
Data Limitations and Challenges .................................................................... 57
Methods and Analytical Techniques ..................................................................... 60
xi
Descriptive Statistical Analysis ...................................................................... 60
Regression Analysis ........................................................................................ 61
Content Analysis ............................................................................................. 63
Validity and Limitations ....................................................................................... 65
Summary ............................................................................................................... 66
IV. DESCRIPTIVE ANALYSIS ...................................................................................... 68
Basic Descriptive Analysis ................................................................................... 68
Ranking the States ................................................................................................. 83
Best States by Various Measures .................................................................... 84
Worst States by Various Measures ................................................................. 89
Summary and Next Steps ...................................................................................... 94
V. QUANTITATIVE & QUALITATIVE ANALYSIS .................................................. 95
Statistical Modeling .............................................................................................. 95
Preparing the Data ........................................................................................... 95
Building the Model ......................................................................................... 99
Qualitative Analyses of Survey Responses ......................................................... 105
Multiple Choice and Likert Responses ......................................................... 108
Narrative Content Analysis ........................................................................... 125
Summary and Next Steps .................................................................................... 132
VI. DISCUSSION & POLICY IMPLICATIONS .......................................................... 134
Review of Findings ............................................................................................. 134
Present Study ................................................................................................ 135
Previous Scholarship Reprised ..................................................................... 137
xii
Policy Implications of Findings .......................................................................... 139
Potential Changes and Adaptations .............................................................. 141
Anticipated Challenges and Resistance ........................................................ 145
Examples for Review and Adaptation .......................................................... 148
Summary ............................................................................................................. 149
VII. PATHWAY TO IMPLEMENTATION & FUTURE RESEARCH ......................... 151
Research Implementation Gap ............................................................................ 151
Detailing the Gap and Defining Policy ............................................................... 152
Theory to Practice ............................................................................................... 153
Policy Implementation and Implementation Science ......................................... 155
Examples of Relevant Frameworks .............................................................. 156
National Implementation Research Network Model .................................... 158
Summary of Recommendations and Actions ...................................................... 161
Future Research Agenda ..................................................................................... 163
Summary and Conclusion ................................................................................... 164
REFERENCES .................................................................................................................... 166
APPENDIX
A. CalPOST (2012) Survey Instrument Summary .............................................................. 181
B. Gustafson Dataset ........................................................................................................... 192
xiii
LIST OF TABLES
Table
II.1 Types and Sources of State-Level Data .................................................................... 52
IV.1 Descriptive Statistics of Comparative State Variables 1995-2009. ......................... 70
IV.2 Descriptive Statistics for LEO Traffic Fatalities by State 1995-2009. .................... 71
IV.3 Top 10 States by Total Number of LEO Traffic Fatalities 1995-2009. ................... 84
IV.4 Top 10 States by Mean LEO Traffic Fatality Rate 1995-2009. .............................. 85
IV.5 Top 10 States by Mean General Public Traffic Fatality Rate 1995-2009. .............. 86
IV.6 Top 10 States by Balance of Mean Traffic Fatality Rates 1995-2009. ................... 87
IV.7 Bottom 10 States by Total Number of LEO Traffic Fatalities 1995-2009. ............. 89
IV.8 Bottom 10 States by Mean LEO Traffic Fatality Rate 1995-2009. ......................... 90
IV.9 Bottom 10 States by Mean Public Traffic Fatality Rate 1995-2009. ...................... 91
IV.10 Bottom 10 States by Balance of Mean Traffic Fatality Rates 1995-2009. ............ 92
V.1 Frequency of Grouped Maximum Daytime Speed Limit. ........................................ 97
V.2 Frequency of Categories of Driver Training Hours. ................................................. 98
V.3 Results of the Gustafson Model. ............................................................................. 102
V.4 Responding States and Function of Responding Agency. ...................................... 105
V.5 Frequency and Mean of Responses to CalPOST (2012) Question 12. ................... 120
V.6 Frequency and Mean of Responses to CalPOST (2012) Question 14. ................... 122
V.7 Frequency and Mean of Responses to CalPOST (2012) Question 14. ................... 124
V.8 States and Agencies Responding to CalPOST (2012) Questions 16 and 17. ......... 125
V.9 Recurring Categories in CalPOST (2012) Survey Responses. ............................... 128
xiv
LIST OF FIGURES
Figure
I.1 Distribution of LEO Deaths 1970-2009. ...................................................................... 2
I.2 LEO Population and Traffic Deaths 1995-2010. .......................................................... 3
I.3 US Population and Traffic Deaths 1995-2010. ............................................................ 4
I.4 Mean LEO Population and Total LEO Traffic Fatalities 2000-2009. .......................... 5
I.5 US Workers and Distribution of On-the-Job Deaths 2010. .......................................... 9
I.6 US LEOs and Distribution of On-the-Job Deaths 2010. ............................................ 10
I.7 Richmond Police Officer Brad Moody and Rico. ...................................................... 11
I.8 Officer Moody’s Patrol Car After His Fatal Collision. .............................................. 12
II.1 Universe of Traffic Collisions and Possible Subsets. ................................................ 49
IV.1 Histogram of Total LEO Traffic Fatalities 1995-2009. ........................................... 69
IV.2 Histogram of Total LEO Traffic Fatalities by State 1995-2009. ............................. 74
IV.3 Total Counts of LEO Traffic Fatalities by State 1995-2009. .................................. 75
IV.4 LEO Traffic Fatalities per 100,000 LEOs by State 1995-2009. .............................. 76
IV.5 Histogram of Mean State LEO Traffic Fatality Rate 1995-2009. ........................... 77
IV.6 Mean Annual General Public Traffic Fatalities by State 1995-2009. ..................... 78
IV.7 General Public Traffic Fatalities per 100,000 Residents by State 1995-2009. ........ 80
IV.8 Histogram of Mean State General Public Fatality Rate 1995-2009. ....................... 81
IV.9 Distribution of Traffic Fatality Rates by State 1995-2009. ..................................... 82
IV.10 Histogram of LEO to General Public Fatality Rate Balance 1995-2009. .............. 83
IV.11 Best States on One or More Mean Measures 1995-2009. ..................................... 88
IV.12 Worst States on One or More Mean Measures 1995-2009. ................................... 93
xv
V.1 Histogram of Maximum Daytime Speed Limit. ....................................................... 96
V.2 Histogram of Minimum Driver Training Hours. ...................................................... 97
V.3 Responses to CalPOST (2012) Question 2 (Agency Type). ................................... 108
V.4 Responses to CalPOST (2012) Question 3 (In-Service Training). ......................... 110
V.5 Responses to CalPOST (2012) Question 5 (Driver Training). ............................... 112
V.6 Responses to CalPOST (2012) Question 6 (Mandatory Seat Belt Use). ................ 113
V.7 Responses to CalPOST (2012) Question 7 (LEO Talk/Text and Drive). ............... 114
V.8 Responses to CalPOST (2012) Question 8 (LEO Speed Limits). ........................... 115
V.9 Responses to CalPOST (2012) Question 9 (LEO Pursuit Laws). ........................... 116
V.10 Responses to CalPOST (2012) Question 10 (Shift Length/Overtime). ................ 116
V.11 Responses to CalPOST (2012) Question 11 (Move Over Laws). ......................... 117
V.12 Responses to CalPOST (2012) Question 12 (Laws/Regulation). ......................... 119
V.13 Responses to CalPOST (2012) Question 14 (Funding/Highway Safety). ............ 122
V.14 Responses to CalPOST (2012) Question 15 (Traffic Fatality Rates). .................. 123
VII.1 Implementation Drivers (NIRN, 2012) ................................................................ 160
xvi
LIST OF ABBREVIATIONS
BCE Before Common Era BLS Bureau of Labor Statistics CALEA Commission on Accreditation for Law Enforcement Agencies CalPOST California Commission on Peace Officer Standards and Training EBP Evidence-Based Practices FBI Federal Bureau of Investigation GPS Global Positioning System IACP International Association of Chiefs of Police IADLEST International Association of Directors of Law Enforcement Standards and
Training LEO Law Enforcement Officer LEOKA Law Enforcement Officers Killed and Assaulted MPH Miles Per Hour NHTSA National Highway Traffic Safety Administration NIOSH National Institute for Occupational Safety and Health NIRN National Implementation Research Network NIRNM National Implementation Research Network Model NLEOMF National Law Enforcement Officers Memorial Fund NTSB National Transportation Safety Board ODMP Officer Down Memorial Page POST Peace Officer Standards and Training SWITRS Statewide Integrated Traffic Reporting System UK United Kingdom US United States
1
CHAPTER I
THE ISSUE & ITS PROBLEMS FOR PUBLIC POLICY
“What we live by we die by.” – Robert Frost (1917, p. 121)
Historically, and in the 1970s especially, the job of a law enforcement officer
(LEO)—whether police, sheriff, or highway patrol—was dangerous. For those who
would be killed in the line of duty, it was likely they would die violently (Klinger, 2004).
Death at the hands of a felon was nearly twice as probable as an accidental death
(National Law Enforcement Officers Memorial Fund (NLEOMF), 2009). This trend
lessened during the 1980s as felonious deaths declined and, by the mid-1990s, accidental
deaths in general surpassed felonious killings as the leading cause of death for LEOs
nationwide (NLEOMF, 2009). In simple language, more officers were dying by accident
than at the hands of criminals.
The changing trend during the 1980s and early 1990s was largely due to the
decline of felonious killings of LEOs (NLEOMF, 2009). This trend is distinct from a
reduction in violence. Criminals were still attempting to kill officers, but the officers
were not as likely to die. Numerous industry initiatives and technological advancements
such as soft body armor (i.e., bullet-proof vests), weapons retention holsters, mobile
communications, and improvements in emergency medical response systems and trauma
centers helped to keep officers alive in the aftermath of what might previously have been
fatal events (Federal Bureau of Investigation (FBI), 2008b; Klinger, 2004). Developing
social and economic conditions in the mid-1990s also contributed to the subsequent
reduction in violence that closed out the decade (Harcourt & Ludwig, 2006; Levitt,
2005). At roughly the same time, accidental deaths began to increase and traffic
2
collisions were the most prevalent of all accidents (Gustafson, 2009; NLEOMF, 2009).
Even so, as one threat diminished, another emerged.
Distribution and Counts of LEO Deaths
For context, consider that “[a]n average of 162 officers a year died in the 2000s,
compared with 160 a year in 1990s, 190 in the 1980s, and 229 in the 1970s” (NLEOMF,
2009, p. 1). On the whole, there has been a downward trend in the total number of officer
deaths over these past decades. A salient issue for this study is the substantial change in
the distribution of the deaths. “[I]n the 1970s, 62 percent of all officer deaths were
felonious killings; in the 1980s, the figure was 54 percent” (NLEOMF, 2009, p. 2). In
the 1990s, for the first time, felonious deaths dropped to just below 50 percent of total
officer deaths (FBI, 1999b). Accidental deaths became the predominant cause and traffic
fatalities trended upward substantially through the 1990s. This trend maintained into the
2000s (FBI, 2009b). Figure I.1 illustrates the change in distribution over time.
Figure I.1 Distribution of LEO Deaths 1970-2009.
38% 46% 50.1% 57.6%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
1970s 1980s 1990s 2000s
Accidental Felonious
3
In the 17 years from 1980 to 1996 there was just one (1988) when more than 50
officers were killed in traffic (National Highway Traffic Safety Administration
(NHTSA), 2010b). In the 14 years thereafter, from 1997 to 2010, there was just one
(2009) when fewer than 50 were killed in traffic. Even allowing for population
differences and regression to the mean, the trend is decidedly upward. Figure I.2
illustrates the trend in LEO traffic fatalities with the backdrop of LEO population
changes for the period 1995-2010.
Figure I.2 LEO Population and Traffic Deaths 1995-2010.
Given this historical trend data, it seems evident that traffic deaths became an
epidemic for LEOs beginning in 1997 and remain an epidemic in 2012. This persistent
increase in traffic fatalities is the basis for this study. Specifically, this research focuses
on the distribution of LEO traffic deaths by state (i.e., why significant variance exists)
and ultimately it aims to identify public policy interventions/solutions to limit these
traffic deaths in order to save lives and public resources.
0
100000
200000
300000
400000
500000
600000
700000
800000
0
10
20
30
40
50
60
70
1995
19
96
1997
19
98
1999
20
00
2001
20
02
2003
20
04
2005
20
06
2007
20
08
2009
20
10
Po
pula
tion
Dea
ths
LEO Population LEO Deaths
4
For an additional point of reference, consider that while officer traffic fatalities
have trended up since the 1990s, traffic deaths in the general driving public have
remained constant or declined (Gustafson, 2009; Gustafson & Cappitelli, 2010; NHTSA,
2010a). Better cars and safety technology (e.g., air bags, skid control technology), better
roadways overall, additional and improved training—all of these advancements would be
expected to lead to lower death rates. This is the case for the driving public at large
(NHTSA, 2010a), but traffic collision actuarial data indicate that this is not the case for
LEOs. This trend in the general population is illustrated in Figure I.3.
Figure I.3 US Population and Traffic Deaths 1995-2010.
The trend in law enforcement traffic deaths poses a serious public policy problem
for policing professionals, and scholars have an opportunity to contribute to improved
knowledge about the factors contributing to this trend. Further, while this trend is evident
nationally, significant variability in LEO traffic fatality rates exists among states
(Gustafson, 2009; Gustafson & Rice, 2010). For example, the 2000-2009 mean LEO
population of North Carolina (20,792) was 6 percent less than Ohio (22,117). However,
0
50000000
100000000
150000000
200000000
250000000
300000000
350000000
0 5000
10000 15000 20000 25000 30000 35000 40000 45000 50000
1995
19
96
1997
19
98
1999
20
00
2001
20
02
2003
20
04
2005
20
06
2007
20
08
2009
20
10
Po
pula
tion
Dea
ths
Population Deaths
5
between 2000 and 2009 (inclusive), North Carolina had 79 percent more LEO traffic
fatalities than Ohio (25 versus 14, respectively). A similar contrast exists in the same
comparison between Pennsylvania (mean 2000-2009 LEO population: 23,488; 2000-
2009 traffic fatalities: 15) and Georgia (mean 2000-2009 LEO population: 21,683; 2000-
2009 traffic fatalities: 27) where Georgia had more than 80 percent more LEO traffic
fatalities despite having 8 percent less LEO population. The overarching LEO traffic
fatality problem and associated state variability as illustrated in these examples and
Figure I.4 provide the genesis for this study.
Figure I.4 Mean LEO Population and Total LEO Traffic Fatalities 2000-2009.
Because of this variance and the potential to effect broad change, I adopt a state-
level perspective in this study and argue for state-level policy/practice/regulatory/legal
interventions. This state-level approach is detailed later in this chapter and built upon
throughout the study.
0
5
10
15
20
25
30
0
5000
10000
15000
20000
25000
NC OH PA GA
LE
O T
raffi
c Fa
talit
ies
LE
O P
opul
atio
n
Average Population Total Traffic Fatalities
6
Author Background—Expertise and Disclosure
I have worked in law enforcement for more than 20 years and spent a great deal of
time driving police cars before I started studying police driving. I have had the
opportunity to talk with people throughout the nation and around the world formally and
informally about the job of LEOs. As a result, I have an educated and informed—but
also inherently biased—perspective. While this study reflects my most objective efforts,
readers should bear in mind my status as an industry insider. Besides my
professional/vocational ties to the research topic, I also have an interest in the problem
and potential solutions as a primary concern for my current employment and role in the
law enforcement industry.
Employment History and Issue Insight
Beginning as an academy cadet in 1992, I worked as a reserve police officer in
one department and then progressed as a police officer, detective, school resource
officer, sergeant, lieutenant, executive officer, and acting chief of police in another
department through 2005. In 2005, I went to work for California Department of Justice,
Commission on Peace Officer Standards and Training (CalPOST) as a Senior Law
Enforcement Consultant (a high-level program-manager responsible for
developing/managing standards and training programs).
I have worked on state-level standards and training issues related to vehicle
operations, training, and policy as part of my portfolio my entire tenure at CalPOST.
From 2007 to 2011, I was primarily assigned to conduct and oversee research on on-duty
LEO traffic deaths. In 2009, enhancements to driver training curriculum for basic law
enforcement academies in California were adopted based on my recommendations. In
7
2012, I was promoted to Bureau Chief of Standards and Evaluation and am currently
responsible for LEO (and public safety dispatcher) standards, evaluation, and research for
the State of California. The substance of this dissertation is encapsulated in my work for
CalPOST and my findings and recommendations will directly impact future standards
and training for LEOs in California. Indirectly, these findings will likely have
significance for LEOs nationwide as California has arguably been the leader in law
enforcement standards since 1959 when it established the first peace officer standards and
training (POST) commission in the US (Conser, Paynich, & Gingerich, 2011).
Personal Experience with Driving and Luck
Like most LEOs, I had thousands of emergency responses during my 10-plus
years of uniformed service. In many ways, I was the officer, supervisor, and manager
that exemplified the problem I describe in this chapter. As an officer I engaged in several
high-speed pursuits with a low benefit to high risk ratio and I frequently drove too fast to
get to calls where an extra 30 seconds or minute did not matter. As a supervisor and
manager I gave tacit approval to officers who drove like I used to drive. I never had a
collision, but I had dozens of close calls. At the time, I felt I owed my success to skill—
the pride and confidence of a young man who learned to drive in big cars on curving
country roads. In retrospect I own the shame of dumb, undeserved luck. My hope is that
this research adds substantially to the evidence that grounds policy and regulatory change
so that future generations of officers have a better foundation than luck from which to
operate. In making this statement I recognize that many excellent regulations and
policies exist across the nation. Even so, these unspecified examples of excellence are
the exception rather than the rule.
8
Dying at Work
Many thousands of people die every day in the United States (Hoyert, 2012).
Death is universal. It often occurs naturally and may even be expected in cases of old age
or prolonged illness. In other instances death is accidental or homicidal—frequently
unforeseen or unavoidable. Death on the job—in some workplace setting—typically falls
into this latter category. Few people go to work facing a likely or known threat of death.
There are exceptions. Many military personnel in time of war, for example, might
reasonably know they face high mortal potential. Nevertheless, with the Bureau of Labor
Statistics (BLS) estimating the 2010 US workforce at 139,064,000 (BLS, 2011) and the
National Institute for Occupational Safety and Health (NIOSH) estimating between 12
and 13 workers die on the job each day (NIOSH, 2012), it is reasonable to conclude that
most people do not die at work. Still, for those workers who do die at work, the most
common cause is what the BLS terms highway incidents—for 2010: 968 or 21.3 percent
of the total 4,547 deaths (BLS, 2011). In plain language this means that about 1 in 5
workers killed on the job in 2010 died in a traffic-related incident like being hit by a car
or involved in a motor-vehicle collision. While these data include individuals in jobs that
have little or nothing to do with driving or traffic (e.g., a bank teller, librarian, or dentist)
and others who primarily drive and/or deal with traffic (e.g., a bus driver, highway
worker, or valet), this aggregate statistic provides a sense of the prevalence and risk of
traffic-related occupational fatalities in the general working population. Figure I.5
illustrates this prevalence and risk. On the left, the chart shows the population of US
workers and a visually-negligible razor-thin slice that represents the proportion killed in
9
2010. On the right, the chart is an exploded view of the workers who died and shows the
proportion of highway deaths to all other on-the-job deaths.
Figure I.5 US Workers and Distribution of On-the-Job Deaths 2010.
By comparison, the risk of on-the-job death for LEOs—especially traffic-related
death—is substantially greater than it is for the average American worker. Figure I.6
illustrates this prevalence and risk. On the left, the chart shows the population of US
LEOs and another visually-negligible slice (nearly 600% thicker in reality) that
represents the proportion killed in 2010. On the right, the chart is an exploded view of
the LEOs who died and shows the proportion of highway deaths to all other on the job
deaths.
U.S. Workers Total Job Deaths
Other Deaths
Highway Deaths
10
Figure I.6 US LEOs and Distribution of On-the-Job Deaths 2010.
Real People in Real Places
In reality, much more is at stake than statistical significance. Real people with
families, friends, and communities are behind the charts and graphs that summarize
industry and policy differences. Brad Moody—pictured with his canine partner Rico in
Figure I.7—is an example of one officer killed in a traffic collision.
Richmond Police Officer Brad Moody, 29 years of age and an 8 year employee
with the Richmond Police Department, was involved in a single patrol-car
accident on October 4, 2008 shortly before 8 a.m. on Marina Bay Parkway just
north of Regatta Blvd. in the City of Richmond. Officer Moody’s vehicle struck a
utility pole in the median while responding to a call that involved a felony assault
with injuries. It had been raining a short time before the call and roads were slick.
…
Officer Moody sustained catastrophic brain injury and he was placed on
life support. On October 5, 2008 a determination was made that Officer Moody’s
Law Enforcement
Officers Total Job Deaths
Other Deaths
Highway Deaths
11
brain functioning was not compatible with life. His preference was to be an
organ-donor, so his family instructed the hospital to move forward with that
process. Following the accident, hundreds of RPD employees visited the hospital
in support of Brad and to grieve with his family.
Officer Moody is survived by his wife, Susan, and his two young
daughters. Officer Moody was a K-9 officer with the department and his dog
“Rico” survived the accident and recovered from minor injuries. Officer Moody
was also a member of the RPD SWAT Team. Brad was held in the highest regard
by his peers. His work ethic was extraordinary, and his commitment to the
community of Richmond and his profession unparalleled. Needless to say, this
was a very difficult time for RPD and Brad’s many friends in the community.
(Richmond, CA – Official Website, 2012, ¶ 1-3)
Figure I.7 Richmond Police Officer Brad Moody and Rico. Photo courtesy of the Richmond Police Department
Brad’s wife Susan—a former police dispatcher—has been a champion of officer-
safety and worked with CalPOST and the International Association of Chiefs of Police
12
(IACP) to highlight the importance of seatbelt use and safe speed following Brad’s death
(CalPOST, 2010). In a 2010 interview I conducted for the IACP (2010) video Is Today
Your Day?, Susan described how Brad always wore his seatbelt when driving off-duty
with his family, but never when he was working. Figure I.8 shows Brad’s patrol car as it
came to rest following his fatal traffic collision.
Figure I.8 Officer Moody’s Patrol Car After His Fatal Collision. Photo courtesy of the Richmond Police Department.
Standing in Susan’s den in April 2010 while my videographer captured footage
showing family photographs of Brad, Susan, and their children, I reflected on my
experience as a LEO driving without a seatbelt even though I always wore it when
traveling off-duty with my wife. It was a stark reminder of how I had been lucky.
During my policing career I responded to several fatal traffic collisions. As a researcher I
have reviewed hundreds of fatal LEO collisions. But somehow, listening to Susan cry as
she described her life after Brad’s collision—the first moments after being notified, the
13
hours in the hospital before his death, the days leading up to his funeral, and then the
years since raising two children without him—put everything in a different perspective.
In light of this tragic example that is illustrative of the many people and stories
behind the statistical summaries presented above, I urge law enforcement practitioners
and scholars to examine and work to reverse this trend. Numerous questions might guide
this effort. What factors account for the increase in law enforcement collisions? What
accounts for state level differences? Which aspects of this problem are most amenable to
policy solutions? Still, many of these critical questions have been displaced by questions
that seek to assign or avoid responsibility or assess and assign the economic costs of
response.
Assessing the Problem
Recent discussions about law enforcement traffic fatality rates often have focused
first on assigning responsibility. This assignment of responsibility naturally leads to a
series of questions. For example, to what extent do LEOs have personal responsibility as
individual drivers? Officer Brad Moody was honored as a hero after his death.
Colleagues, friends, and family extolled Brad’s professionalism and dedication to his job.
But did he have a personal responsibility to drive safely—to get to his destination? He
chose not to wear his seatbelt and he died. He may have died even if he was wearing it.
Is it a personal issue?
Do law enforcement agencies have responsibilities to address and try to prevent
these tragedies as an issue of workplace health and safety? The Richmond Police
Department, for example, allowed Officer Moody to drive without a seatbelt—it was an
accepted practice. The Las Vegas Metropolitan Police Department also accepted LEOs
14
not wearing seatbelts and wantonly speeding—until three officers died in traffic
collisions in 2009. Then Sheriff Gillespie made it an organizational imperative to change
the driving culture through training, policy, and enforcement (Alqadi, 2012; Gustafson &
Cappitelli, 2010).
Are these state-level or national issues? Texas lost 67 LEOs in traffic in the
2000s while Maine, New Hampshire, North Dakota, and Rhode Island lost none
(Gustafson & Rice, 2010). Traffic deaths appear to be an issue for the State of Texas, but
perhaps not for all states. In terms of a national issue, I have already established that
traffic collisions are the leading cause of death for LEOs. On a nationwide scale, there is
no greater threat. But what can be discerned from this aggregated fact?
I contend that the issue of fatal LEO traffic collisions may best be understood as
an industry problem for the law enforcement community as a whole. Questions that
reinforce boundaries (either personal or jurisdictional) as a matter of blame do little to
solve the problem. In fact, interactions related to other public policy risks suggest that an
early emphasis on responsibility may discourage coordinated research and action
(Fischer, 2003). However, questions that view these boundaries as points of demarcation
to identify policies, practices, laws, and regulations that shed light on increased (or
decreased) LEO traffic fatalities, those questions and boundaries offer rich insights.
Since law enforcement training and operational standards are set at the state-level, cross
case (state) analysis—questions that probe differences and similarities relating to
variance in LEO traffic fatalities among states—promises greater understanding.
15
Argue and Avoid—It May Be a Problem, But Certainly Not Mine
Most people can agree on the existence of countless public policy problems from
local to global scale. Consider flood plains and levee infrastructure in New Orleans, LA
or Sacramento, CA. Think of border and immigration issues in Arizona, the state budget
deficit in California, or the drought and wildfires in Colorado and much of the Midwest.
Reflect on the political deadlocks over taxes, healthcare, and social issues in the US
Congress, the economic fragility of the Eurozone, war and political instability throughout
much of the Middle East, and ongoing concerns and debates about global climate change
and nuclear proliferation/security. Many people agree that these are significant public
policy issues that merit serious attention, but that they merit the attention of someone
else. The businesswoman in New York will not likely feel as though she should work on
the levee issues in New Orleans or Sacramento. The legislator in Mississippi will not
likely be compelled to take action on the wildfires in Colorado or the budget in
California. Even most well-intentioned, civically-minded university professors or city
police chiefs will not feel a personal imperative to address global economic, security, or
environmental issues that might very well have a tangible impact on them long term.
Everyone cannot take action on every important issue, and the necessary triage to
determine which issues to take action on plays out in different ways when applied to the
problem of law enforcement traffic fatalities.
Individual LEOs are frequently unaware of the problem or, if they are aware, they
ignore or fail to see the risk. I was an example of this. I was unaware that the most
dangerous aspect of my job was driving the patrol car. While I cannot generalize my
personal experience to the population of LEOs, I can argue that it is difficult to change a
16
system of hundreds of thousands of individuals at the individual-level. Similar dynamics
exist at the agency-level. There are approximately 18,000 state and local law
enforcement agencies in the US (Reaves, 2011). Most of these agencies have never had a
LEO die in a traffic collision. Awareness and an imperative to act are absent for most
agencies. The role of a law enforcement agency is primarily to prevent and investigate
crime—to provide public safety. There are risks involved in this mission and so in some
ways it is natural for LEOs and their agency-level leaders to accept risks (like traffic
collisions) as part of the job.
From the prior examples—Texas with the most LEO traffic deaths and Maine,
New Hampshire, North Dakota, and Rhode Island with the fewest (2000-2009)—logic
for the state-level perspective begins to emerge. With many laws, standards, training,
and regulations for LEOs established at the state-level, states have a clear interest in the
problem. What about the nation as a whole?
Weigh the Costs
I once had occasion to talk about this issue with a number of policymakers in
Washington, DC. After presenting trend data similar to what has been detailed here, a
high-ranking official suggested that with tens of thousands of Americans dying in traffic
each year, 50-80 cops was not a significant enough number to merit federal government
focus and resources. The issue became one of time, money, and priority, as opposed to
just about saving human lives.
As I spoke with the federal policymakers it became clear that a rationale choice
argument was being poorly constructed based on flawed assumptions of transitivity
(Shepsle & Bonchek, 1997). The implication was that it would be (a) unjust, or (b) a
17
poor cost-benefit tradeoff to give special consideration to a small subset of the overall
population that was experiencing a small percentage of the overall problem. This logic is
flawed for several reasons. First, all traffic deaths do not carry the same costs. Second,
risk exposure varies for everyone. Third, the costs to address the problem among LEOs
are unrelated to the costs to address the problem among the general public. To be fair,
the traffic death rate (population/deaths*100,000) among LEOs is typically less than the
traffic death rate among the general population. In 2010, the LEO rate was 8.93 and the
general population rate was 11.28. However, it is important to remember that the LEO
rate only reflects exposure on the job. The LEOs also have the general population risk
exposure when they are not working. Finally, as an occupational group, as described
earlier, the LEO risk is several hundred percent greater than for other workers overall—
the general workforce traffic death rate for 2010 was .70. In plain language, this means
that in the period of one year, the chances are that 9 out of 100,000 LEOs will be killed in
traffic compared to just 1 out of 100,000 workers. In my estimation, these facts support
taking action on the LEO traffic death problem.
Significance of the Problem
For more than a decade, traffic collisions have been the leading cause of death for
LEOs in the US (FBI, 2008b; NLEOMF, 2009). The federal government is aware of this
trend (FBI, 2008b; NHTSA, 2010b) and has identified it as one of the primary
occupational hazards for LEOs (NIOSH, 2009), even as some federal policymakers have
been reticent to take action. Previous research has indicated that law enforcement traffic
deaths pose tremendous financial costs to local, state, and federal governments with each
officer fatality costing approximately $1.7 million on average (Gustafson & Cappitelli,
18
2010). The cost is clearly separate from the immeasurable social and emotional toll
associated with loss of human life. This situation alone creates a moral imperative to
address the problem. The following sections detail how I interpret this moral imperative,
why I think it is worth my time (and yours), and what I hope to accomplish.
Engaged Scholarship—Research for Everyone’s Sake
This study was designed and implemented as a form of engaged scholarship (Van
de Ven, 2007). According to Van de Ven, “[e]ngaged scholarship is a participative form
of research for obtaining the different perspectives of key stakeholders … in producing
knowledge about complex problems” (p. 265). He sees it as a better way to do research
to solve problems—to bridge the theory-practice gap and not just create knowledge, but
also effect change and improvement. Involving insiders and outsiders as Van de Ven
advocates provides a more robust research model, but I think there is more that matters
than good design.
The Moral Imperative
As previously described, I represented a prime example of the problem at issue in
this study earlier in my career—I was simply fortunate not to die. Now, in my capacity
as Chief of Standards, Evaluation, and Research for CalPOST, I have a responsibility to
the LEOs of California to establish responsible and meaningful standards for their
training and to support and recommend principled regulation and law to support their
occupational health and safety. Dozens of them have died since I first began to work on
this issue in 2007, so I carry some significant concept of personal responsibility—not for
the problem—but for the intervention, if not a solution. I think this gives me a different
stake in the research and provides a powerful impetus. Adler and Hansen (2012) have
19
given this topic a great deal of consideration and have summarized the issue most
poignantly:
Researching questions that matter demands passionate conviction. Whether
recognized as such or not, such conviction, combined with profound compassion,
defines true scholarship. Daring to care requires courage—the courage to speak
out and to act. Courage transforms convictions and compassion into action.
Thus, by its very nature, daring to care calls into question the traditional role of
rigid scientific objectivity and invites advocacy to play a vital role within our
scholarly tradition. (p. 128)
And so I choose to be an advocate in addition to a researcher.
Scholarly and Practical Utility
I have worked directly with practitioners in the field, and this study contributes to
a better scholarly understanding of law enforcement regulation and standards setting at
the state level, as well as practitioner knowledge about the effects of various policies and
practices on officer health and safety. By expanding the macro-level, industry-wide
understanding of policy and practice interventions relative to fatal LEO traffic collisions,
this research fills a void in the literature identified in the National Occupational Research
Agenda (NORA) for Public Safety (NIOSH, 2009).
The work also furthers research on restructuring of policing and
professionalization efforts as previous scholarship has called for national-level
assessments and a nexus between theory and practice (Bayley & Shearing, 2000;
McClellan & Gustafson, 2012; Scott, 2009). Research in decades past identified widely
varying law enforcement standards for selection, training, and best practices (e.g.,
20
minimum education required, driver training curriculum, mandatory seatbelt use, or
vehicular pursuit protocols) across states and reported this as problematic (Carte, 1969;
Janeksela, 1981; Vogel & Gamache, 1981). This study contributes to the law
enforcement standards literature as it investigates (a) the extent to which significant
variability still exists among states and (b) to what degree the identified variability
contributes to LEO traffic fatalities.
Partially as a result of the President’s Commission on Law Enforcement and
Administration of Justice (1967), law enforcement standards are generally established at
the state level through POST agencies (Conser, Paynich, & Gingerich, 2011). While the
President’s commission called for a higher level of standardization nationally, little exists
(McClellan & Gustafson, 2012). It is worth noting that the Commission on
Accreditation for Law Enforcement Agencies (CALEA) has created a framework for
national policy standards for law enforcement agencies. However, CALEA is an
independent, nongovernmental organization and its standards are voluntary. A small
percentage of law enforcement agencies participate. Moreover, CALEA allows for
significant variability by allowing participating agencies to meet a standard by
developing a local policy that addresses the standard. In terms of driver training and
operations, CALEA does not provide enough guidance to constitute national standards.
In terms of state variability, in some jurisdictions in the US a person can become a
LEO with as few as 320 hours of training, while in other places more than 1,100 hours of
training is required (Magers & Klein, 2002). Some states require a minimum of an
associate’s degree, while most require a high school diploma or equivalent (International
Association of Directors of Law Enforcement Standards and Training (IADLEST),
21
2005). Specific to law enforcement driver training, some states require none and others
require as much as 60 hours of driver training (IADLEST, 2005). Law enforcement
regulations and standards significantly impact how individual communities are policed in
the US (Mayo, 2006). Assessing the impact of these regulations and standards and the
practices that have come to be associated with them promises to enhance the democratic
process and contribute to law enforcement transparency and accountability (Mayo, 2006).
To that end, descriptive and qualitative assessments of the impacts of standards and
regulations across states expand the existing literature.
Goals, Objectives, and Contributions to Literature
Generally, the goals of this study are to (a) increase understanding of the
characteristics of fatal law enforcement traffic collisions among states, and (b) identify
state-level variables/factors that impact these fatal collisions. This information will be
useful in modifying policies and practices to reduce fatalities.
The specific objectives are to (a) describe the incidence rate and variability for
fatal law enforcement traffic collisions across states, (b) identify associations between
regulations, training, and operational practices and fatal collisions, (c) identify
interactions between these factors that work together to increase or decrease fatal
collisions, and (d) identify potential policy/practice interventions to reduce the
occurrence of fatal law enforcement traffic collisions. Finally, this study aims to produce
information that is hard for law enforcement policymakers to ignore.
Policy Implications and Broad Impacts
This research describes the occurrence of fatal law enforcement traffic collisions
and identifies risks and benefits of laws/regulations and policies/practices across states.
22
In so doing, the findings detailed here provide state and local government actors with
empirical evidence to inform policy and practice. This data-driven model allows for
state, regional, and local customization and has the potential to save lives and reduce
costs. The research benefits individual LEOs, their families, agencies, states, and the law
enforcement industry as a whole.
Intended Audience and Target of This Research
The intended audience for this study is law enforcement policymakers—police
chiefs, sheriffs, commissioners/colonels of state patrols, and POST directors. The target
is uniformed officers—those sworn personnel who drive in emergency response
situations. This research is intended to inform policy and practice related to standards,
training, and operations. While empirical methods and scholarly arguments are featured
throughout this study, the expectation is that they are specified and framed in a way that
is of practical value and import for the practitioner.
Outline for the Study
Chapter 2 reviews highpoints of the relevant history of law enforcement vehicle
operations and explores the scholarly literature related to traffic collisions generally and
law enforcement collisions specifically. It outlines how the issue/problem has come to be
understood to include policy and technological aspects. It then details the gap in the
existing literature wherein this work is positioned. Chapter 3 then lays out the basis for
this research. It explains how the study is conducted in terms of methodology and
specifies the data used, challenges encountered, and remedies employed. It references
prior research that successfully used methods adapted for this study and elucidates the
questions that this study aims to answer/explore/address.
23
Chapter 4 provides a statistical and descriptive analysis of LEO traffic fatalities
by state for the period 1995 to 2009. No detailed analysis of this type has been done
before and the expectation is that the results provide a useful factual basis both for the
premise of this research—the existence of significant state-level differences in law
enforcement traffic fatalities—as well as for practical policy decisions in policing.
Chapter 5 then documents a quantitative and qualitative analysis of the state-level
differences in LEO traffic fatalities detailed in Chapter 4. This chapter is the heart of this
study as it provides the empirical evidence. It has two components. First, it uses a
statistical model to analyze the effects of measureable state-level differences in LEO
traffic fatality rates. Second, it uses qualitative survey data from national law
enforcement policymakers to contextualize, temper, and clarify observed differences.
Next, Chapter 6 offers discussion and policy implications from the analysis and
findings detailed in Chapter 5. This is an interpretive chapter. It uses a policy lense to
make sense of the quantitative and qualitative evidence/analysis and posits state-level
changes to reduce LEO traffic fatalities.
Finally, Chapter 7 provides a framework/process model for implementation of
policy changes and an agenda for future research. This chapter suggests how law
enforcement policymakers can effect meaningful policy changes to reduce LEO traffic
fatalities. It specifies what steps to take and how to go about them. It explicates needs
and opportunities and issues a challenge to leaders to do right by the profession.
24
CHAPTER II
LITERATURE REVIEW
This chapter develops the scholarly basis for the current study through an
analytical review of the literature related to traffic collisions and law enforcement officer
(LEO) traffic collisions specifically. Prefacing this review is the relevant background to
this study, which begins sometime near the turn of the 20th Century when the automobile
was introduced and law enforcement personnel first encountered traffic collisions. But
even this point is somewhat unclear in its specifics because of historical data problems.
There is agreement among three data sources that Officer Alonzo B. Bishop of the
Baltimore City Maryland Police Department died on Tuesday, August 29, 1899, when his
patrol wagon was involved in a collision variously reported as being with “a car”
(Maryland Police & Correctional Fallen Officers Memorial, 2012), “another vehicle”
(NLEOMF, 2012a) or “an automobile” (Officer Down Memorial Page (ODMP), 2012a).
This is likely the first LEO traffic-related fatality in the US
Thereafter, database research reveals a number of discrepancies in the facts
surrounding many officer fatalities that may have been traffic related for more than a
decade after Officer Bishop died. Agreement exists, however, that Policeman James P.
Wylie of the Los Angeles, California, Police Department died on Monday, November 27,
1911, in what was likely the first clear case of an on-duty LEO fatality from a car versus
car traffic collision (Los Angeles Police Department, 2012; NLEOMF, 2012b; ODMP,
2012b).
Throughout this study there are challenges with data, including conflicting counts,
different definitions, lack of law enforcement identifiers, and suspected data input errors.
25
These aspects make it difficult to (a) study the problem of law enforcement collisions
precisely and (b) compare across different reporting institutions (e.g., different states).
Specific data issues are discussed in detail in Chapter 3.
Reporting inconsistencies aside, the aforementioned cases point to the genesis of
the LEO traffic fatality problem addressed by this study. In this respect, from an
epidemiological perspective, it is useful to consider that less than 100 years passed from
the first reported incidence of a LEO traffic-related fatality (1899), to the point at which I
argue the LEO traffic fatality epidemic began (1997). From this period until the present
day there have been many directly and indirectly relevant works of scholarship that
inform this study. These literatures are reviewed in the sections that follow.
Historical Context
For context, it is worth noting that there was no substantive national conversation
or broadly identifiable concern about highway safety prior to the 1960s. Only then did
awareness of a problem begin in earnest. While it is true that the National Transportation
Safety Board (NTSB) came into being in 1926, that division of the federal government
was, at that time, solely focused on air transportation (NTSB, n.d.). During the 1960s,
however, the landscape changed.
The primary impetus for relevant functional and legislative transformation was
likely the change in national awareness and mood regarding highway fatalities. The
American love of the automobile apparently shifted from blind infatuation toward more
thoughtful and informed evaluation. A significant contributor to this shift was Ralph
Nader, an attorney and consumer protection activist/advocate in the early 1960s. Nader
(1965) may have had the broadest impact through his now classic book Unsafe at Any
26
Speed: The Designed-in Dangers of the American Automobile in which he detailed
significant oversights and missed opportunities in safety engineering considerations
among American car manufacturers. In his review of the book, Cutcliffe (1966) noted
that "[w]hen the Senate Commerce Committee opened its hearings on the Highway
Safety Act of 1966, Ralph Nader as author of Unsafe At Any Speed was among the first
witnesses" (p. 445).
Following the Senate hearings, Congress passed the Highway Safety Act of 1966,
which (among many changes and initiatives) mandated seatbelts in automobiles and
created the National Highway Safety Bureau, which later became NHTSA in 1970
(NHTSA, 2006). The framing of the issues and summation of the topic was perhaps most
succinct and poignantly delivered by President Johnson in his signing remarks upon
authorizing the act when he wrote on September 9, 1966:
Over the Labor Day weekend 29 American servicemen died in Vietnam. During
the same Labor Day weekend, 614 Americans died on our highways in
automobile accidents.
Twenty on the battlefield.
Six hundred and fourteen on the highways.
In this century, more than one and a half million of our fellow citizens
have died on our streets and highways: three times as many Americans as we have
lost in all our wars.
Every 11 minutes a citizen is killed on the road.
Every day 9,000 are killed or injured—nine thousand.
Last year 50,000 were killed.
27
And the tragic totals have mounted every year.
It makes auto accidents the biggest cause of death and injury among
Americans under 35.
And if our accident rate continues, one out of every two Americans can
look forward to being injured by a car.
This is not a new problem. Ten years ago in the Senate I told my
colleagues that "the deadly toll of highway accidents demanded action [sic]. And
that this was a responsibility Congress must face. Now, finally, we are facing it.
(NHTSA, 1985, p. 31).
This rhetoric helped catapult the issue of traffic fatalities to the forefront of public
conversation about risks. Contrasting a heavy weekend death toll in Vietnam with a
homeland traffic death toll some 20 times greater captured significant attention. That
same month (September 1966), the National Research Council released Accidental Death
and Disability: The Neglected Disease of Modern Society. In the introduction it noted
that of 107,000 accidental fatalities in 1965, 49,000 resulted from traffic collisions.
Further, the work asserted that, “Public apathy to the mounting toll from accidents must
be transformed into an action program under strong leadership" (p. 5).
These and countless other factors led to changes in many aspects of safety. From
that point forward, highway safety took a new direction in America. Even so, traffic
fatalities continued to rise for several years after Johnson signed the Highway Safety
Act—peaking at 54,589 deaths in 1972 (NHTSA, 1997). Interestingly, while the
awareness of the traffic deaths problem and its peak occurred among the general public in
the mid-1960s and early 1970s, respectively, the timeframes for the LEO-specific
28
epidemic appear to have come much later. Writing for Police Chief Magazine in 2004,
NHTSA Deputy Administrator Otis Cox extolled the highway safety advances of
NHTSA and only incidentally noted that “some of the other news is not all that uplifting.
Traffic crashes are the leading cause of death in the line of duty for law enforcement
officers. In 2003 ... 75 officers died in motor vehicle crashes” (Cox, 2004, ¶ 6).
Perspectives on the Problem
How the problem is constructed—both by industry practitioners/policymakers and
researchers—shapes how it is addressed. As explained in Chapter 1, prior to the 1990s,
accidents in general—and traffic deaths specifically—were the smallest fraction of all
line-of-duty LEO deaths. As such, they were not a priority in the greater scheme of
occupational hazards and there is no documented evidence of a problem realization.
As illustrated by Cox (2004), there was some clear awareness of the loss of
officers in traffic in the early 2000s. Interestingly, in that same work, Cox also made the
point that traffic deaths are not inevitable and accidents (i.e., collisions) can be prevented.
While that outlook has become common in the greater traffic safety community, it has
failed to resonate with many aspects of the law enforcement community. High profile
law enforcement leaders like CalPOST Executive Director Paul Cappitelli and Lexipol
President Gordon Graham have been working to dispel the myth that LEO traffic deaths
are largely unavoidable or a simple industry cost or risk as opposed to a problem to be
fixed (Gustafson & Cappitelli, 2010). Finally, in 2009, NIOSH prioritized scholarly
investigation of LEO traffic injuries and deaths in the National Occupational Research
Agenda (NIOSH, 2009) and, somewhat shockingly, prior to 2002 there was no way to
29
assuredly know if the driver of a vehicle involved in a fatal traffic collision was a LEO
when reviewing national traffic death data from NHTSA (NHTSA, 2008).
In this sense then, the matter of LEO traffic deaths can be viewed retrospectively
as developing into a recognized problem through a series of evolving perspectives. I
argue that (a) in the early days of the automobile, traffic deaths were not recognized as a
problem, (b) after many decades, traffic deaths were recognized as a problem in the
1960s, and finally (c) in the early 2000s, traffic deaths were recognized as a problem for
LEOs, even more so than for the general population. This conceptualization of LEO
traffic deaths might be consistent with the development of perspectives on other risks.
Take cigarette smoking, for example. First it was not recognized as a problem. Then it
was recognized as a problem (i.e., it can lead to cancer). Later it was recognized that
certain occupational groups (e.g., flight attendants) were at greater risk than the general
population. Of course in this example, the greater risk is most easily explained through
increased exposure (which does not explain the LEO traffic fatality problem). Even so,
the example illustrates how some observable trends are viewed differently over time as
they are first recognized and then understood more deeply.
Previous Scholarship
The published work related to this study is vast. Thousands of articles and books
have been written about topics covering traffic safety; vehicle performance; driving and
emergency driving; traffic collisions; law enforcement training, policy, and practice; and
human performance including skills and abilities, as well as more biological determinants
like reaction time and fatigue. Beyond these clearly relevant subjects there are subsets of
meta-disciplines in the social, cognitive, and natural sciences that are clearly informative.
30
For example, consider aspects of culture, attitude, design, cost, and all kinds of
regulation—of roads, cars, and drivers. It quickly becomes apparent that this problem—
LEO traffic fatalities—can be informed and understood from sociological, psychological,
engineering, economic, and political science perspectives. An exhaustive review of the
literature is virtually impossible. It is, however, possible to review key intersections
among relevant literature and establish a base reference point that is substantive.
Relevant Boundaries
The goal in this chapter is to establish meaningful boundaries and identify
sufficient relevant and informative literature to support the study at hand. The research
gap previewed in Chapter 1 and described in detail in this chapter is clear. The need then
is to examine the work that leads up to that gap. There are two primary factors I use to
accomplish this task.
First, I look at the contribution a work makes toward understanding this problem
as a state-level phenomenon. This is more complicated than it may sound. In studying
law enforcement traffic collisions I distinguish among four levels: (a) individual, (b)
agency, (c) state, and (d) national/international.
The individual level is a factor which is unique to the person or event. An
example of this might be driver alertness. Each person may be more or less alert and
alertness may not be consistent across time and place. Another example might be debris
in the road. These are specific occurrences that are not universal or generalizable beyond
the individual person or event.
At the agency level are factors that are common or universal to everyone working
at that organization, but not universal or generalizable to all agencies. A specific law
31
enforcement agency like the California Highway Patrol might have a culture, mission,
rules, equipment, or procedures that differentiate it from other agencies. Consequently,
while sharing many basic commonalities each organization will have unique features.
Next is the state-level—the focus of this study. A state-level factor is one that is
common/uniform statewide, but varies from state to state. Training is a good example.
Some states require LEOs to have a specific course of driver training that other states do
not. The amount of training is a state-level factor. There can also be agency-level
training or individual level training. In this respect, training could be a factor at multiple
levels. In this case, it is probably clear enough what training is required at each level.
Other factors are more complicated.
For example, accept for a moment that highway speed is an important factor in
whether or not a vehicle collision is fatal—it is, and this will be discussed at length.
When considering highway speed, it is clear that individual collisions happen at different
speeds and one might expect that this is an individual-level phenomenon. While this is
true, there are strong state-level determinants of highway speed like enforcement and
regulatory law (i.e., speed limits). In Montana, for example, there was no maximum
speed limit for a long period of time and even now, with a maximum speed limit,
enforcement of that limit is arguably not as aggressive as it is in some other states.
Weather also is an example of the complex nature of traffic accidents at many levels.
Accept for a moment that weather, like speed, is an important factor in the occurrence of
fatal traffic collisions. Again, the argument can be made that weather is unique in any
given collision circumstance—it is an individual-level factor. Then again, it is clear that
weather does vary by state. Clearly, there is a great deal more inclement weather that
32
impacts driving (e.g., rain, snow, freezing temperatures) in the State of Alaska than in the
State of Arizona. So perhaps weather, like speed, is a state-level phenomenon that
deserves consideration.
My second primary factor is the contribution a work makes toward understanding
this problem as a (state-level) public affairs issue. While highway speed continues to
meet this second standard since maximum speed limits are set by states, weather does
not. No amount of lobbying, legislation, or rulemaking can change the rainfall in
Washington or the sunshine in Florida. Unless there is a state-level policy or law that
impacts when or how LEOs perform in certain weather conditions (e.g., a law that
prohibits LEOs from driving when it rains), previous literature regarding weather will be
less informative.
Problems and Solutions—Prior Understandings and Responses
With these parsing factors in mind, it is next useful to consider how scholars have
understood and tried to address this issue. The problem, for example, may be understood
as a technical or engineering problem with the automobile itself, as Nader (1965)
suggested. Alternatively, it may be understood as a cognitive training problem—drivers
just lack skill. These different conceptions of the problem naturally lead to different
solutions. An engineering problem requires an engineered solution. A training problem
requires a training solution. Bolman and Deal (1997) referred to this approach as framing
and noted that “[f]rames filter out some things while allowing others to pass through
easily. Frames help us order experience and decide what to do” (p. 12). While these
conceptions present necessarily artificial boundaries and divisions, they provide a useful
33
heuristic for understanding how scholars have approached traffic collisions in both the
general public and law enforcement specific settings.
As described in more detail later, this research assumes that LEO traffic collisions
are a subset of the universe of traffic collisions and have many of the same properties.
While there are undeniably factors that only apply to the LEO subset (e.g., driving with
emergency lights and sirens activated), the precursors to a general population collision
(e.g., high driving speeds) are also precursors to a LEO collision.
LEO Fatalities as a Technical Problem. Anyone who has seen an array of car
commercials over some number of years—especially a Volvo commercial—has likely
seen traffic safety posed as a technical problem. The implication of some of these
commercials is that cars simply need to be better engineered; then you and yours will be
safe on the road. In many respects this social construction may have evolved directly
from Nader’s (1965) indictment of the auto industry. In the span of a modest lifetime,
technical fixes to vehicle safety have evolved from the requirement of a lap (seat) belt to
anti-lock brakes to air bags to recent models that will apply the brakes, monitor traffic in
the blind spots, or make minor steering adjustments automatically to assist the
presumably unaware driver (Müller & Stajic, 2011). This is a well-documented history
(see MacGregor, 2009; Wetmore, 2009). Whitelegg (1983), following Nader, detailed a
broad complaint about roadway safety. Beyond technical issues of vehicular engineering,
he further described a systemic problem to include not just the industry, but also
government and consumers. He argued that society’s reliance on cars was central to the
problem. Whitelegg viewed essentially everything as flawed—from land use to public
transportation to the layout of highways.
34
In terms of a technical problem for law enforcement, there has been a growing
dialogue throughout the industry that police cars are really just vehicles adapted to
policing purposes. This may seem like an overtly silly observation. But consider
garbage trucks, fire engines, limousines, parking enforcement vehicles (i.e., those small,
three-wheeled vehicles common in most cities), school buses, and delivery vans (e.g.,
FedEx-type vehicles). Each of these vehicles is technically engineered to an occupational
need/task. Clearly, they were designed with a specific function in mind. This is not the
case with the American police car. While car companies may market a police package, it
typically includes a more robust electrical system (to support all the lights and
equipment), heavy-duty brakes, and speed-rated tires. But these are options available to
anyone and engineered to a somewhat generic need. The same vehicle used by your local
police or sheriff’s department also is marketed to taxi companies, rental car companies,
and private citizens. So the discussion began between a frustrated law enforcement
official and an automotive executive that an engineered police vehicle was needed
(Carbon Motors, 2011).
As a result, Carbon Motors was born in 2003. The prototype vehicle was debuted
at the IACP Convention in San Diego, CA in 2008 (Jonsson, 2008). The car has features
that include integrated control panels, specially designed seats that allow space for an
officer’s gun belt, and a heads-up display that allows the officer to see critical
information while keeping their eyes on the roadway (Carbon Motors, 2011). These are
clearly technical fixes/responses to problems that have been identified as having some
nexus to LEO traffic collisions.
35
LEO Fatalities as a Human Performance Problem. Vila (2000, 2006) has
reported on the effects of fatigue as it contributes to law enforcement accidents (errors in
general) and traffic collisions specifically. This has been his primary area of research for
more than a decade and as a former LEO, Vila has uncommon insights. Not surprisingly,
Vila found that the more fatigued officers are, the more likely that they will make
judgment errors. While this finding may not be surprising at an intuitive level, consider
that there are no laws or standards that regulate how much fatigue is too much in law
enforcement.
This situation is best portrayed through a hypothetical example. Shift work and
overtime are endemic to law enforcement. LEOs working 8, 10, and 12-hour shifts are
common in agencies throughout the nation. Also common is for LEOs to make arrests or
write reports at the end of their shift. This creates overtime—hours beyond the 8, 10, or
12 just worked. An officer, for example, might be scheduled to work a 10-hour shift
from 2:00 PM to Midnight. At 11:30 PM, the officer encounters a vehicle on the road
driving erratically. The officer pulls this vehicle over, determines the driver to be under
the influence of alcohol, and at 11:45 PM arrests the driver. A number of procedures
follow: the driver is taken to the hospital for a blood draw; the vehicle is inventoried,
towed, and stored; the driver is booked at the county jail, which may be some 30-minutes
travel time from where the arrest occurred; the officer files paperwork at the jail to justify
the incarceration; the officer returns to the station to write the arrest report; evidence has
to be cataloged and booked-in; equipment needs to be returned; the patrol car has to be
refueled; and then, sometime around 3:00 AM the officer goes off-duty. The 10-hour
shift then becomes13 hours on duty. Presumably, as is common in many parts of the
36
nation, the officer might have had a commute before and after work. For this scenario,
maybe the hypothetical officer drives an hour each way to and from work. Now the
hypothetical LEO has had a 15-hour workday. The officer probably has personal cares
and such before and after work—perhaps a spouse and children to spend a little time with
before or after work, and time to unwind. In all, this officer may be up for 18 or 20
hours. The officer may have to do this several times a week. Many officers have
required court appearances in the middle of the morning when they would normally be
sleeping. This scenario is not far-fetched or uncommon. Imagine if, instead of a 10-hour
shift with a driving-under-the-influence arrest at the end, the officer had worked a 12-
hour shift and had a robbery or domestic violence incident at the end—they could be
awake and at work even longer. Shifts like that can lead to serious sleep deficits. And,
unfortunately, there are no standardized rules to prevent this from happening.
Vila (2000, 2006, 2009) also studied sleep deficits and found that the disruptions
to Circadian rhythms endemic to law enforcement shift work have a lasting and
increasing toll on performance and that overall shift length (e.g., 8, 10, or 12 hours) is a
contributor to fatigue. He noted this physiological response has been studied relatively
well in other industries like aviation and transportation, and more recently among
emergency room doctors (2009). The human body is limited in the number of hours it
can be deprived of sleep and safely complete complex tasks (2000, 2006, 2009).
Consequently, truck drivers and pilots have limits on the number of hours they can work
and are required to keep logs to show when they drive or fly and when they sleep. But no
such regulation exits for LEOs (or emergency room doctors). Simply put, “Whom do
you want out there doing the job? The person who ‘grabbed a few hours of sleep,’ or the
37
person who made it a priority to come to work rested and ready to handle life and death
situations?” (Vila & Gustafson, 2011, p. 12).
James and Vila (2012) also studied distraction as it relates to law enforcement
traffic collisions. As technological innovations have advanced, there have been more
demands on a LEO’s attention in the patrol vehicle. In the early days of patrol cars the
vehicle was just that—a conveyance. Then, progressively over time, technologies were
added—lights, sirens, radio(s), scanner(s), weapons (e.g., shotgun, patrol rifle, baton),
safety cage, radar, stolen vehicle locators, video cameras, computers, and more.
According to James and Vila, “‘Multitasking’ is a myth, the brain must shift from one
focus of attention to another, and back. Switching requires cognitive resources that tend
to diminish as an officer progresses through the work shift” (p. 17). Even if officers
manage to avoid focusing their attention on these technologies (i.e., actively use them),
they still have to look through/around them in order to view the roadway and check their
blind-spots as they drive.
These research findings are representative of the intersection of the technological
and human performance aspects of the problem. Where Carbon Motors is attempting to
design away the problems of distractions, James and Vila are researching interventions to
help officers cope with distractions. This contrast is one of the best examples of two
different ways of viewing and responding to the same set of circumstances viewed
through a different problem construction.
Another example of this technological/human performance bifurcation exists with
the issue of cellular telephones in vehicles. Bener, Özkan, and Haigney (2006)
researched the effects of mobile phone use and found the distraction of mobile phone use
38
while driving increased collision rates by statistically significant amounts. Similar
findings have prompted many states to ban mobile phone use while driving—except with
the technological solution of a hands-free device (e.g., Bluetooth headset). However,
Abdel-Aty (2003) found that drivers are distracted when talking on a cellular phone—
hands-free or otherwise—and his results were confirmed in a meta-analysis of 125
studies on the same topic (Mccartta, Hellingaa, & Bratimana, 2006).
Finally, many authors have considered effects of speed on traffic collision and
fatality rates among drivers in general (Friedman, Barach, & Richter, 2007; Friedman,
Hedeker, & Richter, 2009; Garber & Gadiraju, 1992; Pant, Adhami, & Niehaus, 1992).
These researchers have universally found that higher mean speeds associated with higher
speed limits (e.g., a 65 MPH limit versus a 55 MPH limit) result in higher collision
fatality rates. Using a human performance construct, the problem is drivers—LEOs or
the general public—have a diminishing capacity to manage higher speeds. Later, I
reconceptualize this issue (speed) as a regulatory/public affairs issue.
LEO Fatalities as a Managerial Control Problem. Many practitioners and
scholars (e.g., Batiste, Wagers, & Ashton, 2011; IACP, 2011; Gustafson, 2009;
Gustafson & Cappitelli, 2010; Schultz, Hudak, & Alpert, 2010) have examined LEO
traffic collisions and fatalities as an issue of command and control management and/or
policy. The general premise in this perspective is that law enforcement executives,
managers, and supervisors must set expectations (both in policy and practice) that hold
LEOs accountable for driving responsibly. Because studies (Gustafson, 2009; NHTSA,
2010b) have shown that approximately half of all LEO traffic collision fatalities involve
either (a) excessive speed while not assigned to an emergency call or (b) the LEO driver
39
not wearing a seatbelt, there is an objectively reasonable sense that the fatalities are
avoidable and that there is a lack of will or discipline on the part of decision-makers.
Schultz, Hudak, and Alpert (2009, 2010) investigated traffic collisions and
fatalities resulting from police pursuits. They found that more permissive pursuit policies
or regulations (e.g., granting officers discretion to pursue for minor traffic violations) are
linked to higher collision rates and fatalities. Some law enforcement agencies prohibit
pursuits and others restrict pursuit initiation to only violent offenders (Alpert & Smith,
2008). In this view of the problem, restricting or prohibiting pursuits might be part of the
solution.
LEO Fatalities as a Cognitive/Behavioral Problem. Since the 1930s, scholars
have worked to develop a psychological theory of traffic collisions separating roadway or
vehicular variables from human factors (Sorensen, 1994). Consistent with other literature
reviewed, a series of authors have, over several decades, developed a theory of traffic
collisions based on human characteristics of driver behavior (Fell, 1975; Fell & Tharp,
1969; Salminen & Lähdeniemi, 2002). Fell’s (1975) model includes driver perceptions
as a factor in traffic collisions—for example, where a driver has to perceive the proximity
of an oncoming car. Clearly, there is no state-level contribution to the individual
perceptions of drivers and perception might arguably border on the physiological. Other
components of Fell’s model, however, include exposure and risk-taking. In the case of
LEOs, exposure could be the number of hours officers work and risk-taking could include
the necessity (or prohibition) of engaging in vehicular pursuits (as described above).
Other authors have identified distinct behaviors such as aggression and risk-taking
that contribute to collision involvement (Castro, 2008; Dorn et al., 2005, 2010). From
40
this perspective, individual personality/behavioral/cognitive factors increase collision
likelihood. Responses to the cognitive/behavioral problem have included training,
screening (for purposes of prevention), and, from the technological aspect, surveillance
(Castro, 2008; Dorn & Barker, 2005; Dorn et al., 2010). In the case of training, the
premise is that officers can gain knowledge, skills, and abilities that will enable them to
monitor and manage their behavior. With respect to screening, which is widely used in
the UK, the idea is to identify risky or problem drivers before they are put into an
emergency driving role. Last, with respect to surveillance, technologies like in-car video
cameras and global positioning system (GPS) monitoring of speed provide for various
social control systems (Giddens, 1991). One system might include concertive control,
whereby coworkers monitor each other and adopt the organizational goal (e.g., safe
driving) as their own (Barker, 1993). Other systems might include panoptic control in the
Foucaultian sense whereby workers monitor themselves for fear of constant surveillance
(Sewell, 1998) or strictly bureaucratic (i.e., managerial) control (Taylor, 1911)—
consistent with managerial control described previously.
LEO Fatalities as a Socio-Cultural Problem. Mention of control systems
naturally leads to a discussion of LEO fatalities as a socio-cultural problem. Research in
this area is relatively new and developing. The premise from this perspective is that a
driving culture may develop which encourages either better or worse driving behavior.
Colloquially, this has been described as a continuum ranging from a culture of safety—
one concerned with getting to the destination without loss of life or property—to a culture
of speed—one primarily orientated to getting to the destination fast and first if possible
(Gustafson & Cappitelli, 2010). Wehr, Alpert, and Rojek (2012) have studied this as an
41
agency-level phenomenon, as have Alpert, Rojek, and Porter (2012) and there is evidence
to suggest there is an industry-wide culture among LEOs wherein they (officers) have a
sense that the rules of the road are for the public, but not for them. Responses to the
socio-culture problem construction mirror those described for both cognitive and
managerial problems. The possible interventions include training, discipline, regulation,
and monitoring.
LEO Fatalities as a LEO Problem. Some scholars have investigated LEO traffic
fatalities as a LEO phenomenon. This frame of reference suggests that there are inherent
or enculturated differences between LEOs and other drivers. This perspective is closely
related—if not intertwined—with the cultural orientation described above. NHTSA
(2010b) published Characteristics of Law Enforcement Officers’ Fatalities in Motor
Vehicle Crashes which explored demographic and situational factors in all fatal LEO
traffic collisions in the US from 1980 to 2008. NHTSA found that:
The LEO and non-LEO groups show substantially different characteristics at
crash time, first harmful event, roadway function class (rural/urban), emergency
use, fire occurrence, rollover, most harmful event, impact point, vehicle
maneuver, crash avoidance maneuver, age, sex, person type, seating position,
restraint use, and air bag availability and deployment. (p. i)
The NHTSA study, however, failed to provide substantive analysis exploring why these
differences might exist. The work primarily provides a statistical description of LEO
fatalities (nationwide) through data mining (similar to Chapter 4 of this study with the
primary difference that this study subdivides the population by state whereas the NHTSA
study considers LEOs as a homogenous group).
42
NHTSA (2010b) compared variables from the Fatality Analysis Reporting System
(FARS) database for LEO and non-LEO traffic fatalities, which are collected by traffic
investigators at the scene of each fatal traffic collision. Some of those variables listed
above as substantially different are more meaningful than others. The variable
emergency use, for example, refers to the use of emergency lights and/or siren at the time
of the fatal collision (NHTSA, 2008). Clearly, one would expect that LEOs and the
general public would differ in this characteristic because the general public does not have
emergency lights and sirens in their vehicles. Age, sex, and person type are also variables
where significant differences are to be expected. The age range of the general population
spans 0 to 100-plus years of age. The age range of the LEO population is much narrower
since LEOs almost universally must be adults (age 18 or older) and usually stop working
(i.e., cease to be LEOs) by the time they are in their 50s or 60s even when working a full
career. In terms of the variable sex (i.e., gender), females and males are pretty evenly
distributed in the general population (US Census Bureau, 2012) whereas males are
significantly over-represented in the LEO population (FBI, 2009b). Finally, the variable
person type is the actual selection variable used to differentiate LEOs from the general
public in the FARS database (NHTSA, 2008) and the NHTSA (2010b) study.
Research currently in progress at the University of California at Berkeley and
CalPOST is seeking to add detail and analysis to the question of demographic and
situational factors identified by NHTSA (2010b) (Rice & Gustafson, 2012). The
Berkeley study analyzes a much larger sample of collisions by including injury as well as
fatal collisions as collected at the state-level in California (Rice & Gustafson, 2012).
43
Research Gap and Need
As described above, research on traffic collisions and fatalities is well-developed.
Many previous studies have investigated the factors that lead to traffic collisions and
traffic fatalities as a small percentage of collisions. Studies of LEO traffic deaths have
developed over the past decade, but remain far less comprehensive. State level studies
have been conducted, but not specific to LEOs. Zlatoper (1991), for example, studied
traffic fatalities at the state level and found increased spending on law enforcement and
highway safety reduced fatality rates. Multiple authors have considered the effects of
driver training (e.g., Christie, 2001; Gustafson, 2009), finding that, in general, driver
training programs have some collision reduction effects. Wagenaar, Maybee, and
Sullivan (1988) examined traffic collision fatalities in eight states and found significantly
lower fatality rates after mandatory seat belt laws were implemented. Additionally,
Rivara and Mack (2004) and Hutson, Rice, Chana, Kyriacou, Chang, and Miller (2007)
conducted nationwide descriptive studies of traffic fatalities resulting from law
enforcement pursuits. Rivara and Mack failed to provide state-level descriptions, and
both studies included civilian deaths as well as LEO deaths resulting from pursuits.
In each of these studies, authors have aggregated individual data over time (i.e.,
multiple years) and space (i.e., multiple cities, counties, or states) to construct a larger
trend. Few studies, however, have looked solely at LEOs and these are typically limited
samples as they tend to examine individual departments or occasionally an individual
state (Gustafson, 2009). Additionally, few studies have considered state-level factors.
None have considered LEO traffic fatalities across states (i.e., cross case analysis) and
little is known descriptively about these occurrences, let alone about causal mechanisms
44
or significant factors. The gap, therefore, is twofold. First, a comprehensive description
of LEO traffic fatalities and related variables by state is critical. Second, an investigation
of the role of state-level variables in law enforcement traffic fatalities is needed. These
data and findings will provide state policymakers a factual anchor point from which to
base regulatory standards. This leads to a necessary understanding of LEO fatalities as a
public policy problem.
LEO Fatalities as a Public Policy Problem
Based on my understanding of the problem and the many approaches that have
been adopted in studying and addressing the issue, LEO traffic fatalities are first and
finally a public policy problem. I have described its enormous scale that cuts across
states, demographics, and disciplines, and its rooting in the public realm (both in terms of
cost and impact). I have also described its enduring nature over the course of more than
15 years. For these reasons, this research adopts the perspective of the policy sciences.
The policy sciences somewhat ambiguously refer to those fields and disciplines
concerned with “issues of the most intractable nature” (deLeon, 1988, p. 2). They are
interdisciplinary and solution oriented (Lasswell, 1951). As such, they provide exactly
the right framework for this problem. As I have described this problem based on the
variance in LEO traffic fatality rates and the practices that lead to those fatalities, a policy
solution will advance standardization from a macro perspective.
A State-Level Approach. Because there are no national law enforcement
standards in the US—an outcome of the feature I colloquially refer to as the miracle of
federalism—and because there are approximately 15,000 law enforcement agencies in the
US (FBI, 2009a), a state-level approach is the most logical and practical way to effect
45
change. The significant variability described at the state level suggests something is
happening among variables at this level of policy.
The State as a Site for Study and Change. Practically speaking, the reasons for
identifying states as sites for study and change are twofold. First, data are readily
available and many differences are categorically distinct and within the control of
policymakers. Second, as already detailed, the majority of law enforcement standards are
established at the state-level and the scale is amenable to change. Effecting policy and
practice modifications in 50 states and the District of Columbia is—while daunting—
much more feasible than trying to do the same with 750,000 LEOs or even 15,000 law
enforcement agencies individually. There is leverage to be had at the state level.
Moving On to the Research
Chapter 1 outlined the problem and this chapter has reviewed previous research
and historical context relevant to its understanding. Next, Chapter 3 specifies (a) the data
for this research—state-level variables that can be impacted by policy, (b) research
questions that interrogate the policy aspects of the variables, and (c) methods used in the
analysis. Additional literature is introduced to support the research questions, data, and
particularly the methodology for determining what aspects of state-level policy, law, and
practice explain the variance in LEO traffic fatalities.
46
CHAPTER III
RESEARCH QUESTIONS, DATA, & METHODS
There are many worthwhile questions to explore related to law enforcement
officer (LEO) traffic fatalities. There are also many data sources and methods available
for analysis. This chapter hones the universe of possibilities to succinct state-level
questions, relevant and consistent data, and reliable, proven methods. The goal of this
research is to produce clear and viable policy and implementation recommendations for
law enforcement practitioners.
Research Question and Sub-Questions
This research centers on the following questions:
Research Question 1
What accounts for different law enforcement traffic fatality rates by state in the
US?
Research Question 2
Which of the explanations for different collision rates are amendable via policy
and practice adaptation?
Research Question 3a
What policy and practice adaptations are most likely to reduce collision rates?
Research Question 3b
To what extent are identified policy and practice adaptations feasible given social
and institutional complexities such as labor and management rights, budget limitations,
agency culture, and related factors?
47
Relationships Among Research Questions and Sub-Questions
As alluded to earlier, the primary research question could generate multiple
answers with varying degrees of relevance for policymakers and practitioners. For
example, a finding that population density, weather, or topography effects law
enforcement traffic fatality rates may be less notable than a finding that duration of law
enforcement driver training or highway speed limits does so as well. This issue
illustrates the importance of the first sub-question. Those explanations that can be
influenced by policy and practice have decidedly more value than those that cannot.
Colorado, for example, cannot change its weather or topography; however, it can change
its LEO driver training programs and its highway speed limits. Sub-questions that
investigate the efficacy and feasibility of policy and practice adaptations follow
intuitively. Reducing speed limits by one-half might prove highly effective for lowering
traffic fatality rates, but it is almost certainly not feasible given social and economic
expectations for transportation. Therefore, each question must hone the answer from the
previous question in order to be relevant to the identified problem so that answers to
subsequent questions will facilitate clear and appropriate recommendations for policy and
practice. In the following section, I provide a rationale for investigating these questions in
terms of regulation and economic investment and constraint.
Rationale for Applying a Regulatory and Economic Framework
The variables described in Chapter 2 that contribute to collision fatalities can be
grouped into state-level factors in terms of varying law enforcement traffic fatality rates
under the umbrellas of regulation and welfare economics. These concepts are described
48
below, but generally refer to rules, laws, and regulatory policies (i.e., regulation) and
spending or investment (i.e., welfare economics).
Regulation, as standards setting and as a barrier to entry (Gerber & Teske, 2000;
Teske, 2004) into law enforcement and law enforcement vehicle operations or pursuits,
establishes requisite circumstances for acceptability and, depending on stringency, lowers
collision rates and improves highway safety (Mashaw & Harfst, 1987). For example,
local LEOs in the United Kingdom (UK) who engage in vehicular pursuits are required to
have more than 10 times the amount of driver training than a local LEO in California
(Gustafson, 2009). In general, the UK regulates roadways and transportation (to include
law enforcement) much more stringently than the US and, perhaps as a result, traffic
collision and fatality rates are significantly lower in the UK than in the US (Hedlund,
2007). In this research then, states that more or less actively regulate law enforcement
driver training, vehicle operations, speed limits, pursuits, or seatbelt use, should
experience different traffic fatality rates as a result of regulatory impact.
Welfare economics (Weimer & Vining, 2005), defined here simply as choice via
available funding, drives standards-setting and infrastructure decisions and therefore
impacts traffic fatality rates. For example, if any given municipality were to build
pedestrian overpasses at every roadway intersection instead of simply painting
crosswalks in the roadway, vehicle-versus-pedestrian collisions could be reduced
dramatically. The cost of such a choice would likely be unbearable, and therefore
welfare economics leads to other choices (Miller, 1992). In this respect, welfare
economics may frequently work in conjunction with regulation wherein more stringent
49
regulations are desirable, but not affordable. In this research, state spending on law
enforcement traffic operations and infrastructure is expected to influence traffic fatalities.
Propositions
Fatal traffic collisions are a set of the universe of traffic collisions. As such, the
factors that lead to traffic collisions also lead to fatal traffic collisions and LEO traffic
collisions and fatal LEO traffic collisions. Additionally, seat belt use is a variable that
does not affect the occurrence of traffic collisions, but does affect the likely mortality of
traffic collisions. Therefore, it is possible to have traffic collisions with and without seat
belt use, with and without fatalities, and with and without LEO involvement as illustrated
in Figure II.1.
Figure II.1 Universe of Traffic Collisions and Possible Subsets.
This research is premised on the overarching proposition that directional
relationships exist among key state-level variables and the dependent variable law
enforcement traffic fatalities (e.g., increased spending on highway safety results in
decreased LEO traffic fatalities). These key variables (discussed in Chapter 2) include
50
training (Castro, 2008; Dorn et al., 2010; Gustafson, 2009; IADLEST, 2005); standards
(Carte, 1969; IADLEST, 2005); speed (Friedman, Barach, & Richter, 2007; Friedman,
Hedeker, & Richter, 2009; Garber & Gadiraju, 1992; Pant, Adhami, & Niehaus, 1992);
and spending (Zlatoper, 1991). Based on the literature reviewed, the following
propositions are posited:
Proposition 1
LEO traffic fatalities are a function of economic investment (i.e., spending).
States that invest more on law enforcement and highway safety will have lower law
enforcement collision fatality rates.
Proposition 2
LEO traffic fatalities are a function of regulatory standards (e.g., laws, policies).
More stringent law enforcement requirements will result in lower law enforcement
collision fatality rates. These standards can be examined in several instances. (a) LEO
driver training (amount/duration): States that require more law enforcement driver
training will experience lower LEO collision fatality rates. (b) Seat Belts (i.e., mandatory
versus optional use): States that require officers to wear seat belts will have lower law
enforcement collision fatality rates. (c) Pursuits: States that limit or otherwise govern law
enforcement pursuits will experience lower LEO collision fatality rates. (d) Speed: States
that have either (i) lower overall speed limits or (ii) restrictions on LEO speeds will
experience lower LEO collision fatality rates. (e) Exposure (as it relates to fatigue):
States that limit the amount of time LEOs can work will experience lower LEO collision
fatality rates. (f) Distractions (e.g., mobile phone use): States that prohibit mobile phone
use while driving (for LEOs) will experience lower LEO collision fatality rates.
51
Concepts and Definitions
Several concepts (e.g., regulation and training) are used as variables in this
research design. Definition and operationalization of these concepts notably impact the
analysis. Explanations of these concepts as relevant to this study include the following
five conditions (a) Regulation/standard: in general, an identifiable state-level law, rule, or
policy requiring or prohibiting some action (relevant to the dependent variable).
Colorado, for example, has a law that requires LEOs to wear seat belts when driving.
California does not have this law. A finding that LEO traffic fatalities are higher in
California than Colorado might be partially attributed to this regulation/standard. (b)
Law enforcement officer (LEO): an individual authorized by law to enforce laws and
make arrests (IADLEST, 2005). This analysis focuses on uniformed, front-line law
enforcement (i.e., police, sheriff, highway patrol) as opposed to specialized law
enforcement (e.g., FBI special agents or alcoholic beverage control officers). As defined
here, LEOs also are restricted to full-time paid personnel. These distinctions are common
in the law enforcement community. (c) Fatal traffic collision: an instance where an
individual dies as a result of any vehicular movement (e.g., colliding with another
vehicle, hitting an object, overturning, or a vehicle hitting a pedestrian). (d) Economic
investment: funds (i) allocated to highways and transportation or (ii) spent on highway
safety and enforcement. (e) Driver training: required number of hours of entry-level
training required of LEOs.
Data Sources, Types, Description, and Issues
Data for this research were sourced from many governmental agencies and two
nonprofit entities. The data presented represent both quantitative and qualitative
52
measures and while most quantitative measures represent actuarial counts, some
estimates are also employed. Several issues were encountered during the acquisition and
processing of the data, including factual discrepancies in counts, even among federal
government sources. These data and issues are detailed in the following sections.
Sources
This research covers the 50 states and Washington, DC, for the period 1995-2009.
Data were obtained from the sources listed in Table II.1.
Table II.1 Types and Sources of State-Level Data.
Data Type Source
Area in Square Miles US Census Bureau, 2009
LEO Driver Training Hours IADLEST, 2005
LEO-Involved Traffic Deaths NHTSA, 2012
LEO Traffic Deaths FBI, 1995b; 1996b; 1997b; … 2009b
ODMP, 2012c
NLEOMF, 2012c
LEO State Populations FBI, 1995a; 1996a; 1997a; … 2009a
Highway Receipts/Expenditures FHA, 1995; 1996; 1997; … 2009
Policy and Opinion Information CalPOST, 2012
Populations FBI, 1995b; 1996b; 1997b; … 2009b
Speed Limits Federal Highway Administration (FHA), 2004b
Traffic Deaths NHTSA, 2012
Urban/Rural Roadway Mileage FHA, 1995; 1996; 1997; … 2009
53
Types and Description
As already detailed, the study contains data for the 50 states and Washington, DC
for the period 1995-2009. As such there are 51 cases and 15 observations for each case,
which results in a dataset for the study with a total of 765 observations. The exception to
this sample size is the survey data, detailed below, which covers a one-time snapshot and
has missing cases. The variables used (and listed in Table II.1) are described below.
Area in Square Miles is a continuous variable that indicates the land mass of each
case (i.e., state) in this study.
LEO Driver Training Hours is a discrete variable that indicates the minimum
number of hours of basic (i.e., entry-level) driver training a LEO must have in each state
according to published regulatory and training standards (IADLEST, 2005). Many states
do not require driver training, which is reported as a zero (0) and some states do not
specify driver training (which is treated as missing data).
LEO-Involved Traffic Deaths is a discrete variable collected by NHTSA in its
FARS Encyclopedia that is determined by the code “5” in the “Special Use” category of
the FARS database (NHTSA, 2008, p. V-85). This is a vehicle-related factor in the
database, and therefore any fatal traffic collision where a police vehicle was involved will
be captured by this variable whether it was the officer who died or another person in
another involved vehicle. This variable does not relate to the status of the driver (e.g.,
on-duty or off duty, peace officer).
LEO Traffic Deaths is a discrete variable that has been developed from three
sources. The FBI (1995b-2009b) Law Enforcement Officers Killed and Assaulted
(LEOKA) publication identifies counts by jurisdiction and manner of death of LEOs
54
killed in the line of duty. When the FBI identifies such a death, the next of kin of the
deceased officer is available for a tax-free death benefit from the Public Safety Officers'
Benefits Program of the Bureau of Justice Assistance in an amount currently set at
$323,035.75. The FBI carefully investigates cases it identifies as line of duty deaths and
this only applies to public safety officers, which are generally defined as sworn LEOs.
The other two sources are the ODMP and the NLEOMF. These nonprofit organizations
are primarily established to recognize and honor fallen officers. These organizations are
much more lenient in defining officers and their databases include federal, tribal, military
police, and other officers not necessarily recognized as LEOs in the FBI’s publication.
Therefore, overall counts of deaths are higher as recorded by the ODMP and NLEOMF.
What is notable for this study is that the ODMP and NLEOMF provide a summary of the
events leading to the death of each LEO in their respective databases. I used these two
sources in conjunction with the FBI (1995b-2009b) LEOKA publications to review every
LEO traffic fatality (N=840) contained in this variable. By doing this I was able to
maintain my metric of uniformed, fulltime, frontline LEOs. Issues with this variable are
described in the following section.
LEO State Populations is a discrete variable that reports the number of state and
local LEOs in each state (FBI, 1995a-2009a). This is understood to be a count of fulltime
paid LEOs as defined by the FBI (1995a-2009a) in its Crime in the United States
publication.
Highway Receipts/Expenditures represents two discrete variables. Receipts is a
measure of total available funds reported to the FHA (1995-2009) by state and local
governments. Expenditures is a measure of funds allocated by state and local
55
governments to highway safety and enforcement as reported to the FHA (1995-2009).
Both are reported in 1,000s of dollars.
Policy and Opinion Information refers to state-level survey data collected by
CalPOST (2012). CalPOST produced these survey data via distribution of the instrument
to the population of US POST agencies (N~51) through CalPOST and the IADLEST and
US state highway enforcement agencies (e.g., California Highway Patrol, Colorado State
Patrol) through CalPOST and the International Association of Chiefs of Police (IACP)
Division of State and Provincial Police. Thirty-six responses were received by CalPOST
(2012). The population of the survey is approximated (N~51) when bifurcated into these
two categories because in some cases (like Hawaii and Washington, DC) the POST
agency and the highway traffic agency are one in the same. Therefore, the total
population is not 102 (i.e., two responses for each state). Since not all states responded to
the survey, the true population is unknown.
The survey posed questions about regulatory factors (e.g., statewide seatbelt laws,
speed laws for LEOs, cellular phone use while driving), budgets, and also presented
practitioners with a series of Likert items which asked them to respond on a five-point
scale (Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree) to statements
regarding the effects of policy, law, and spending. Finally, the CalPOST survey also
presented state-level difference data regarding law enforcement training and LEO fatal
traffic collisions from Gustafson and Rice (2010) and asked practitioners to report their
agreement on a five-point Likert scale (No, Doubtful, Maybe, Probably, Yes) to
statements like: “The LEO traffic fatality rate in my state is a matter of chance”
56
(CalPOST, 2012, Question 15). A summary of the CalPOST survey instrument is
provided in Appendix A.
Populations is a discrete variable that refers to counts and/or estimates of LEOs
and residents by states as reported by the FBI (1995a-2009a).
Speed Limits is a discrete variable that reports the maximum daytime speed limit
on federal highways by state as reported by the FHA (2004b).
Traffic Deaths is a discrete variable that reports counts of persons killed in vehicle
collisions in the FARS database (NHTSA, 2012). This variable is aggregated by state
and year in this study although FARS does provide data on each fatal incident at the
person, vehicle, and collision levels (e.g., three people could die in two vehicles involved
in one collision and FARS data is parsed to allow analysis from each of these levels).
FARS only contains data for collisions where one or more deaths occur within 30 days of
the collision. For example, if a person is involved in a traffic collision and spends five
weeks (i.e., 35 days) in the hospital before dying from injuries sustained in the collision,
the collision is not reported in FARS (and would not be included in this variable).
Urban/Rural Roadway Mileage is a discrete variable that indicates aggregated
miles of urban and rural roadway by state as reported by the FHA (1995-2009).
All of the quantitative data from these variables (not contained in the CalPOST
survey) are included in the Gustafson Dataset presented in Appendix B. Inclusion of
these data allows any interested researcher the ability to verify, test, or recreate all the
analyses and assertions presented in this research.
57
Data Limitations and Challenges
As mentioned in Chapter 1, data in this research did not always match across
sources. This problem may best be explained through a scenario. An on-duty LEO was
driving a vehicle on routine patrol when she was involved in a traffic collision with a
legally drunk driver. The LEO died as a result of the traffic collision. The ODMP would
report this fatality as a vehicular assault. The NLEOMF and the FBI would report it as
an automobile accident. If the scenario is modified slightly, the reporting outcomes
change. For example, imagine that the LEO tried to pull over the drunk driver and the
drunk driver tried to get away and in so doing intentionally rammed the LEOs car and
killed the LEO. The ODMP would still report this as a vehicular assault. The NLEOMF
would report it as an automobile accident and the FBI would report it as a homicide. In
my research, I would locate the LEO death in the first scenario in all three sources. In the
second scenario, the LEO death would only surface in the ODMP and NLOMF
databases. The FBI does not have a searchable database, so I would need to read
thousands of pages of homicide data incident by incident to have any chance at locating a
vehicle-related homicide.
Other issues include the lack of common definitions—of LEO, of traffic collision,
or of death, for example. The ODMP and the NLEOMF tend to recognize anyone who
has a badge and a gun in a public or semi-public setting as a LEO. This might include a
tribal officer. In some places tribal officers are effectively police officers and in other
places they are casino security. Because they are uncommon and they tend not to fall
under the umbrella of state-level regulation at issue in this research, these cases have
been excluded. Other LEO definition issues alluded to previously relate to the
58
employment status of LEOs. In some states, like North Carolina for example, individuals
can volunteer and be deputized without law enforcement training. These individuals
might have a uniform and a badge and a gun and a police car, but they frequently fail to
meet a professional standard. To the extent I have been able to identify these types of
personnel, they have been excluded. They appear to represent a small percentage of
cases in the data. Finally, some states have uniformed law enforcement that drive marked
vehicles, but would not be considered front line or general law enforcement. Examples
include fish and game wardens, department of motor vehicles police, and housing
authority police. These cases have been excluded.
With regard to defining a traffic collision, there are also challenges in the data.
As already detailed, NHTSA only recognizes a fatal collision if the death occurs within
30 days of the collision. The other data sources do not use this metric. For instance, in
many cases officers died in one year (e.g., 1995) from injuries sustained in a collision in
another year (e.g., 1985). I have made case-by-case determinations. I have excluded (a)
deaths resulting from collisions that occurred outside the study timeframe (i.e., 1995-
2009), (b) deaths that appeared unrelated to the collisions (e.g., a case where a LEO
developed an infection during post-collision surgery and subsequently died a short time
later), and (c) deaths where the LEO lived for a year or more on life-support following
the collision.
I have included cases where a pedestrian LEO is killed by a motor vehicle. This
is most common on highways where the LEO makes a traffic enforcement stop and is run
over by passing traffic. There are several aspects of training and regulation that relate to
this situation, so inclusion in general is justified. However, some cases are simply hard to
59
define. In one instance a LEO had a collision with a fixed object. The LEO got out of
the patrol vehicle after the collision, but then the patrol vehicle rolled backward over the
LEO resulting in death. One data source listed this as a pedestrian collision and the other
two as a vehicle collision. This case was included (as a vehicle collision). In another
case, a LEO had a motorist stopped and approached the driver on foot. The driver
attempted to flee the LEO. The LEO grabbed onto the car and reached in the driver’s
window. The LEO was drug beside the car and died. This case was excluded. I
reviewed hundreds of these cases and subjectively tried to create a dataset that
consistently identified front line, general LEOs killed in traffic.
Finally, there are what appear to be capture and entry issues with the FARS data.
In 2002, NHTSA began including a person-level variable that assuredly identified a
LEO’s involvement in a fatal collision (NHTSA, 2008). As already detailed, the only
source in the database is the vehicle-level variable that identifies a police car. NHTSA
has FARS analysts that receive fatal traffic collision reports from law enforcement
agencies in each state, though the formats vary among states. In California, for example,
the standardized traffic collision investigation form (CHP 555) has multiple fields that
can be checked to identify a police motorcycle. However, when that form is entered in to
the Statewide Integrated Traffic Reporting System (SWTRS), the coding may not capture
all the detail. This in turn is sent to the FARS analyst for entry. Because I have reviewed
so many cases in great detail, and because there are generally not a lot of cases in each
state each year, I have noticed discrepancies where I know that there was a fatal LEO
collision, but it does not show up in the FARS dataset. To deal with this sort of error I
have created maximum variables that report the highest number of fatalities across
60
multiple sources. When I use this method in analyses in Chapters 4 and 5 it will be noted
accordingly.
Methods and Analytical Techniques
This research employs a mixed methods design. The methodology has three
components: (a) descriptive statistical analysis, (b) cross-sectional regression analysis,
and (c) qualitative content analysis. The unit of analysis is a state and the unit of
observation is both individual LEOs and states. In terms of descriptive analysis, means,
rates, counts, standard deviations, and comparisons are provided longitudinally across,
between, and among states. In terms of regression analysis, state law enforcement fatal
collisions (rates and counts) is the dependent variable of interest analyzed against state
regulatory, spending, and demographic independent variables. Finally, in terms of
content analysis, a comparative methodology is used in conjunction with Likert response
comparisons (and descriptive statistics) with the survey data. Each of the analytical
techniques is discussed in detail in the following sections.
Descriptive Statistical Analysis
Chapter 4 provides descriptive statistical analysis of LEO fatal collisions by state
for the years 1995-2009 and establishes mean fatal collision rates for each state. The
LEO fatal (collision) death rate for each state is calculated by dividing the number of
LEOs killed in traffic collisions by the LEO population and multiplying by 100,000 (a
simple mathematical technique for reducing leading zeros). The mean rate is the
arithmetic average of the yearly rates (1995-2009).
Counts, rates, means, and standard deviations are provided for LEOs numerically
and in graphical form to provide useful comparative data within, between, and among
61
states. These data are also reported for the general population for comparison and
contextualization. On the one hand, consistently high rates across LEO and general
populations in a state might be indicative of a more dangerous traffic environment
overall. A LEO rate, on the other hand that is consistently higher than the general
population rate might be indicative of a significant LEO-specific factor. These
comparison techniques are consistent with other studies of collision fatalities (Clarke &
Zak, 1999).
Regression Analysis
Cross-sectional regression analysis is presented in Chapter 5. The relatively low
frequency of LEO traffic fatalities by state presents some analytical challenges, though
several methods offer potential analytical power and sophistication to address this issue.
Diggle, Heagerty, Liang, and Zeger (2002) provide justification for using random-effects
versions of Poisson and negative binomial regression to analyze longitudinal data. This
accounts for the repeated-measures as used in this study (i.e., 15 (annual) measures from
each state). Still, other variables remain fixed in some states throughout the study period
(e.g., highway speed limits and driver training hours). Multiple models were developed
and tested to identify the best fit for a variety of analyses. These are specified as they are
reported in Chapter 5.
In general, these methods address the problem of clustered data. For instance,
imagine 100 research participants provide monthly blood pressure measurements over the
course of a year (i.e., 12 measures for each case, n=1200). If a simple linear regression
was used to analyze these data to see how certain covariates predict diastolic blood
pressure, various coefficients and standard errors would be generated by the model. The
62
problem is that the standard errors in a simple linear regression model assume that all
observations are independent. This would mean that any two measurements from one
participant would be no more alike than any other two measurements from any other
participant(s) in the study. This assumption is clearly inaccurate since a single person’s
blood pressure in a given month is likely to be similar to their blood pressure in another
month. In other words, a single person’s blood pressure one month is not independent
from that same person’s blood pressure in another month. They are related and the data
would reflect that (i.e., smaller variance within the case than across cases). A visual
inspection of the data from this hypothetical 100-person study would likely reveal this
difference between within case variance and across case variance. There will be
clustering. The same is true with the state-level data in this study. Any two data points
(i.e., values) on any variable (e.g., LEO death rate, LEO population, highway spending,
urban roadway mileage) will be less variable (i.e., more clustered) with a state (e.g.,
Colorado) than between states (e.g., Florida and Oregon). The random intercept model is
a regression model with an extra component (grouping—i.e., by state) to allow each state
to have its own intercept to account for these repeated measures and clustering.
Compared with simple linear regression, a random intercept model will have similar
coefficients and larger (and more accurate) standard errors (Rabe-Hesketh & Skrondal,
2008).
Besides clustering, the other analysis challenge for this data is over-dispersion—
greater variability than would be expected in a population. In one sense, this problem
bolsters my premise that LEOs across states are not a single homogeneous population.
Both random intercept Poisson and negative binomial models address over-dispersion
63
(Diggle, Heagerty, Liang, & Zeger, 2002; Rabe-Hesketh & Skrondal, 2008). I used both
in the analyses described in Chapter 5 and found no benefit to the negative binomial
model. I selected Poisson because it is the classic count model for traffic collision data
representing a discrete model beginning at zero (De Smet, 2008; Hayter, 2006). In terms
of traffic collisions, it reflects that there cannot be less than zero collisions and the
observations increase in whole integers (i.e., one collision, then two, etc.) (Weiss, 2008).
Content Analysis
Because p-values (i.e., the statistical significance of various models presented in
Chapter 5) can only reveal the presence, strength, and direction of relationships among
variables, this research relies on other data to provide core policy-actionable information
and findings. The CalPOST (2012) survey data serves this purpose as two questions
solicited open-ended, narrative responses and others allowed open-ended comment (e.g.,
to clarify a Likert response). The CalPOST survey was targeted to individuals with
known expertise and designed based on knowledge established through earlier relevant
studies (Gustafson & Rice, 2010). It posed questions parallel to the propositions in this
research and focused respondents on the specific variables of interest consistent with
good survey design (Frankfort-Nachmias & Nachmias, 2008). As such, from a grounded
practical theory perspective (McNabb, 2002), I have aligned the survey data with
observable trends in the statistical findings to establish potential causal mechanisms (i.e.,
reasons for differences in LEO fatal collision rates among states that are grounded in the
observations). A summary of the CalPOST survey instrument (i.e., questions and design)
is presented in Appendix A.
64
The interview data were analyzed using a version of the constant comparative
method of qualitative research (Glaser & Strauss, 1967; Strauss & Corbin, 1998). First, I
fully reviewed all of the narrative responses in the survey data and coded sections of the
text grouping common statements. Many respondents, for example, wrote statements
indicating a linkage between the driving public and LEOs (i.e., suggesting that LEOs
drive in relationship to the public they police—a relationship between LEO and general
public fatalities). The goal of this type of coding is to “open up the inquiry” (Strauss,
1987, p. 29) since little is known collectively about how law enforcement leaders
understand the issue of LEO traffic fatalities. As new themes emerged (e.g., response to
emergency calls), I articulated new codes and read and reread the narrative data until I
was able to identify no additional themes. This process of integration allowed me to
reshape and compare and contrast categories (e.g., pursuits, responses, vehicle-related,
LEO-related, public-related, etc.) and develop a conceptual scheme of themes across the
interviews (Lindlof & Taylor, 2002). These integrated categories generated an
interpretation of theoretical and practical relevance, changing the nature of categories
from collections of statements or events into meaningful constructs for policy action as
detailed in Chapters 6 and 7 (Lindlof & Taylor, 2002).
This method provides a form of process tracing (Gerring, 2007) wherein
statistically significant variables are linked to state-level policies and practices, thereby
identifying potentially causal mechanisms. This approach also provides for triangulation
of propositions and trends identified through previous literature and data analysis to
enrich overall understanding of the issue (Frankfort-Nachmias & Nachmias, 2008; Yauch
& Steudel, 2003).
65
Validity and Limitations
Issues of validity in research are twofold—internal and external—with internal
pertaining to the study itself and external pertaining to a larger population that is not
considered by the study (Gerring, 2007). Because this research involves a population-
level inquiry of all US LEOs killed in traffic collisions, issues of external validity are
significantly diminished. They are not entirely mitigated due to the data issues described
above. However, threats to external validity are unlikely.
The more challenging issues for this research are internal validity (e.g., issues of
cause and effect among independent and dependent variables, bias, and history) and
construct validity (i.e., measurement and the proposed relationships between and among
variables) (Van de Ven, 2007). In this study, selection bias—the idea that officers who
select law enforcement jobs in different states may have innate differences among them
outside the measured variables—is a significant possibility. The use of both statistical
assessment and content analysis of qualitative survey data are intended to uncover any
confounds of this nature. Additionally, the population-level starting point of the study
reduces the likelihood of innate differences remaining undetected.
History may be another threat since this research design clearly does not capture
all variables that may exhibit significant change over time (by state) for the study period
1995-2009. It is possible that regulatory variables changed during that period and
increased or decreased overall collision rates without the model (or the survey
respondents) recognizing the effects of the changes. I have taken every reasonable
precaution to identify and address these issues; however, this is an unavoidable reality of
grounded practical research. In no way does this design—in either its quantitative or
66
qualitative components—presume to explain all observed variability. Rather, the design
aims to identify actionable variables that are known and observable.
Finally, the construct validity of this study is a potential threat. One of the
scholarly contributions of this research is its linkage of state-level factors to individual
collision contributors such as training hours or speed limits to LEO fatality rates. While
these are theoretically appropriate across the literature, they have not been applied in this
manner previously. Because of this, the construct validity of this study cannot be taken
for granted.
Finally, this research design has limitations. Its focus on state-level factors
necessarily limits its scope. As described above, many individual-level factors contribute
to collision rates. As such, this design cannot—even with the best concepts, measures,
and data—explain all variability among states. While many individual-level factors are
considered as they are influenced by state-level factors, this is a macro study that offers
broad, state-level findings. Other limitations are the known restrictions of statistical
analysis available given a modest number of observations with many non-events (i.e.,
zeros—states without any LEO traffic fatalities in a given year). These are primarily
issues of data, time, and LEO experience and the expectation is that this study provides a
foundation for continued, future research that can expand and enhance the knowledge
gained here.
Summary
This chapter has detailed the research question, sub-questions, propositions, and
foundational rationale. It has also detailed the type, source, and description of data used
in this study, as well as the analytical methods and techniques. Finally, it has reviewed
67
the known limitations and provided a basis for the validity of the research. Next, Chapter
4 provides the first scholarly descriptive statistical analysis of LEO traffic fatalities by
state. Thereafter, Chapter 5 provides the quantitative and qualitative analyses described
in this chapter.
68
CHAPTER IV
DESCRIPTIVE ANALYSIS
This chapter describes in practically useful statistical detail LEO traffic deaths by
state for the study period 1995-2009. Where previous studies have examined LEO traffic
deaths nationally as a homogenous population (i.e., NHTSA, 2010b), this analysis
explores LEO traffic deaths in heterogeneous groups (differentiated by state). To begin,
all states are described in terms of counts, rates, means, standard deviations, and
minimum and maximum values on the variables of interest. Next, states are ranked best
and worst on key variables in terms of counts and rates.
Basic Descriptive Analysis
The variables used to describe states and fatal LEO traffic collisions in this
chapter come from the sources described in Chapter 3. Table IV.1 summarizes the
variables used here as a whole. Specifically, LEO Deaths refers to LEO deaths by state
by year (1995-2009), resulting in 765 observations (Obs). The Mean is 1.10, that is on
average each state (and the District of Columbia) had 1.1 officer fatalities per year. The
Minimum (Min) is 0, meaning that the lowest observed number of LEO deaths in a single
state was zero. The Maximum (Max) is 10, meaning that the highest observed number of
LEO deaths in a single state was 10. The Standard Deviation (Std. Dev.) is 1.63. Since
the standard deviation is significantly greater than the mean, it is an indication that the
data have significant variability—the premise for this research. As a rule, 99.7 percent of
the data should fall within 6 standard deviations (Weiss, 2008). In a normal distribution
these standard deviations would fall equally above and below the mean. Though as
previously noted, traffic collisions do not follow a normal distribution. Rather, they
69
follow a Poisson distribution, which is skewed to the right, and bounded by a discrete
zero on the left (i.e., there cannot be less than zero traffic collisions) (Weiss, 2008). Half
of the data still fall below the mean; however, in this study approximately half of the
observations are zero (i.e., of the 765 observations of 51 states over 15 years, 374 were
instances where there were no LEO deaths). Figure IV.1 illustrates the distribution of the
variable LEO Deaths.
Figure IV.1 Histogram of Total LEO Traffic Fatalities 1995-2009.
Total LEO Deaths (TLD) refers to LEO deaths summed by state (i.e., 51
observations) for the entire study period 1995-2009 and show a mean of 16.47, a standard
deviation of 18.46, and minimum and maximum values of 0 and 86, respectively. LEO-
Involved Fatals (LEO Inv. Fat.) is a variable described in Chapter 3 that includes not just
LEOs, but also non-LEOs killed in a collision with a LEO. Maximum LEO-Involved
Fatals (Max LEO Inv. Fat.) is a logic-generated variable that reports the highest number
between the variables LEO Deaths and LEO-Involved Fatals. This variable (Maximum
LEO-Involved Fatals) is used extensively in Chapter 5 to increase statistical power for
050
100
150
200
250
300
350
400
Freq
uenc
y
0 2 4 6 8 10LEO Deaths
70
modeling state-level differences. LEO Rate (LR) reports the mean 1995-2009 LEO
traffic fatality rate per 100,000 LEOs and is calculated by dividing the 15-year total
number of LEOs killed in traffic (840) by the 15-year LEO population of the US
(9,867,944) and multiplying by 100,000. General Public Rate (GPR) reports the mean
1995-2009 general public traffic fatality rate per 100,000 population and is calculated by
dividing the 15-year total number of people killed in traffic (619,992) by the 15-year
population of the US (4,038,549,298) and multiplying by 100,000. Balance is a ratio
variable that is calculated by dividing the LEO Rate by the General Public Rate. Balance
has a mean of .55, which indicates that on average, the LEO Rate is just over half of the
General Public Rate.
Table IV.1 Descriptive Statistics of Comparative State Variables 1995-2009.
Variable Obs. Mean Std. Dev. Min Max
LEO Deaths 765 1.10 1.63 0 10
TLD 51 16.47 18.46 0 86
LEO Inv. Fat. 765 2.22 3.10 0 20
Max LEO Inv. Fat. 765 2.42 3.09 0 20
LR 51 8.51 6.09 0 25.47
GPR 51 15.35 6.60 6.87 36.28
Balance 51 0.55 0.26 0 1.37
Table IV.2 provides the 15-year mean, standard deviation, total LEO deaths, and
other descriptive statistics for each state. This table provides a great deal of information
about how each state experiences LEO deaths. California has a mean of 5.73 LEO deaths
71
per year and an annual minimum and maximum of 2 and 9 LEO deaths, respectively. By
comparison, New Hampshire has a mean, standard deviation, minimum, maximum, and
total LEO deaths of zero since there were no LEO traffic fatalities from 1995 to 2009
(inclusive).
Table IV.2 Descriptive Statistics for LEO Traffic Fatalities by State 1995-2009.
State Mean Std Dev Min Max Total
Alabama 1.73 1.22 0 4 26
Alaska 0.13 0.35 0 1 2
Arizona 1.27 1.10 0 4 19
Arkansas 1.20 1.82 0 7 18
California 5.73 2.15 2 9 86
Colorado 0.53 0.64 0 2 8
Connecticut 0.33 0.49 0 1 5
Delaware 0.20 0.41 0 1 3
District of Columbia 0.20 0.41 0 1 3
Florida 3.87 2.67 1 8 58
Georgia 2.27 1.33 0 5 34
Hawaii 0.40 0.51 0 1 6
Idaho 0.27 0.59 0 2 4
Illinois 1.40 1.72 0 6 21
Indiana 1.73 0.96 0 3 26
Iowa 0.13 0.35 0 1 2
72
Table IV.2 (con’t.).
State Mean Std Dev Min Max Total
Kansas 0.27 0.59 0 2 4
Kentucky 0.53 0.64 0 2 8
Louisiana 2.27 2.02 0 8 34
Maine 0.13 0.35 0 1 2
Maryland 1.33 1.45 0 5 20
Massachusetts 0.80 0.86 0 2 12
Michigan 1.33 1.40 0 4 20
Minnesota 0.60 0.74 0 2 9
Mississippi 1.20 0.68 0 2 18
Missouri 1.67 1.35 0 4 25
Montana 0.40 0.63 0 2 6
Nebraska 0.13 0.35 0 1 2
Nevada 0.53 0.64 0 2 8
New Hampshire 0.00 0.00 0 0 0
New Jersey 1.27 0.88 0 3 19
New Mexico 0.73 1.10 0 4 11
New York 2.07 1.10 0 4 31
North Carolina 2.20 1.42 0 5 33
North Dakota 0.00 0.00 0 0 0
Ohio 1.20 1.15 0 4 18
73
Table IV.2 (con’t.).
State Mean Std Dev Min Max Total
Oklahoma 1.20 1.01 0 3 18
Oregon 0.47 0.74 0 2 7
Pennsylvania 1.40 1.24 0 5 21
Rhode Island 0.00 0.00 0 0 0
South Carolina 1.60 1.30 0 5 24
South Dakota 0.13 0.35 0 1 2
Tennessee 2.20 1.61 0 5 33
Texas 5.60 2.92 2 10 84
Utah 0.47 0.52 0 1 7
Vermont 0.13 0.52 0 2 2
Virginia 1.33 1.18 0 3 20
Washington 0.47 0.52 0 1 7
West Virginia 0.20 0.41 0 1 3
Wisconsin 0.60 0.83 0 3 9
Wyoming 0.13 0.35 0 1 2
Figure IV.2 illustrates the distribution of the variable Total LEO Deaths (TLD),
which sums the 1995-2009 deaths by state (i.e., 51 observations). The shape of this
distribution is similar to the distribution illustrated in Figure IV.1, which provided finer-
grain detail by de-aggregating years within states. Shown in this format, Figure IV.2
clearly illustrates the outliers. At the bottom of the distribution, New Hampshire, North
74
Dakota, and Rhode Island each had zero fatalities. At the top of the distribution,
California and Texas had 86 and 84 fatalities, respectively. This begins to further
illustrate the significant variability among states in terms of LEO traffic fatalities.
Figure IV.1, presents a near-classic Poisson distribution. The left side has a great
number of nonevents (i.e., state-years—a year of observation for a single state: n=765—
without a LEO death). The right side of the distribution becomes a razor-thin tail as it
shows few state-years where any state had six or more LEO deaths. Figure IV.2 also
illustrates a near-classic Poisson distribution. On its left side there are a few zeroes
before climbing to large numbers of observations. On its right side, however, there are
the two outliers (California and Texas). These two states account for more than 20
percent of the events (i.e., LEO deaths) and they feature at or near the 6th standard
deviation from the mean.
Figure IV.2 Histogram of Total LEO Traffic Fatalities by State 1995-2009.
Figure IV.3 reformats the information in Figure IV.2 and illustrates the total LEO
traffic fatalities by state (alphabetically) 1995-2009. This again provides a sense of the
02
46
8
Freq
uenc
y
0 10 20 30 40 50 60 70 80 90Total LEO Traffic Fatalities by State
75
distribution, but beyond variance, it allows for consideration of regional or state-specific
contexts. The limitation is that these counts say nothing about the population exposure.
For example, Ohio and Oklahoma have the same number of LEO deaths over the study
period (18), but lacking respective population details, not much can be determined.
Figure IV.3 Total Counts of LEO Traffic Fatalities by State 1995-2009.
0 10 20 30 40 50 60 70 80 90LEO Traffic Fatalities
WYOMINGWISCONSIN
WEST VIRGINIAWASHINGTON
VIRGINIAVERMONT
UTAHTEXAS
TENNESSEESOUTH DAKOTA
SOUTH CAROLINARHODE ISLANDPENNSYLVANIA
OREGONOKLAHOMA
OHIONORTH DAKOTA
NORTH CAROLINANEW YORK
NEW MEXICONEW JERSEY
NEW HAMPSHIRENEVADA
NEBRASKAMONTANAMISSOURI
MISSISSIPPIMINNESOTA
MICHIGANMASSACHUSETTS
MARYLANDMAINE
LOUISIANAKENTUCKY
KANSASIOWA
INDIANAILLINOIS
IDAHOHAWAII
GEORGIAFLORIDA
DISTRICT OF COLUMBIADELAWARE
CONNECTICUTCOLORADO
CALIFORNIAARKANSAS
ARIZONAALASKA
ALABAMA
76
Figure IV.4 provides this essential information by transforming the LEO death
counts into a mean rate (i.e., LEO deaths per 100,000 LEO population) for the study
period (1995-2009). With this additional information it becomes clear that Oklahoma’s
rate is nearly three times Ohio’s.
Figure IV.4 LEO Traffic Fatalities per 100,000 LEOs by State 1995-2009.
0 5 10 15 20 25Mean Rate of LEO Traffic Fatalities
WYOMINGWISCONSIN
WEST VIRGINIAWASHINGTON
VIRGINIAVERMONT
UTAHTEXAS
TENNESSEESOUTH DAKOTA
SOUTH CAROLINARHODE ISLANDPENNSYLVANIA
OREGONOKLAHOMA
OHIONORTH DAKOTA
NORTH CAROLINANEW YORK
NEW MEXICONEW JERSEY
NEW HAMPSHIRENEVADA
NEBRASKAMONTANAMISSOURI
MISSISSIPPIMINNESOTA
MICHIGANMASSACHUSETTS
MARYLANDMAINE
LOUISIANAKENTUCKY
KANSASIOWA
INDIANAILLINOIS
IDAHOHAWAII
GEORGIAFLORIDA
DISTRICT OF COLUMBIADELAWARE
CONNECTICUTCOLORADO
CALIFORNIAARKANSAS
ARIZONAALASKA
ALABAMA
77
Figure IV.5 illustrates the distribution of these mean 15-year rates by state (n=51)
in a histogram. The range spans 0 to more than 25 (LEO deaths per 100,000 LEOs)
beginning with the previously mentioned states with no LEO traffic deaths and
continuing up to Montana, which has the highest LEO death rate in the nation.
Figure IV.5 Histogram of Mean State LEO Traffic Fatality Rate 1995-2009.
Given this background on LEO death counts and rates, it next becomes instructive
to contextualize these numbers. This is most readily done using the general population
counts and rates. Intuitively one might expect that a LEO death rate in Nevada or New
York would correspond in some fashion with the general population rates in those
respective states. After all, LEOs and the general public drive on the same roads and in
the same weather, so many state-level factors are naturally controlled in this type of
within-state comparison. This statement does not assume that the rates would or should
be the same, but rather posits that there could be a relationship between the rates since
many other aspects of state-level variability do not factor in the within-state comparison.
010
2030
Per
cent
0 5 10 15 20 25LEO 15-year Mean Rate
78
Figure IV.6 illustrates the mean annual general public traffic fatalities by state for
the entire study period. Although this is a mean, in terms of comparative distribution,
this figure might be compared with Figure IV.3 (total LEO traffic fatalities).
Figure IV.6 Mean Annual General Public Traffic Fatalities by State 1995-2009.
0 1,000 2,000 3,000 4,000Mean Total Traffic Fatalities
WYOMINGWISCONSIN
WEST VIRGINIAWASHINGTON
VIRGINIAVERMONT
UTAHTEXAS
TENNESSEESOUTH DAKOTA
SOUTH CAROLINARHODE ISLANDPENNSYLVANIA
OREGONOKLAHOMA
OHIONORTH DAKOTA
NORTH CAROLINANEW YORK
NEW MEXICONEW JERSEY
NEW HAMPSHIRENEVADA
NEBRASKAMONTANAMISSOURI
MISSISSIPPIMINNESOTA
MICHIGANMASSACHUSETTS
MARYLANDMAINE
LOUISIANAKENTUCKY
KANSASIOWA
INDIANAILLINOIS
IDAHOHAWAII
GEORGIAFLORIDA
DISTRICT OF COLUMBIADELAWARE
CONNECTICUTCOLORADO
CALIFORNIAARKANSAS
ARIZONAALASKA
ALABAMA
79
Clearly, California and Texas have the most overall traffic fatalities. While it is
not as obvious, New Hampshire, North Dakota, and Rhode Island also appear to be on the
far left side of the graph, indicating generally lower numbers of overall traffic fatalities.
Next, Figure IV.7 converts counts to rates and illustrates the mean general public
traffic fatality rate by state for the study period (1995-2009). This graph is directly
comparable to the LEO version presented in Figure IV.4.
80
Figure IV.7 General Public Traffic Fatalities per 100,000 Residents by State 1995-2009.
Consistent with the previous comparison, Figure IV.8 illustrates the distribution
of these mean 15-year rates by state (n=51) in a histogram. The range spans
approximately 7 to more than 35 (general public deaths per 100,000 residents). With this
information it becomes clear that, overall, LEO traffic death rates are substantially lower
0 10 20 30 40Mean Rate of Total Traffic Fatalities
WYOMINGWISCONSIN
WEST VIRGINIAWASHINGTON
VIRGINIAVERMONT
UTAHTEXAS
TENNESSEESOUTH DAKOTA
SOUTH CAROLINARHODE ISLANDPENNSYLVANIA
OREGONOKLAHOMA
OHIONORTH DAKOTA
NORTH CAROLINANEW YORK
NEW MEXICONEW JERSEY
NEW HAMPSHIRENEVADA
NEBRASKAMONTANAMISSOURI
MISSISSIPPIMINNESOTA
MICHIGANMASSACHUSETTS
MARYLANDMAINE
LOUISIANAKENTUCKY
KANSASIOWA
INDIANAILLINOIS
IDAHOHAWAII
GEORGIAFLORIDA
DISTRICT OF COLUMBIADELAWARE
CONNECTICUTCOLORADO
CALIFORNIAARKANSAS
ARIZONAALASKA
ALABAMA
81
on average than general public traffic death rates. While approximately 80 percent of the
mean LEO traffic death rates fell below 15 deaths per 100,000 LEOs, approximately 55
percent of the mean general public traffic death rates fell above 15 deaths per 100,000
residents.
Figure IV.8 Histogram of Mean State General Public Fatality Rate 1995-2009.
This comparison leads to a question of balance—the ratio between LEO and
general public traffic death rates. Figure IV.9 illustrates this ratio balance with a stacked
horizontal bar graph. On the left side is the gernal public traffic death rate ratio and on
the right side in the LEO traffic death rate ratio. If the two sides are equal, then the ratio
would be 50:50 and indicate that the general public and LEOs die in traffic collisions at
the same rate. This is not the case. In all but two cases (Hawaii and Indiana) the general
public rate is greater than the LEO rate.
010
2030
Per
cent
0 10 20 30 40Public 15-year Mean Rate
82
Figure IV.9 Distribution of Traffic Fatality Rates by State 1995-2009.
Figure IV.10 illustrates the rate balance in a histogram where the balance is
calculated by dividing the overall mean LEO traffic death rate by the overall mean
general public traffic death rate. If they are equal, the result is one (i.e., a number divided
by itself equals one). If the LEO rate is greater (as in the cases of Hawaii and Indiana)
0 10 20 30 40 50 60 70 80 90 100Percent Distribution of Fatality Rates
WYOMINGWISCONSIN
WEST VIRGINIAWASHINGTON
VIRGINIAVERMONT
UTAHTEXAS
TENNESSEESOUTH DAKOTA
SOUTH CAROLINARHODE ISLANDPENNSYLVANIA
OREGONOKLAHOMA
OHIONORTH DAKOTA
NORTH CAROLINANEW YORK
NEW MEXICONEW JERSEY
NEW HAMPSHIRENEVADA
NEBRASKAMONTANAMISSOURI
MISSISSIPPIMINNESOTA
MICHIGANMASSACHUSETTS
MARYLANDMAINE
LOUISIANAKENTUCKY
KANSASIOWA
INDIANAILLINOIS
IDAHOHAWAII
GEORGIAFLORIDA
DISTRICT OF COLUMBIADELAWARE
CONNECTICUTCOLORADO
CALIFORNIAARKANSAS
ARIZONAALASKA
ALABAMA
Public LEOs
83
the number is greater than one. If the public rate is greater—as in the majority of cases
(i.e., 49 of 51)—the number is less than one.
Figure IV.10 Histogram of LEO to General Public Fatality Rate Balance 1995-2009.
Overall, these descriptive statistics illustrate the variability of LEO traffic deaths
among states and within states compared to the general population. They account for
exposure in terms of population through rates and time and intermittent fluctuations
through means. The next section draws comparisons among states across a number of
measures.
Ranking the States
There are several measures available to rank the states in terms of LEO traffic
fatalities. As detailed above, the most common measures are counts and rates. For the
purposes of this research rates are considered for the general and LEO populations, as
well as the balance or ratio between these two rates. When considering the rankings
presented in this section it is useful to review the 15-year nationwide means presented in
Table IV.1 as a base rate for comparison.
010
2030
40
Per
cent
0 .5 1 1.5Balance - LEO to Public Rate
84
Best States by Various Measures
The term best is highly subjective. Even so, when considering deaths, fewer—by
rate or count—is, without a doubt, better. Here I present four analyses, each with
different results for the best states. Each study ranks states on one of four variables: (a)
Total LEO Deaths (TLD), (b) LEO Rate (LR), (c) General Public Rate (GPR), and (d)
Balance (Bal). The first variable in each table is the ranking variable, however all four
variables are presented in each study.
First, Table IV.3 details the top 10 states by total number of LEO traffic fatalities
(TLD) for the study period 1995-2009. As previously mentioned, the least number of
fatalities are Rhode Island, North Dakota, and New Hampshire. The only variable
differing among these states is the General Public Rate.
Table IV.3 Top 10 States by Total Number of LEO Traffic Fatalities 1995-2009.
Rank State TLD LR GPR Bal
1 Rhode Island 0 0.00 7.69 0.00
1 North Dakota 0 0.00 16.41 0.00
1 New Hampshire 0 0.00 13.09 0.00
2 Vermont 2 13.67 24.56 0.56
2 Maine 2 6.16 14.31 0.43
2 Nebraska 2 4.04 16.02 0.25
2 Wyoming 2 10.52 32.68 0.32
2 South Dakota 2 10.53 21.94 0.48
2 Alaska 2 11.46 12.78 0.90
2 Iowa 2 2.67 15.19 0.18
85
Next, Table IV.4 details the top 10 states by lowest LEO Rate (LR) for the study
period 1995-2009. This again results in a three-way tie among New Hampshire, Rhode
Island, and North Dakota. However, some of the other top 10 states change in this
measure of best. Notably, New York—a state with a relatively large population and
significant number of total LEO deaths—has the third-place finish for overall lowest
LEO death rate.
Table IV.4 Top 10 States by Mean LEO Traffic Fatality Rate 1995-2009.
Rank State LR GPR Bal TLD
1 New Hampshire 0.00 13.09 0.00 0
1 Rhode Island 0.00 7.69 0.00 0
1 North Dakota 0.00 16.41 0.00 0
2 Iowa 2.67 15.19 0.18 2
3 New York 3.34 8.37 0.40 31
4 Illinois 3.94 10.86 0.36 21
5 Kansas 3.97 17.70 0.22 4
6 Nebraska 4.04 16.02 0.25 2
7 New Jersey 4.17 8.96 0.47 19
8 Connecticut 4.25 9.20 0.46 5
Next, Table IV.5 details the top 10 states by lowest General Public Rate (GPR)
for the study period 1995-2009. In this measure, Massachusetts—a state heretofore
unrepresented in the top 10—comes in first overall with a rate of 6.87 general public
traffic deaths per 100,000 residents.
86
Table IV.5 Top 10 States by Mean General Public Traffic Fatality Rate 1995-2009.
Rank State GPR Bal TLD LR
1 Massachusetts 6.87 0.72 12 4.93
2 Rhode Island 7.69 0.00 0 0.00
3 New York 8.37 0.40 31 3.34
4 District of Columbia 8.80 0.56 3 4.96
5 New Jersey 8.96 0.47 19 4.17
6 Connecticut 9.20 0.46 5 4.25
7 Washington 10.38 0.46 7 4.80
8 Hawaii 10.52 1.37 6 14.42
9 Illinois 10.86 0.36 21 3.94
10 Maryland 11.79 0.77 20 9.13
Last, Table IV.6 details the top 10 states by lowest overall Balance (Bal)—the
quotient of the LEO rate divided by the general public rate. This again results in the
three-way tie among Rhode Island, New Hampshire, and North Dakota. These first place
finishes may not be as meaningful one would expect. Because the LEO rate is zero, the
product of any balance calculation will be zero. In this case, it may be more interesting
and telling to look at the states in second and subsequent places where there is actually a
LEO rate (greater than zero) for a dividend. In this case, Pennsylvania—in eighth
place—may be most notable. Like New York, it is a sizeable state. With 21 total LEO
deaths, Pennsylvania still manages to have a LEO Rate (6.1) that is approximately one-
third of the General Public Rate (17.82).
87
Table IV.6 Top 10 States by Balance of Mean Traffic Fatality Rates 1995-2009.
Rank State Bal TLD LR GPR
1 Rhode Island 0.00 0 0.00 7.69
1 New Hampshire 0.00 0 0.00 13.09
1 North Dakota 0.00 0 0.00 16.41
2 Iowa 0.18 2 2.67 15.19
3 Kansas 0.22 4 3.97 17.70
4 Nebraska 0.25 2 4.04 16.02
5 West Virginia 0.28 3 6.34 22.55
6 Kentucky 0.32 8 7.01 21.79
7 Wyoming 0.32 2 10.52 32.68
8 Pennsylvania 0.34 21 6.10 17.82
Figure IV.11 summarizes the best states (alphabetically) overall that appeared in
one or more of these top 10 studies and graphically illustrates the 4 variables on a visual
scale. The Balance variable was transformed by multiplying it by 20 so that it would be
on a consistent scale for visual assessment. Arguments can be made for and against
different measures and it may be objectively impossible to identify a single winner. One
other factor worth noting is the variable Maximum LEO-Involved Fatals which features
prominently in the analyses in Chapter 5. As previously explained, that variable
identifies—in addition to LEO deaths—members of the general public killed in LEO
traffic collisions. Only one state in this research had no reported instances where a
member of the general public was killed in a collision with a LEO: New Hampshire.
88
Both North Dakota and Rhode Island each had one instance where a member of the
general public was killed in a LEO-involved collision.
Figure IV.11 Best States on One or More Mean Measures 1995-2009.
The next section provides contrast with the best states section which identified top
10 states in each of four analyses. It repeats these analyses, but looks at the bottom 10
states across each of the same four variables/measures.
0 5 10 15 20 25 30 35
WYOMINGWEST VIRGINIA
WASHINGTONVERMONT
SOUTH DAKOTARHODE ISLANDPENNSYLVANIA
NORTH DAKOTANEW YORK
NEW JERSEYNEW HAMPSHIRE
NEBRASKAMASSACHUSETTS
MARYLANDMAINE
KENTUCKYKANSAS
IOWAILLINOISHAWAII
DISTRICT OF COLUMBIACONNECTICUT
ALASKA
TLD LR
GPR Bal(*20)
89
Worst States by Various Measures
This section ranks the states from the bottom beginning with number 10—the
worst, and then working up the ranks (e.g., 9, 8, 7…) to identify the lesser of the worst.
As in the previous section it begins first with the total number of LEO traffic fatalities
(TLD). Table IV.7 details the bottom 10 states by highest total LEO deaths for the study
period 1995-2009.
Table IV.7 Bottom 10 States by Total Number of LEO Traffic Fatalities 1995-2009.
Rank State TLD LR GPR Bal
10 California 86 7.95 12.81 0.62
9 Texas 84 11.76 16.46 0.71
8 Florida 58 9.66 18.92 0.51
7 Georgia 34 10.85 20.64 0.53
7 Louisiana 34 14.85 22.59 0.66
6 Tennessee 33 15.82 21.81 0.73
6 North Carolina 33 11.18 18.54 0.60
5 New York 31 3.34 8.37 0.40
4 Indiana 26 16.79 15.59 1.08
4 Alabama 26 17.41 24.82 0.70
Next, Table IV.8 details the bottom 10 states by highest LEO Rate (LR) for the
study period 1995-2009. This is an interesting and potentially significant statistic in this
description—particularly with these worst states. With the exception of Montana, all of
these states have more than 10 LEO deaths—enough to give some meaning to the rate.
90
Likewise, with the exception of Montana, all of these states are geographically located in
the lower, southern portion of the nation.
Table IV.8 Bottom 10 States by Mean LEO Traffic Fatality Rate 1995-2009.
Rank State LR GPR Bal TLD
10 Montana 25.47 26.38 0.97 6
9 Mississippi 23.90 36.28 0.66 18
8 Arkansas 22.83 23.70 0.96 18
7 New Mexico 19.37 26.97 0.72 11
6 Alabama 17.41 24.82 0.70 26
5 Oklahoma 17.02 21.30 0.80 18
4 Indiana 16.79 15.59 1.08 26
3 South Carolina 16.69 25.45 0.66 24
2 Tennessee 15.82 21.81 0.73 33
1 Louisiana 14.85 22.59 0.66 34
Table IV.9 details the bottom 10 states by highest General Public Rate (GPR) for
the study period 1995-2009. In this study, Mississippi comes in last (i.e., worst) overall
with a rate of 36.28 general public traffic deaths per 100,000 residents. This is more than
twice the national average for the study period. At this rate, a resident of Mississippi is
500 percent more likely to die in a traffic collision than a resident of Massachusetts.
91
Table IV.9 Bottom 10 States by Mean Public Traffic Fatality Rate 1995-2009.
Rank State GPR Bal TLD LR
10 Mississippi 36.28 0.66 18 23.90
9 Wyoming 32.68 0.32 2 10.52
8 New Mexico 26.97 0.72 11 19.37
7 Montana 26.38 0.97 6 25.47
6 South Carolina 25.45 0.66 24 16.69
5 Alabama 24.82 0.70 26 17.41
4 Vermont 24.56 0.56 2 13.67
3 Arkansas 23.70 0.96 18 22.83
2 Louisiana 22.59 0.66 34 14.85
1 West Virginia 22.55 0.28 3 6.34
Last, Table IV.10 details the bottom 10 states by highest overall Balance (Bal)—
the quotient of the LEO rate divided by the general public rate. In this study, Hawaii
features as the worst overall with a Balance of 1.37, meaning that the LEO traffic death
rate is 37 percent higher than the general public traffic death rate. Hawaii’s relatively
small LEO population makes this rate particularly sensitive to even one additional or
fewer death. However, at second from the bottom and with sizeable populations,
Indiana’s Balance of 1.08 is more remarkable. In this case, LEOs die in traffic collisions
at a rate eight percent higher than the general public. While the rate itself (16.79) is
much lower than many other states, it raises questions about why it is comparatively so
high.
92
It is also notable that Massachusetts makes this list as the least bad of the worst
(i.e., 10th from the bottom), since it was noted for its very low rate elsewhere. In this
case, with a Balance of .72, LEOs are killed in traffic at nearly three-quarters of the rate
the general public is killed in traffic. Referring back to Table IV.1, the mean Balance for
the nation as a whole is .55. This study highlights the distance between rates. So while
Massachusetts has low rates overall in both categories, a modest difference between two
low rates appears rather large.
Table IV.10 Bottom 10 States by Balance of Mean Traffic Fatality Rates 1995-2009.
Rank State Bal TLD LR GPR
10 Hawaii 1.37 6 14.42 10.52
9 Indiana 1.08 26 16.79 15.59
8 Montana 0.97 6 25.47 26.38
7 Arkansas 0.96 18 22.83 23.70
6 Alaska 0.90 2 11.46 12.78
5 Oklahoma 0.80 18 17.02 21.30
4 Utah 0.78 7 10.54 13.52
3 Maryland 0.77 20 9.13 11.79
2 Tennessee 0.73 33 15.82 21.81
1 Massachusetts 0.72 12 4.93 6.87
Finally, consistent with the previous section, Figure IV.12 summarizes the worst
states (alphabetically) overall that appeared in one or more of these bottom 10 studies.
The figure provides a visual scale of the 4 variables. The Balance variable was again
93
transformed by multiplying it by 20 so that it would be on a consistent scale for visual
assessment. As before, arguments can be made for and against different measures and it
is difficult to identify a single loser.
Figure IV.12 Worst States on One or More Mean Measures 1995-2009.
Taken together, these analyses provide measures that reasonably identify areas of
concern for states. Death rates more than twice the national average—whether LEOs or
the general public—are arguably worthy of careful consideration. Instances where LEOs
are being killed at rates near that of the general public—counter to the national trend—
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
WYOMINGWEST VIRGINIA
VERMONTUTAH
TEXASTENNESSEE
SOUTH CAROLINAOKLAHOMA
NORTH CAROLINANEW YORK
NEW MEXICOMONTANA
MISSISSIPPIMASSACHUSETTS
MARYLANDLOUISIANA
INDIANAHAWAII
GEORGIAFLORIDA
CALIFORNIAARKANSAS
ALASKAALABAMA
TLD LR
GPR Bal(*20)
94
also provide reasonable grounds for further assessment. That said, a simple ranking on
any one measure likely oversimplifies the complexities of the many factors involved.
Summary and Next Steps
This chapter provided a basic statistical description of LEO traffic fatalities by
state for the period 1995-2009 and several comparative measures. Chapter 5 takes this
baseline information and utilizes more sophisticated statistical analyses and additional
variables to model LEO traffic deaths at the state level. It then introduces quantitative
and qualitative survey data from expert practitioners across the US representing 29 of the
51 cases (i.e., states) in this research. These combined data and analyses provide a more
substantive level of explanation for the observed differences and variability reported in
this chapter.
95
CHAPTER V
QUANTITATIVE & QUALITATIVE ANALYSIS
This chapter represents the heart of this research as it offers analytical evidence in
support of the observations, descriptions, and propositions detailed in earlier chapters. It
has two primary components. One, it develops a statistical model to analyze the effects
of certain measureable state-level differences in LEO traffic fatality rates. Two, it uses
qualitative CalPOST (2012) survey data from national law enforcement policymakers to
contextualize, temper, and potentially explain the observed state-level differences.
Statistical Modeling
As described in Chapter 3, this analysis uses a random effects Poisson regression
model to investigate state-level differences in LEO traffic fatalities. All modeling was
done utilizing the Stata/IC 11.2 software package. Where appropriate I will detail the
particular specifications of my programming syntax to facilitate replication of the results.
Preparing the Data
Several steps are involved in preparing a rational statistical model like the one
used here. Data preparation is an important step. The various counts (e.g., traffic deaths
and population) are discrete variables—they increase in whole integers. The rates and
means discussed in Chapter 4 are continuous—there are an infinite number of divisions
(i.e., decimal places) between each observed and theoretical value. Still other variables
are more complex. Speed (i.e., miles per hour (MPH)), for example, is a continuous
variable, but the Speed Limit used in this analysis is discrete. While monetary variables
are truly discrete (e.g., a dollar can only be divided into 100 parts), they are sometimes
treated as continuous variables—especially in larger quantities (e.g., state highway
96
receipts in billions of dollars). These variables are instrumental to the selection of the
model. As already described, the Poisson distribution is for discrete counts. The
outcome (i.e., dependent) variable must be discrete for any result from a Poisson model
to be valid. This means, for example, that the rates and means discussed in Chapter 4
would not be appropriate outcomes to model using Poisson. The raw counts are required
for proper modeling.
Sometimes discrete variables can complicate a model and lessen the significance
of an analysis when there are too few observations or there are large clusters with a few
dispersed observations. There are two examples of this in this data—Maximum Speed
and Driver Training Hours—Figures V.1 and V.2 illustrate the distribution of these
variables, respectively.
Figure V.1 Histogram of Maximum Daytime Speed Limit.
010
2030
40
Per
cent
55 60 65 70 75Maximum Speed Limit MPH
97
Figure V.2 Histogram of Minimum Driver Training Hours.
To address this clustering and dispersion issue for the speed variable, speeds
below 65 MPH are grouped together (Stata syntax: recode speed 55/60=65). This has the
effect of creating a comparison among states with one of three speed limit categories,
rather than comparing five speed limits, wherein two are not significantly represented.
Table V.1 details the new frequency and percent of the grouped variable.
Table V.1 Frequency of Grouped Maximum Daytime Speed Limit.
MPH Frequency Percent Cumulative
55-65 335 43.79 43.79
70 255 33.33 77.12
75 175 22.88 100
For the driver training variable, I addressed the clustering and dispersion issue by
converting the discrete variable into a categorical variable and treating clusters as groups
or categories (Stata syntax: recode dtrain 0/15=1 16/21=2 24=3 25/39=4 40=5 41/99=6).
010
2030
40
Per
cent
0 20 40 60Minimum Driver Training Hours
98
There are many states that have a requirement of zero driver training hours. Then there
are several states that require somewhere between 16 and 21 hours. This pattern
continues up to a maximum requirement of 60 hours. When working with speed, it is
clear that 60 MPH is the same in each state. When working with training hours, there is
no expectation of consistency because of a lack of the data to describe the content and
quality of the training hours. So beyond the clustering and dispersion, it makes little
sense to design the analysis such that the model assumes that the states requiring 60 hours
of training are getting 300 percent of what the states requiring 20 hours of training get.
Rather, it makes more sense to design the model such that it accounts for states that do
not require this training, states that require some training, and states that require more
training. This has the effect of creating a comparison among states with one of several
(six) categories of training requirements, rather than attempting a comparison of actual
numbers. Table V.2 details the frequency and percent of each category of required driver
training.
Table V.2 Frequency of Categories of Driver Training Hours.
Category Frequency Percent Cumulative
0-15
16-21
24
25-39
40
41+
155
60
75
55
210
75
24.60
9.52
11.90
8.73
33.33
11.90
24.60
34.13
46.03
54.76
88.10
100.00
99
Notably, nine states failed to report a number for this variable (i.e., minimum
required driver training hours) to IADLEST (2005), which creates missing data and
reduces the sample size for analysis. In my analyses, I developed models both with the
data missing and by recoding missing data into the zero category. This second procedure
was premised on the assumption that states that did not report a number might not
regulate driver training, which is in effect a requirement of zero hours. In this case,
recoding these missing data weakened the model and may also have misclassified them
(i.e., incorrectly placing them in the zero category), so ultimately the changes were not
included.
Building the Model
Before I had any formal training in statistical modeling I had the uninformed
belief that somehow a person could just enter all the data into a program or system, push
a button or two, and then the answer would be produced. In retrospect, this was a
severely naïve and wishful perspective. Even so, now, after the benefit of several courses
and training, I, as a non-statistician social scientist, still hold out hope that the variables I
have carefully selected based on theory and prior research studies will all—each and
everyone—show a statistically significant impact in my model. This might be like a
person hoping they will win a lottery or raffle contest with tens of thousands of ticket
holders. It could happen, but it rarely does. Such is the case with this model.
Over a period of approximately three months I estimate that I developed 200-300
model variations. These variations included many types of common—and occasionally
creative—manipulation. Variables were included and excluded, transformed, and parsed.
In some instances, I developed models that included and excluded monetary variables
100
relative to highway spending. I also developed models that only included the largest
states that had consistent LEO deaths year after year. I also included and excluded
different assumptions relative to exposure (e.g., state population versus LEO population)
and time (e.g., years as discrete single-increment spans versus transformations of years
through division and squaring). I also changed the design of the model (e.g., simple
linear regression versus Poisson regression versus random effects Poisson regression).
This process of model manipulation is important and worth noting. Commonly,
statistical packages like Stata provide automatic model selection. The researcher simply
inputs the data and the software package conducts its own analysis and chooses a model
that fits. Beyond this, automatic stepwise regression has become a popular technique for
allowing the software to select variables. The software package runs models (in a
manner not unlike I personally did) and selects the best predictor variables based on a
brute force approach to modeling. The problem is that the researcher is not involved in
this automated process. The process is not informed by theory or prior literature. As a
result, the researcher takes a great deal for granted in trusting that the software package
has made appropriate choices. This type of automated analysis has been criticized for
more than two decades (Derksen & Keselman, 1992). Beyond the moral high ground, I
argue that there is tremendous value in working through the variables to actually see the
impact of inclusion, exclusion, and manipulation. Through this process the researcher
comes to know and understand the model.
The Gustafson Model. The final model presented here provides at least two
significant findings. It does not support all the propositions laid out in Chapter 3.
Though, a lack of support is not the same as countermanding or otherwise disproving the
101
proposition. The propositions may all be true, but not verifiable given the available data
and analysis. I discuss this further in Chapters 6 and 7 and call for future research. Here
I present the available evidence based on the current data.
What I am calling the Gustafson Model is a random effects Poisson regression
model (and this type of model is alternatively referred to as cross-sectional time series or
repeated-measures). It uses the variable Maximum LEO-Involved Fatals (which is the
greater of the two variables LEO Deaths and LEO-Involved Fatals in each observation in
the Gustafson Dataset) as the dependent or outcome variable. The independent or
predictor variables from the Gustafson Dataset are Maximum MPH Speed Limit (grouped
as described previously), Total Traffic Fatalities, Minimum Hours of Driver Training
(recoded as described previously), and Highway Safety/Enforcement Spending.
Population of LEOs is used as the exposure variable and State is used as the grouping
variable. Given this structure, I executed the model (Stata syntax: xtpoisson
maxleoinvfatals i.speed ttfatals i.dtrain enforce, exp(leopop) i(state) irr nolog normal).
The syntax instructs Stata to represent specific categorical variables (noted with an i. in
the syntax) using two or more binary indicator variables in the model. Another notation
(noted as i(state) in the syntax) tells Stata that the collision counts from each state are
grouped or clustered (i.e., that repeated counts from one state are more similar to each
other than they are to counts from other states). The model uses an extra regression term,
a so-called random effect, to estimate the variance due to this clustering and permits the
correct estimation of model coefficient standard errors. Using a standard Poisson
regression without this random effect would produce grossly inaccurate coefficient
standard errors. The results are presented in Table V.3 and discussed in the next section.
102
Table V.3 Results of the Gustafson Model.
Predictor IRR P-Value 95% Confidence Interval
Max Speed
70 1.42257 0.032 1.031176 1.962523
75 1.641206 0.005 1.162477 2.317084
TT Fatals 1.000206 0.012 1.000045 1.000367
D Training
2 .6394005 0.084 .3849968 1.061913
3 1.004694 0.984 .6394463 1.57857
4 .7102082 0.143 .4494188 1.122329
5 .9384354 0.714 .6684123 1.317542
6 .9372467 0.784 .5897664 1.489457
Enforce S .9999992 0.000 .9999988 .9999996
Results of the Gustafson Model. The results presented in Table V.3 provide an
estimation of the impact of a one-unit change in the predictor variable on the outcome
variable (per LEO fatality rate) as an Incidence Rate Ratio (IRR in the table). A
Probability (P-value in the table) of attaining the observed incidence rate ratio due to
sampling error alone is reported. Finally, the upper and lower limits of a 95 percent
confidence interval are provided, which indicates (with 95 percent certainty) the range
within which the true incidence rate ratio exists). In the variables that are grouped (i.e.,
speed and driver training), the first category is treated as the baseline. Thus, the second
103
and subsequent categories provide test statistics to indicate their influence on the outcome
variable relative to the baseline group.
First the results for Maximum MPH Speed Limit are presented. The baseline is
considered to be states with a maximum speed limit of 65 MPH or less. The first result
reports that states with a maximum speed limit of 70 MPH (a one-unit increase over the
baseline) experience a 42 percent higher rate of Maximum LEO-Involved Fatals. The
second result reports that the next unit increase (i.e., states with a maximum speed limit
of 75 MPH) results in a 64 percent higher rate of Maximum LEO-Involved Fatals over the
baseline. The probabilities for these statistics are both significant at the 95 percent level.
The next result is for Total Traffic Fatalities. The findings show for each
additional total traffic fatality in a state, the rate of Maximum LEO-Involved Fatals will
increase by 2/100s of a percent. This statistic is significant at the 95 percent level and
can roughly be interpreted as in increase of one LEO-involved traffic fatality for every
additional 5,000 total traffic fatalities in a state. Since no single state has ever
experienced 5,000 traffic fatalities in a year—let alone an increase of 5,000 traffic
fatalities—this finding appears to have no practical significance.
Next are the results for Minimum Hours of Driver Training. None of the results
are statistically significant at the 95 percent level. However, I left this variable in the
model for a number of reasons. First, according to the literature and theory, training
should be a predictor. Second, as described above, there were missing data (nearly 20
percent), which may reasonably have impacted the results. Third, the first category was
almost significant and in fact did result as significant in other models that did not
function as well overall. The model as designed now establishes zero hours of required
104
driver training as the baseline and the first category then becomes 16-21 hours of required
driving training. If this model has a probability that was three and half percent less, the
interpretation would be that states that require 16-21 minimum hours of driver training
will have a 44 percent lower rate of Maximum LEO-Involved Fatals.
Since this was not the case, I interpret the present results as indicating that
minimum required driver training hours does not appear to be a significant predictor of
state-level variability in LEO fatal collisions. This finding may result from incomplete
data, confounding of training standards with unmeasured differences among states, and/or
an inadequate number of observations. Minimum required driver training hours may
very well be a significant predicator (of collision involvement) in an officer or
department-level analysis within one or more states.
Last are the results for Highway Safety/Enforcement Spending. This variable is
reported in 1,000s of dollars. Therefore, it reports that for each additional $1,000 of
spending on highway safety and enforcement in a state, the rate of Maximum LEO-
Involved Fatals will decrease by 8/100,000,000s of a percent. This statistic is significant
at the 95 percent level and can roughly be interpreted as a decrease of one LEO-involved
fatal for each additional $1.3 billion in highway safety and enforcement spending in a
state. Given this interpretation, there does not appear to be any practical significance.
These results are discussed further in Chapters 6 and 7. The next section focuses
on the qualitative component of this research. The insights and opinions of expert
practitioners are analyzed in order to provide context and other variable information that
is not otherwise available in existing quantitative data.
105
Qualitative Analyses of Survey Responses
A total of 29 states responded to the CalPOST (2012) State-Level Differences in
Law Enforcement Officer Traffic Fatalities survey. Table V.4 indicates the specific states
and agencies that responded, and indicates whether the responding agency identified its
function as highway enforcement (HE), standards and training (ST), or both (B).
Table V.4 Responding States and Function of Responding Agency.
State Agency HE ST B
Arizona POST X
California Commission on POST X
Colorado POST Board X
Connecticut State Police X
Delaware State Police X
Florida Highway Patrol X
Georgia Department of Public Safety X
Hawaii Honolulu Police Department X
Iowa Iowa Law Enforcement Academy X
Idaho POST X
Indiana Indiana Law Enforcement Academy X
Kansas Kansas Law Enforcement Training Center X
Kansas Kansas Highway Patrol X
Massachusetts Massachusetts State Police X
Massachusetts Municipal Police Training Committee X
106
Table V.4 (con’t.).
State Agency HE ST B
Maryland Police & Corrections Training Commissions X
Maine State Police X
Minnesota POST Board X
Missouri POST X
North Dakota Highway Patrol X
Nebraska Nebraska Law Enforcement Training Center X
New Hampshire State Police X
New Hampshire Police Standards and Training Council X
Nevada Highway Patrol X
Nevada POST X
New York State Police X
Oregon Department of Public Safety Standards & Training X
South Dakota South Dakota Law Enforcement Training Academy X
Tennessee POST X
Tennessee Department of Safety and Homeland Security X
Texas Department of Public Safety X
Vermont State Police X
Washington* State Criminal Justice Training Commission X
West Virginia Division of Justice and Community Services X
Wyoming Highway Patrol X
107
Table V.4 (con’t.).
State Agency HE ST B
Wyoming POST X
*Did not answer questions (effective non-response)
It is reasonable to assume that either a highway enforcement agency or a
standards and training agency would be able to respond to the questions CalPOST posed
in the survey, although there could be different perspectives on the same issues (e.g., a
standards and training agency might place more value on training or a highway
enforcement agency might view general public collision rates as more or less related to
LEO collision rates—each perspective could be a function of the respondent’s proximity
to the issue). There is a well-balanced distribution of responses from each agency type
and there are several responses (a) from each agency type within the same state, and (b)
agencies that represent both functions. Figure V.3 illustrates the agency type/function
breakdown. An analysis of responses indicating other revealed that these agencies best
fit in the both category.
108
Figure V.3 Responses to CalPOST (2012) Question 2 (Agency Type).
The survey respondents also reflect a diversity of states. There are states from the
North, South, East, and West. There are states with more urban areas and states with
more rural areas. There are several of the largest and smallest states—both by land mass
and population. For these reasons, it appears to be a representative sample on most
measures. The one area that does not appear to be well-represented is that of states with
the highest LEO traffic fatality rates. Only 2 of the bottom 10 states responded to the
survey compared with 7 of the top 10 states (ranked by mean LEO traffic fatality rate).
Multiple Choice and Likert Responses
The CalPOST (2012) survey posed multiple choice yes/no questions and Likert
scale questions that asked respondents to gauge their agreement or disagreement to a
statement across a five-point scale. The CalPOST survey also included sections for
narrative comments to allow respondents to clarify their scale responses or provide
additional detail. An analysis of the multiple choice and Likert responses is provided
50.0%
33.3%
5.6%
11.1%
Please confirm if you are responding as a standards and training agency or as a highway enforcement agency:
Standards & Training
Highway Enforcement
Both
Other
109
here. Notations and analysis of the narrative comments are included in this section when
appropriate. The full narrative responses are analyzed in the next section.
To begin, Figure V.4 illustrates the ratio of state respondents indicating a state-
level requirement for in-service or refresher driver training. In-service or refresher driver
training is considered any training beyond the basic, entry-level training required to
initially become a LEO. In terms of a general public comparison, consider getting a
driver’s license. In most states a person is required to take a test (usually written and
behind-the-wheel driving) to initially be licensed for vehicle operation. Subsequently,
many states require a written test (and occasionally a driving test) periodically to ensure
that the licensed driver still has the requisite knowledge and/or skill to drive. This is the
type of question being asked of states—essentially whether or not they bring LEOs back
for follow up training.
As a note, because some states had two respondents in the CalPOST (2012)
survey data (i.e., one from the highway enforcement agency and another from the
standards and training agency), the ratios presented here must be interpreted as ratios of
respondents and not states. Since some states have two responses and others have just
one, the ratios will not consistently report a percentage of the 29 represented states. Even
so, with regard to this question, it is clear by the extreme nature of the ratio that most
states do not have a state-level requirement that LEOs receive in-service or refresher
driver training. Many respondents did comment that agencies typically engage in this
training, although it is not required.
110
Figure V.4 Responses to CalPOST (2012) Question 3 (In-Service Training).
There were five states that indicated that in-service or refresher driver training
was required: Colorado, Indiana, Minnesota, Nevada, and Tennessee. Based on my own
expert knowledge, California should have responded that in-service or refresher training
is required. Colorado noted that in-service training was required, but in the comments
section indicated it was not mandated. In the cases of Nevada and Tennessee, there were
two responses for each state, but they were not the same within each state. The standards
and training agency in Nevada reported that in-service training was not required while the
highway enforcement agency reported that it was. In Tennessee, the standards and
training agency reported that in-service training was required while the highway
enforcement agency reported that it was not. These sorts of discrepancies regarding a
stark yes or no question are troubling. There is no way to know if one or more of the
respondents was simply misinformed or if the question was misunderstood or understood
differently by different respondents. Many of the respondents indicating that in-service
14.3%
85.7%
Does your state mandate in-service (refresher) driver training for LEOs?
Yes No
111
training was not required noted that the training was commonly occurring, but not a state-
level requirement.
Question 4 of the CalPOST (2012) survey asked those states that indicated that in-
service or refresher driver training was required to specify how often, how much, and
what kind (of refresher driver training) was required. The responses varied, including a
frequency ranging from annually to every five years, amounts ranging from two to eight
hours, and types of training including classroom, behind-the-wheel, and at the discretion
of individual agencies within the state. This information, coupled with the discrepancies
noted above, suggests a wide variety of practices in place among states.
The next question (and several others) in the CalPOST (2012) survey asked
respondents to rate to three statements in terms of their agreement or disagreement on a
Likert scale. The scale ranged from (a) strongly disagree, (b) disagree, (c) neutral, (d)
agree, and (e) strongly agree. Throughout this section these responses are graphed on a
five-point scale ranging from one to five, with one corresponding to strongly disagree,
and continuing incrementally through five, representing strongly agree.
Specifically, CalPOST (2012) Question 5 made statements indicating that (a)
additional in-service or refresher driver training, (b) entry-level driver training, and (c)
basic training in general, contributes to lower LEO traffic fatality rates. Overall, there
was substantial agreement with each of these statements as shown in Figure V.5.
112
Figure V.5 Responses to CalPOST (2012) Question 5 (Driver Training).
The means responses to the statements about in-service, entry-level, and basic
driver training were 4.23, 4.14, and 3.97, respectively. A mean of 2 would indicate solid
disagreement with the statement whereas a mean of 4 indicates agreement. A mean
greater than 4 moves into the category of strong agreement. These responses show that,
overall, there is a strong belief that different types of driver training contribute to lower
LEO traffic fatality rates. Narrative comments associated with this question confirmed
that there was value in in-service refresher training and that most agencies the
respondents knew of were mandating it, but the mandate was just not present at the state-
level. There was a point of departure, however. There was a contrast in the comments
between respondents who emphasized skills training (e.g., how to drive a car fast) and
respondents who emphasized decision-making training (e.g., how to know when not to
drive fast).
The next question (6) in the survey asked if states had a mandatory seat belt law
for on-duty LEOs. Figure V.6 illustrates the responses to this question. While about
1.00 2.00 3.00 4.00 5.00
Additional "basic training" (hours) contributes to lower LEO traffic fatality rates.
Additional "entry-level driver training" (hours) contributes to lower LEO traffic fatality rates.
Additional "in-service (refresher) driver training" (hours) contributes to lower LEO
traffic fatality rates.
Please indicate your response to the following statements by selecting whether you: 1-Strongly Disagree, 2-Disagree, 3-are Neural, 4-Agree, 5-Strongly Agree
113
two-thirds of respondents indicated that their state does have a mandatory seat belt law
for on-duty LEOs, several made comments indicating that there were exemptions for law
enforcement personnel.
Figure V.6 Responses to CalPOST (2012) Question 6 (Mandatory Seat Belt Use).
These comments about exemptions and seat belt requirements highlighted
ideological differences. There was one line of responses that detail LEO exceptionalism
as it emphasized the exemption to a state’s mandatory seat belt law. There was another
clear rhetorical argument among three respondents that emphasized the applicability of
the seat belt law to everyone. Another notable trend emerged in this section where
respondents pointed to individual agency policy. This lack of consensus turned out to be
a recurring theme.
Question 7 in the survey asked whether or not states had a law prohibiting LEOs
from talking or texting on a cellular telephone while driving. The overall response was
mostly negative as illustrated in Figure V.7. Some states had a prohibition against
texting, but not talking. Many of the states that did have laws prohibiting LEOs from
34.3%
65.7%
Does your state have a "mandatory seatbelt law" that requires on-duty LEOs to wear a seatbelt?
No Yes
114
talking or texting also had exemptions for what was typically referred to as job-related
use of the cellular telephone. There were fewer narrative comments associated with this
response; however, the same three themes noted in the last question were present: (a)
LEOs have an exemption, (b) the law applies to everyone, and (c) agency policy decides
the issue.
Figure V.7 Responses to CalPOST (2012) Question 7 (LEO Talk/Text and Drive).
Question 8 inquired about speed limits on LEOs when operating as an emergency
vehicle (i.e., lights and sirens). The response to this question was largely negative. The
comments of those states that did indicate that they had some kind of speed limit
suggested that the limit was generally established through due regard for the general
public. There was no indication that any state had a specific limit on speed.
74.3%
25.7%
Does your state have a "cellphone law" that prohibits on-duty LEOs from talking on a cellphone and/or texting while driving?
No Yes
115
Figure V.8 Responses to CalPOST (2012) Question 8 (LEO Speed Limits).
In similar fashion, Question 9 asked if there were any state laws governing police
pursuits—requirements or prohibitions. Overall, the answers to this question were
predominantly negative. The comments associated with this question were similar to the
comments following the speed limit question. In general, those states with pursuit laws
had broad language about due caution. There were eight comments that referred to
agency policy as the level where this decision is made. One state (Minnesota) indicated
all agencies were required to adopt statewide emergency vehicle operations and pursuit
policies. Another state (California) indicated that agencies are required by state law to
have a pursuit policy, but agencies determine what the actual policy consists of
individually.
91.2%
8.8%
Does your state have any "speed limits" (laws, regulations, or policies) that restrict the overall speeds on-duty LEOs can drive when using lights and
sirens?
No Yes
116
Figure V.9 Responses to CalPOST (2012) Question 9 (LEO Pursuit Laws).
Responses to Question 10 of the CalPOST (2012) survey were the most uniform
of any single question. Just two states (New York and North Dakota) indicated that there
were any statewide limits on shift length, overtime, or secondary employment.
Figure V.10 Responses to CalPOST (2012) Question 10 (Shift Length/Overtime).
New York’s response is worthy of highlighting. The respondent noted that the
law had been in place for at least 30 years and set 3 standards. First, LEO secondary
77.1%
22.9%
Does your state have any "pursuit laws" (i.e., laws, regulations, or policies) that specify what LEOs can, cannot, or must do--or when they can or cannot
engage--in pursuits?
No Yes
94.3%
5.7%
Does your state have any "shift length/overtime or secondary employment laws" (i.e., laws, regulations, or policies) that specify or otherwise limit how
much LEOs can work?
No Yes
117
employment was limited to a maximum of 20 hours per week. Second, total shift length
(including overtime) was limited to 16 hours except in emergencies. Third, a minimum
of 8 hours was required between shifts. Two other themes emerged from the comments
regarding working hours. First, there were several notations about this being an issue of
department policy. Second, this was identified as a collective bargaining or contractual
work conditions issue. Each of these themes suggests a non-state-level perspective is
seen as the norm in many of the responding states.
Question 11 asked if states had a mandatory move over law. These laws typically
mandate that drivers approaching an emergency vehicle stopped on the shoulder of a
highway must move a lane away if possible (i.e., move over) or slow down in cases
where a lane change is not possible. Laws of this type are intended to reduce the
incidence of LEOs (and other emergency personnel) being struck in traffic—either as
pedestrians or while in a vehicle (writing a ticket, talking on the radio, etc.)—when
stopped on the roadway. Figure V.11 illustrates the responses to this question.
Figure V.11 Responses to CalPOST (2012) Question 11 (Move Over Laws).
5.9%
94.1%
Does your state have a "mandatory move over law" that requires motorists to slow and/or move over for emergency vehicles stopped on the side of the road?
No Yes
118
The predominantly affirmative responses (94.1 percent) to this question indicate
that mandatory move over laws have become common. Comments following up this
question indicated that most states have added this law in the last five years. There was
one response indicating the law was added in 2001, but none earlier than that. The trend
is decidedly a 21st Century legal change. While LEO pedestrian traffic fatalities account
for less than eight percent of the cases in this research (i.e., 66 out of 840), it is not
immediately clear how many LEOs are killed while sitting in vehicles stopped on the side
of the roadway since these are classified as vehicle collisions. This research did not
consider this distinction during data collection; however, it may be reasonable to estimate
LEO traffic deaths while stopped on the roadway as being about as frequent as pedestrian
deaths (i.e., approximately eight percent of overall LEO traffic-related fatalities).
Following this series of legal/regulatory questions, Question 12 of the CalPOST
(2012) survey asked respondents to agree or disagree (on the five-point scale previously
described) with six statements. These six statements indicated that (a) mandatory move
over laws, (b) shift length/overtime or secondary employment laws, (c) pursuit laws, (d)
speed limits, (e) cellular telephone laws, and (f) mandatory seat belt laws, each
contributes to lower LEO traffic fatality rates.
Figure V.12 illustrates the mean responses. Overall, respondents agreed with the
statement that mandatory seat belt laws contribute to lower LEO traffic fatality rates.
There were no disagreements and only one neutral response to this statement. Next,
respondents generally agreed with the statement regarding mandatory move over laws.
Again, there were no disagreements and three neutral responses to this statement. The
other four statements generated significantly more neutral statements and some
119
disagreement. Even so, most respondents continued to agree or strongly agree. So the
variability became apparent among the expert opinions as the majority of respondents
agreed or strongly agreed (with statements that laws relating to shift length/overtime,
pursuits, speed limits, and cellular telephones contribute to lower LEO traffic fatality
rates) while others were neutral or, in five cases, disagreed. There were too few
comments on these issues to identify a theme among respondents.
Figure V.12 Responses to CalPOST (2012) Question 12 (Laws/Regulation).
Table V.5 reports the frequency and mean of responses to each issue statement in
Question 12 (CalPOST, 2012). Again, the range of response options included strongly
disagree, disagree, neutral, agree, and strongly agree (indicated as SD, D, N, A, and SA,
1.00 2.00 3.00 4.00 5.00
"Mandatory seatbelt laws" (for on-duty LEOs) contribute to lower LEO traffic fatality rates.
"Cellphone laws" (prohibition for on-duty driving LEOs) contribute to lower LEO traffic
fatality rates.
"Speed limits" (for on-duty LEOs) contribute to lower LEO traffic fatality rates.
"Pursuit laws" contribute to lower LEO traffic fatality rates.
"Shift length/overtime or secondary employement laws" contribute to lower LEO
traffic fatality rates.
"Mandatory move over laws" contribute to lower LEO traffic fatality rates.
Please indicate your response to the following statements by selecting whether you: 1-Strongly Disagree, 2-Disagree, 3-are Neural, 4-Agree, 5-Strongly Agree
120
and weighted as 1, 2, 3, 4, and 5 respectively, in the table). Four of the five
disagreements came from the State of Wyoming where the highway enforcement agency
and the standards and training agency were uniform in their disagreements with the
statements that speed limits and pursuit laws contribute to lower LEO traffic fatality
rates. Indiana also disagreed with statement that cellular telephone laws (prohibiting
LEOs from talking/text while driving) contribute to lower LEO traffic fatality rates.
Table V.5 Frequency and Mean of Responses to CalPOST (2012) Question 12.
Issue SD D N A SA Mean
Mandatory Move Over 0 0 3 10 22 4.54
Shift Length/OT/Secondary Employment 0 0 12 19 4 3.77
Pursuit Laws 0 2 15 10 8 3.69
Speed Limits 0 2 14 13 6 3.66
Cellular Telephones 0 1 7 16 11 4.06
Mandatory Seat Belts 0 0 1 8 26 4.71
Question 13 from the CalPOST (2012) survey asked about overall agency budgets
for fiscal year 2011-2012. The responses covered a wide range from as little as $400,000
for West Virginia POST, to more than $677,000,000 for the New York State Police.
Beyond this broad budgetary range, it is clear that, overall and in general, highway
enforcement agencies have more personnel and larger budgets than standards and training
agencies. It is also clear, both from my personal industry knowledge and from some
narrative comments in the survey, different funding mechanisms are in place to support
both highway enforcement and standards and training budgets among states. For
121
example, California POST is funded by fines and penalties (e.g., some of the revenue
generated from a speeding ticket is directed to standards and training support for LEOs).
By contrast, the California Highway Patrol does not receive revenue from tickets (this is
likely to alleviate the appearance that the agency might bolster its funding by increasing
enforcement efforts). Additionally, the purpose of agency budgets varies significantly
among states. In California, POST reimburses agencies for training (i.e., a direct support
model). In other states the POST agency establishes training standards, but does not fund
the training. In still other states the POST agency actually provides the training (i.e., a
direct provision model). In terms of the highway enforcement agencies, there are also
broad differences in mission and scope. In Texas and Georgia, for example, the Highway
Patrol is a specific division of a larger Department of Public Safety, which has broad
functions beyond highway enforcement. In California and New York, the Highway
Patrol and State Police are standalone entities that have also have broader missions
(beyond highway enforcement), but not with the breadth of the departments of public
safety, not the specificity of a highway enforcement-only division. Because of all these
differences, it is difficult to compare agency budgets.
Question 14 of the CalPOST (2012) survey asked respondents to indicate their
level of agreement or disagreement with three statements regarding sufficiency of
funding and overall commitment to highway safety and enforcement. The frequencies
and mean responses are illustrated in Figure V.13. On average, there was general
agreement that funding for LEO driver training and highway enforcement was sufficient.
On the issue of each state’s commitment to highway safety, there was agreement
(trending toward strong agreement) that states are strongly committed.
122
Figure V.13 Responses to CalPOST (2012) Question 14 (Funding/Highway Safety).
This survey item is an example where the mean does not adequately reflect the
responses. Table V.6 reports the frequencies and means for each statement. The range
on issues of funding spans from strong disagreement to strong agreement. Similar to the
responses to the budget question, these responses suggest that there are different funding
dynamics among states. Comments related to this question focused on the need for
additional funding.
Table V.6 Frequency and Mean of Responses to CalPOST (2012) Question 14.
Issue SD D N A SA Mean
Sufficient Funding for Training 6 10 7 9 2 2.74
Sufficient Funding for Enforcement 2 3 7 18 4 3.56
Commitment to Highway Safety 0 0 1 15 18 4.50
Question 15 on the CalPOST (2012) survey asked respondents for a more
speculative response to three statements on scale that included No, Doubtful, Maybe,
0.00 1.00 2.00 3.00 4.00 5.00
My state has a strong commitment to highway
safety.
My state sufficiently funds highway/traffic
enforcement.
My state sufficiently funds LEO driver
training.
Please indicate your response to the following statements by selecting whether you: 1-Strongly Disagree, 2-Disagree, 3-are Neural, 4-Agree, 5-Strongly Agree
123
Probably, and Yes. The statements indicated that (a) there are other state-level factors not
referenced that contribute to LEO traffic fatality rates, (b) there is a relationship between
LEO and general public traffic fatality rates, and (c) the LEO traffic fatality rate is a
matter of chance. Mean responses to these statements are illustrated in Figure V.14.
Figure V.14 Responses to CalPOST (2012) Question 15 (Traffic Fatality Rates).
This is another area where there was a broad distribution of opinions that are not
adequately expressed by the mean response rates. Table V.7 reports the frequencies and
means for each statement. In this format it becomes clear that expert practitioners do not
agree about the presence of other state-level factors, the existence of a relationship
between LEO and general public traffic fatality rates, or simple chance as an explanation.
1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
The LEO traffic fatality rate in my state is a matter of chance.
The LEO and General Public traffic fatality rates in my state are
related.
There are other state-level factors that
contribute to LEO traffic fatality rates.
Please indicate your response to the following statements by selecting whether you would reply: 1-No, 2-Doubtful, 3-Maybe, 4-Probably, 5-Yes
124
Table V.7 Frequency and Mean of Responses to CalPOST (2012) Question 14.
Issue SD D N A SA Mean
Other Factors Contribute to LEO Rates 5 3 10 9 7 3.29
LEO and General Public Rates Are Related 10 10 4 4 6 2.59
LEO Traffic Fatality Rate is Chance 6 13 12 2 2 2.46
This portion (i.e., yes/no and multiple choice Likert scale questions) of the
CalPOST (2012) survey essentially forced qualitative opinions of state representatives
into categories of convenience. This is an understandable and necessary contrivance to
make sense of a broad range of experiences and insights. Optional comments associated
with these responses occasionally provided additional, more nuanced clarification of the
perspectives that were presented. Overall, this section has illustrated some of the basic
areas of agreement (e.g., the need for LEO driver training) and areas of disagreement
(e.g., the utility of regulation and the interpretation of traffic fatality rates) among law
enforcement leaders in 28 states (since Washington did not complete the survey).
The next section culls thematic trends from unbounded responses to open-ended
questions. Finer-grain details about how practitioners make sense of LEO traffic
fatalities at the state-level are revealed through content analysis. This adds a measure of
explanation to the agreements and disagreements described in the previous section. The
findings also point to some of the underlying issues that impact variability in LEO traffic
death rates among states.
125
Narrative Content Analysis
This section uses a form of constant comparative analysis (Glaser & Strauss,
1967; Strauss & Corbin, 1998) described in Chapter 3 to identify categories and themes
across the non-standardized, narrative responses. Questions 16 and 17 of the CalPOST
(2012) survey asked open-ended questions specifically requiring this form of narrative
response. Question 16 asked whether or not and why respondents understood there to be
a relationship between LEO and general public traffic fatality rates in their respective
states. Question 17 asked respondents if they thought there were other (policy-relevant)
state-level factors contributing to LEO traffic fatality rates that had not been addressed in
the survey and also invited general commentary on the issue.
Notably, the response rate dropped off in these two questions. Where there had
consistently been 34 or 35 responses to each question, there were 28 responses to
CalPOST (2012) Question 16 and 15 responses to Question 17. As such, the first point of
analysis is to identify states and agencies responding to these questions. This information
is presented in Table V.8
Table V.8 States and Agencies Responding to CalPOST (2012) Questions 16 and 17.
State Agency Q16 Q17
Arizona POST X
California Commission on POST X X
Colorado POST Board X X
Connecticut State Police X X
Delaware State Police X X
126
Table V.8 (con’t.).
State Agency Q16 Q17
Florida Highway Patrol
Georgia Department of Public Safety X X
Hawaii Honolulu Police Department X
Iowa Iowa Law Enforcement Academy
Idaho POST X X
Indiana Indiana Law Enforcement Academy X
Kansas Kansas Law Enforcement Training Center X X
Kansas Kansas Highway Patrol X X
Massachusetts Massachusetts State Police X X
Massachusetts Municipal Police Training Committee X X
Maryland Police & Corrections Training Commissions X
Maine State Police
Minnesota POST Board X X
Missouri POST X
North Dakota Highway Patrol X X
Nebraska Nebraska Law Enforcement Training Center X
New Hampshire State Police X X
New Hampshire Police Standards and Training Council X
Nevada Highway Patrol
Nevada POST X X
127
Table V.8 (con’t.).
State Agency Q16 Q17
New York State Police X X
Oregon Department of Public Safety Standards & Training X X
South Dakota South Dakota Law Enforcement Training Academy X X
Tennessee POST
Tennessee Department of Safety and Homeland Security X X
Texas Department of Public Safety X X
Vermont State Police X X
Washington State Criminal Justice Training Commission
West Virginia Division of Justice and Community Services
Wyoming Highway Patrol X
Wyoming POST
Overall, respondents appeared to take the opportunity presented in these two
open-ended questions to reply broadly to the issue of traffic fatalities—both among LEOs
and the general public. Some respondents wrote specifically to the questions of
differences between LEOs and the general public and state-level factors, but most seemed
to address the issue(s) they found most salient. Some of the narrative responses to
Questions 16 and 17 (CalPOST, 2012) were quite long (e.g., a half page of single-spaced
text). Many other responses were short (e.g., one or two sentences).
In order to utilize the largest amount of content available for this analysis, I
included comments from the earlier part of the survey in addition to the responses from
128
the last two open-ended questions. This approach provided me with more than 4,500
words of text to analyze. As described in Chapter 3, I read and reread this text and
identified categories and themes across the respondents. Table V.9 reports the 26
recurring categories identified among respondents.
Table V.9 Recurring Categories in CalPOST (2012) Survey Responses.
# Category
1 Accountability (within the agency and to the public)
2 Agency Policy (as the guiding standard)
3 Alcohol-Impaired Public (as a threat to LEOs on the roadway)
4 Basic Driver Training (as a foundation for LEO driving)
5 Decision-Making (in terms of deicing how to drive)
6 Differences (between LEOs and the public)
7 Discipline (in terms of managing LEO driving)
8 Distraction (for LEOs while driving)
9 Due Regard/Due Caution/Due Care (as a standard for safe driving)
10 Emergency Driving (as a routine for LEOs)
11 Exceptions And Exemptions (in terms of what does not apply to LEOs)
12 Fatigue (as a LEO-specific problem)
13 General Public Driving Culture (in terms of the environment LEOs drive in)
14 High Speed (as a routine for LEOs)
15 In-Service Driver Training (as a refresher for skills and decision-making)
16 Knowledge And Skill (as two components of driving)
129
Table V.9 (con’t.).
# Category
17 Lack Of Data (as a limit to knowing what matters for LEO driving)
18 Lack Of Funding (as a barrier to adequate training)
19 LEO Culture (as a factor in collision rates)
20 Mandatory Versus Voluntary (in terms of what is or should be a standard)
21 Pursuits (as a routine for LEOs)
22 Responding To Calls (as a routine for LEOs)
23 Roadway Conditions (as a collision factor for all drivers)
24 Seat Belts (as a requirement or option for LEOs)
25 Similarities (between LEOs and the general public)
26 Weather (as a collision factor for all drivers)
Given the narrative context in these categories, all 26 categories were integrated
into 1 or more of 4 broad themes: (a) LEO Exceptionalism, (b) Agency Sovereignty, (c)
Training, and (d) External Control Loci. For example, LEO Culture can be viewed (a) as
a function of LEO Exceptionalism in that it differentiates LEOs from the general public
and other subcultures. At the same time, LEO Culture can also be viewed (b) as a
function of Agency Sovereignty when a strong leader like Las Vegas Metropolitan Police
Department Sheriff Doug Gillespie takes steps to reform the culture (Alqadi, 2012;
Gustafson & Cappitelli, 2010) as described in Chapter 1. Finally, LEO Culture can be
viewed (c) as a function of Training as desired behaviors and attitudes become
enculturated through modeling and practice. This is one example of the framework of
130
thematic logic used to group the identified categories to move from a list of terms to
concepts for sense-making and, potentially, theory-building. Each of the themes is
discussed in the following subsections.
LEO Exceptionalism. The term exceptionalism is frequently used to draw
distinctions among nations (Fredrickson, 1995). Even so, conceptually it is valuable and
applicable to sub-national groups as a means to label and highlight the presence of actual,
imagined, or socially-constructed distinguishments. An example of actual
exceptionalism is the US military capacity. The US is distinguished from all other
nations in terms of its fighter jets, naval ships, weaponry, for example. Illustrations of
imagined exceptionalism can be found at schools and colleges around the world where
the home team is thought to be the best of the best, when in reality the home team (in
whatever manifestation it takes) almost certainly falls somewhere under the normal curve
in whatever measure is thought to be exceptional. Finally, examples of socially-
constructed exceptionalism are present in many places where bias or racism flourish.
Consider a workplace where women are paid less than men for no other difference than
gender. The difference in wages is real (distinct from something imagined); however, the
premise of the exceptionalism (i.e., an underlying belief that men are somehow better
than women) is socially-constructed.
The narrative comments from the survey respondents variously touched on each
of these conceptualizations of exceptionalism, some more directly than others. For
example, there were comments that suggested that the general public needs speed limits
for safety, but LEOs can just use due caution. The same arguments were applied to
cellular telephone use, suggesting that LEOs were somehow less impacted by distraction.
131
These appear to be imagined instances of exceptionalism that would put LEOs in a
different category than other humans in terms of the laws of physics or human
performance. In terms of actual exceptionalism, there were many statements noting that
LEOs have special training and performance expectations that both prepare them for, and
expose them to, more hazardous driving conditions. Finally, there were many statements
about legal exemptions whereby various laws, rules, and regulations are not applicable to
LEOs. These are clearly social constructions of exceptionalism.
LEOs on the roadway are undoubtedly different from other motorists. In some
instances the exceptionalism is both real and sensible. In other instances, it appears as
though LEOs, as related by these law enforcement leaders, use what is allowed (i.e., by
exemption) and treat is as what should be (i.e., as practice).
Agency Sovereignty. Agency sovereignty is the idea that each law enforcement
organization has the right to set its own standards and policies. It is the miracle of
federalism that allows subdivisions of government to act autonomously. Best practices,
validated evidence and findings, and the greater good of all can be set aside in favor of
local control and avoidance of regulatory oversight and restriction. This theme was
produced and reproduced throughout the survey responses. Quite simply there was a
propensity among respondents to downplay the role of the state. While the survey did not
mention the idea of national standards (i.e., federal law/regulation), it is certain it would
be almost totally objectionable to the respondents.
Training. Training was referenced by virtually all respondents as a critical
component to reducing LEO traffic fatality rates—conceived of in two bifurcations.
There was (a) the concept of basic, entry-level training to establish a requisite level of
132
competency, and (b) the idea of in-service or refresher training that would maintain LEO
competency over time on a recurrent basis. Next, there was the concept of (a) skills
training in the dynamics and performance of driving, and (b) decision-making training on
the judgment aspect of emergency responses and pursuits. The decision-making
component was powerfully advocated by a few respondents, and absent from the
comments of many others.
External Control Loci. Finally, and practically, are external control loci. These
outside influences are not so much matters of determinism as issues beyond the LEO or
the law enforcement agency. No respondent wrote in terms in inevitability, but many
made note of factors outside of their control or influence. These factors included issues
like weather and the general public, and occasionally funding. Several respondents
mentioned LEOs being killed in traffic by alcohol-impaired drivers. In many instances
there is nothing a LEO can do to avoid a fatal collision. As such, external control loci are
important realities to acknowledge. There is a balance between in recognizing areas of
opportunity for influence and these other areas where awareness may be all that is
possible. The respondents appeared clear and reasonable in this respect.
Summary and Next Steps
This chapter has reported quantitative and qualitative findings from the data. The
Gustafson model revealed statistically significant LEO traffic fatality predictor variables
relating to speed and highway safety/enforcement spending. The survey analysis
revealed several state-level trends in terms of training and regulation and the content
analysis identified themes relating to LEO exceptionalism, agency sovereignty, the
importance of training, and the reality of external control loci.
133
The next chapter takes these findings and provides a discussion and analysis in
terms of policy-relevant implications. Options for state-level change to reduce LEO
traffic fatality rates are identified. The identification of potential policy changes sets up
the last chapter, which provides a process model for implementation, as well as an agenda
for future research.
134
CHAPTER VI
DISCUSSION & POLICY IMPLICATIONS
This chapter discusses findings and identifies policy implications from this
research and the analyses reported in previous chapters. It brings together knowledge
gained from prior literature and research, as well as new knowledge from this original
research, to synthesize practical policy changes to reduce law enforcement officer (LEO)
traffic fatalities. This is an interpretive chapter that offers prospective state-level policy
options for law enforcement leaders and others concerned with law enforcement policy
relative to vehicle operations and training.
The first section begins with a summary of findings from this research and a brief
review of previous findings of relevance. The next section discusses specific
implications of the findings and identifies potential policy changes and adaptations to
reduce LEO traffic fatalities. Next is a candid analysis of anticipated challenges and
resistance to recommended policy initiatives including discussion of local control,
labor/management politics, and economic environments.
Review of Findings
This research set out to determine what accounts for different law enforcement
traffic fatality rates by state. The study further aimed to identify which of the identified
explanations are most amenable to policy change and practice adaptation. Finally, this
research sought to identify which policy and practice changes are most likely to reduce
LEO collision rates (and thereby fatalities) and to what extent the identified changes are
feasible given social and institutional complexities. These findings are presented in the
sections that follow.
135
Present Study
In many respects, the findings from this research are not so remarkable overall.
There was no breakthrough that identified a singular solution, but showed what was
suspected at the outset. The propositions were entirely practical and appreciably
intuitive. The first proposition simply posited that increased state investment in law
enforcement and highway safety would result in lower LEO traffic fatality rates. This
proposition was statistically validated and qualitatively supported by practitioner
commentary as reported in Chapter 5. Not surprisingly, funding helps to address large-
scale problems.
The second proposition posited that more stringent regulatory requirements
(relative to LEOs and highway safety) would result in lower LEO traffic fatality rates.
This proposition had many subordinate components that put forward various examples of
regulatory requirements for LEOs thought to be qualifying for this proposition: (a) driver
training, (b) mandatory seat belt use, (c) pursuit limitations, (d) speed limitations (either
for LEOs specifically or lower overall), (e) work hours restrictions, and (f) prohibitions
on distractions (i.e., cellular telephone use). Given the available data, only driver training
and overall state speed limits were tested statistically. Higher speed limits were shown to
be statistically significant predictors of higher LEO traffic fatality rates. Driver training
(as modeled) did not produce significant results statistically. The survey data from
practitioner experts across the nation did reveal support for each of these propositions to
varying degrees. There was overwhelming agreement on the efficacy of mandatory seat
belt use, while there was greater variability of opinions on the effects of limiting speeds.
136
In sum, there was little disagreement with these propositions. Where there was a lack of
agreement, there were mostly neutral responses.
So the findings are modest and predominantly simple. The real surprise in this
research is that the law enforcement community has done little with all the knowledge
that already exists. The most remarkable finding may be the degree of disconnect
between or among stated items of importance and actual policies, practices, and
requirements (i.e., the substance of the second proposition).
Apparently, among agencies there appears to be disbelief and resistance to
common facts. Desired training outcomes and actual training practices seem
incongruent. Risk versus gain policies and practices seem misaligned. There is a great
deal of rhetoric in the law enforcement community about getting to the destination safely,
but there seems to be a dearth of directive policy to make this happen. And while some
critics might argue that these policy needs are met at the local level, the fact remains that
essentially everything else that is considered universally critical for law enforcement is
prescribed at the state-level. So why should the primary factors (e.g., seat belt use, speed
limits, work hour maximums) that lead to the leading cause of death for LEOs (i.e.,
traffic fatalities) be left to local control? That is not how these issues are handled for the
general public or bus drivers or truckers. Cities and counties across the nation do not
establish rules for seat belt use or maximum speeds or maximum work hours on a case-
by-case basis. This approach would be ineffective. States establish these regulations in
order to standardize best practices and promote safety. The most notable finding in this
research is that the core problem of LEO traffic fatalities is more an issue of
implementation than one of knowledge. It is a political quandary and a problem of human
137
resistance. It is an issue of federalism and local control. It is a problem of information
sharing and networking—a failure to disseminate and adopt practices that are known to
be effective. It is an issue of will and prioritization. It is an issue of organizational
behavior. As such, it is an issue of public policy and as long as policymakers and their
constituents are amenable to the status quo, then the problem will persist.
Previous Scholarship Reprised
Speed kills; higher speed limits correlate with increased traffic fatalities
(Friedman, Barach, & Richter, 2007; Friedman, Hedeker, & Richter, 2009; Pant, Adhami,
& Niehaus, 1992). This has been studied repeatedly. When vehicles are stopped (i.e., the
absolute absence of speed) fatal traffic collisions are nonexistent. The faster vehicles
travel, the greater the rate of traffic fatalities. As a note, this year the State of Texas
decided to raise its maximum speed limit to 85 MPH on a 41-mile section of toll road
(“85-mph Speed,” 2012). In a recent conversation with an official from Utah, I learned
that the speed limit there was raised to 80 MPH since I began this research (Doug Larsen,
personal communication, October 9, 2012). These increases will have direct impacts in
terms of increased traffic collisions and fatalities.
State occupational (process) regulation serves the public interest and may improve
quality through continuing education requirements (Teske, 2004). Speaking to the issue
of purposeful regulation, Teske noted that “states that want greater professionalism in all
aspects ... should require it in both entry barriers and in maintenance requirements”
(Teske, 2004, p. 150). Applied to LEOs, this could be interpreted to mean that the
performance and discipline that is required of recruits should also be required of
incumbent, veteran officers. There is a vast literature on regulation and it is too often
138
conflated with market economics, when as administrative law, it serves non-market
purposes (Stewart, 1983), which many economists might say is the proper role of
government. Simply speaking, from human experience, the solution to most large scale
public issues—like educational standards, occupational licensing, wildfires, or water
quality—is (state) regulation. The primary take away from these findings is that
regulation is an appropriate and effective means to address systemic public problems like
LEO traffic fatalities.
Human beings can only focus attention on one thing at a time. Put another way,
multitasking is a myth (James & Vila, 2012; Loukopoulos, Dismukes, & Barshi, 2009;
Rosen, 2008). Some people may switch between tasks or refocus attention better than
others, but the operative words here are switch and refocus. Distraction is a challenge for
everyone who drives; for LEOs, it is compounded by the greater number of distractions
and demands for attention (e.g., radios, computers, scanners).
Seat belts save lives (Richens, Imrie, & Copas, 2000). Mandatory seat belt laws
(that have penalties) increase seat belt use and thereby save lives (Richens, Imrie, &
Copas, 2000; Wagenaar, Maybee, & Sullivan, 1988). As with speed, the effects of seat
belt use (and nonuse) have been studied repeatedly. The evidence is unambiguous.
Fatigue impacts driving performance and the more fatigued a driver is, the more
likely they will have a collision (Vila, 2000, 2006, 2009). Driving while tired is like
driving while alcohol-impaired (Williamson, Feyer, Mattick, Friswell, & Finlay-Brown,
2001). A person who has been awake for 20 hours has approximately the same
performance capacity as a person who is legally drunk (i.e., a .08 percent blood-alcohol
content) (Dawson & Reid, 1997). Socially, Americans have displayed diminishing
139
tolerance for drunk driving; fatigue has not received the same scrutiny. At the scene of a
fatal collision, however, dead is dead whether the cause was fatigue or alcohol.
Training improves performance—especially in highly complex tasks (Schneider,
1985). In order to drive well and maintain desired behaviors, drivers need training in
both skill development and cognitive behaviors, which may benefit from periodic
refresher training (Dorn, 2010; Dorn & Barker, 2005). Professionals regularly train and
practice their skills. LEOs should train and practice not only their psychomotor skills,
but also their (driving) decision-making skills. Training these skills is different from
everyday driving. An accountant balancing their checkbook would not consider it to be
training to perform an audit. Just because LEOs drive on a daily basis does not mean
they are honing their skills, learning good habits, or getting feedback on their
performance—important aspects of training.
Highway traffic enforcement reduces traffic fatalities (DeAngelo & Hansenyz,
2010). The premise is quite simple. Traffic laws are established to promote safety in
order to reduce injuries and fatalities. Enforcement of traffic laws in fact accomplishes
the intended result of reducing traffic fatalities. This suggests that there should be some
kind of enforcement for LEOs. This might come in the form of supervision or
management or strong leadership or in more enforcement-oriented terms like traffic
citations or disciplinary measures. Whatever the actual policy design, an accountability
system will support the desired outcome (i.e., traffic safety).
Policy Implications of Findings
The logical question to ask—and the point of this research—is simply this: given
the findings that have been presented, what are the state-level policy implications? While
140
there are many downstream implications, they all flow from the overarching implication
that, for the most part, current state-level policies and practices (e.g., laws, regulations,
standards) do not comport with known methods to reduce LEO traffic fatalities. Frankly,
current policies ignore the facts. In more politic language, one might say that current
policies are positioned with respect to cultural and social norms as opposed to a
paternalistic position of oversight and enforcement.
Carried further, this quagmire reduces to a relatively dichotomous ideological
debate. When this sort of debate comes about with regard to the general public, it makes
some reasonable sense. This country was founded on the ideal of individual freedom. So
while the government may know that cigarette smoking can lead to cancer and other
health problems and the likely possibility of an untimely, unpleasant, and costly death for
an individual, the American ethos of civil liberties allows the individual to choose their
own path. The government adopts an informative role to advise individuals of risks, but
ultimately leaves choice to the individual. When the question is expanded beyond the
individual to include groups of individuals or entire populations, the debate changes
somewhat and there is frequently more agreement about the role of government to protect
groups from the choices of individuals. Using the cigarette smoking example, while
individual choice is preserved, there are restrictions on place to protect members of the
public who might not consent to the exposure.
There is a tremendous literature and debate regarding these issues of liberty and
the role of government dating from 350 BCE with Aristotle (1943), to modern times with
Dewey (1927), to the latest volume of most any political science or public affairs journal.
This inquiry is outside the scope of this research. What is worth noting is that in terms of
141
government regulating itself or its own agents, there has not been a great deal of
controversy. This is to say that a person who freely chooses to work for the government
would not be considered oppressed when required to follow its established rules. Further,
there has been little controversy around the idea of the government establishing rules to
protect its workforce or ensure the health and safety of government workers. No one
cries fowl when firefighters are required to wear turnout coats when battling a fire or
military pilots are required to sleep before flying a mission. Requirements of these types
are intrinsically understood as necessary and responsible for the safe conduct of
operations.
Given this perspective, I outline a series of policy changes and adaptations that
comport with the findings detailed in this research and elsewhere with regard to
improving traffic safety and reducing fatal collisions for LEOs. After detailing these
policy proposals I follow up with a discussion of anticipated challenges and resistance to
their adoption. Finally, I highlight known examples where substantially comparable
policies have been enacted that serve as examples if not models for development.
Potential Changes and Adaptations
In this section I outline seven independent policy actions that can be implemented
at the state-level to reduce LEO traffic fatalities. I advocate them as a group and expect
that taken as a whole have complimentary effects. Although, each can improve LEO
traffic safety on its own and I encourage policymakers to start with what is feasible rather
than take an all or nothing approach. Incremental change is preferable to inaction. The
following are the recommendations in no specific order.
142
Slow Down. Limit the maximum speeds LEOs are allowed to drive. The Carbon
Motors police car referenced in Chapter 2 has an advertised top speed of greater than 150
MPH (Carbon Motors, 2011). Why? When would a policymaker ever want a LEO to
drive in excess of 150 MPH? A standard should be thoughtfully established and
enforced.
Perhaps an absolute maximum of 100 MPH should be set. Or perhaps the
standard should be something dynamic that recognizes changing environments. The
policy could dictate a LEO maximum that is 20 MPH over the speed limit posted on the
roadway (e.g., 45 MPH in a 25 MPH zone or 85 MPH in a 65 MPH zone). Alternatively,
the policy could be designed with performance expectations in mind. In some cases, it
may be appropriate that highway enforcement LEOs have a higher speed limit than
municipal police LEOs. In any event, exceptions can be specified. If any LEO is
actively pursuing someone who just committed murder, for example, the LEO might be
exempted from all speed limits. In this case, the meaning of the policy could be
interpreted as an indication that catching the murderer is more important than the safety
justifications that support the limit in the first place. These are values questions that
policymakers can address.
Create and Standardize Requirements. Specify how LEOs should drive in
specific situations. From emergency driving to pursuits to vehicle operations training,
LEOs (and the motoring public) will benefit from consistency and standardization.
Should LEOs respond with lights and sirens (i.e., emergency response) exempted from
speed limits and stop signs to a report of a burglary when there is no indication that a
suspect is in the area? There are some agencies that allow this type of response for any
143
felony report. What is gained versus what is risked with this type of response? Does
arriving at the scene 30 seconds or a minute sooner change the outcome enough that it
justifies the risk of collision associated with an emergency response?
What about pursuits? There was an entire television series dedicated to police
pursuits titled World’s Scariest Police Chases. The show also spawned a video game.
Does it make sense to chase everyone for any violation of law? This kind of policy
should not be decided on a community by community basis. High speed pursuits are a
form of group risk exposure that should be regulated and standardized because it creates
danger not just for the individual LEO (or even the fleeing suspect), but for the motoring
public at large.
Concentrate and Stay Focused. Limit what technologies and activities LEOs can
use and do while driving. Distractions are dangerous and many cannot be avoided in a
police car. But some distractions, like cellular telephones and particularly texting, can be
managed if not entirely avoided while driving. Emergency communications have been
delivered via radio and in-car computer for many years. It is highly unlikely that a LEO
needs to talk on a telephone while driving. It is almost unimaginable that a LEO would
have to text while driving. These activities are difficult and dangerous for humans—
LEOs included. Statewide policies should prohibit these activities. If exceptions need to
be specified, they should be narrow and specific as opposed to a blanket exemption for
on-duty LEOs.
Don’t Be A Dummy… Require LEOs to wear seat belts while driving. Seat belts
save lives—period. There is no normative reason for a LEO to drive without a seat belt.
Could it make sense for a LEO patrolling a dark alley at low speed looking for a suspect
144
not to have a seat belt on? Yes; this is not a normal situation. There may be many
specific instances when a LEO could justifiably have a vehicle in motion and not have
the seat belt on—in the half-block before pulling up to an in-progress call, for example.
Again, these exceptions should be specified rather than covered by a broad exemption for
all LEOs all the time.
Get Some Rest. Limit work periods and manage rest for LEOs. Fatigue reduces
human performance and increases risk of accidents in general and traffic collisions
specifically. Limits should be established for (a) total shift length, (b) consecutive work
days, (c) total weekly work hours, and (d) allowable hours of secondary employment.
Minimum rest periods should be required between shifts. Exceptions can be made for
emergencies (e.g., disasters, in-progress crimes, etc.). This has been done for commercial
truck drivers and pilots and makes sense for LEOs making life and death decisions too.
Train for Safety and Success. Design and mandate basic and refresher driver
training relevant to performance expected and outcomes desired. Broadly speaking,
LEOs are taught to shoot their weapons accurately (i.e., accuracy and skill) and to know
when to shoot or not shoot (i.e., deadly force decision-making). The same two elements
are essential to driving. LEOs need to know how to operate a vehicle proficiently in the
conditions they are expected to drive. This may mean driving in ice or snow or off road.
It might also mean driving at high speeds on interstates or in heavy traffic congestion.
LEOs may also need to improve skills for emergency and pursuit driving—how to
properly clear an intersection against a red light or how to manage a fleeing vehicle.
Likewise, LEOs also need to know when, where, and under what circumstances they
145
should or should not utilize these skills. They may know how to drive at 75 MPH, but
they also need to know they should not do so in a school zone when children are present.
Design Accountability. Create systems to hold LEOs, supervisors, and managers
accountable for their actions and inactions. Improper driving has consequences for
members of the public; it should also have consequences for LEOs. Education,
enforcement, and engineering (commonly known as the 3 E’s) have long been referred to
as the foundation of highway safety for the public. A standardized state-level system that
educates LEOs on expectations, rights, and responsibilities; enforces these expectations,
rights, and responsibilities; and engineers (whether through policy, training, or
technology) these expectations, rights, and responsibilities will improve traffic safety
outcomes.
Anticipated Challenges and Resistance
It is clear that state-level policy change almost always involves real challenges
and resistance. Some challenges and resistance are objectively valid and others may have
a more rhetorical or ideological basis that is subject to interpretation. Several examples
of anticipated challenges and resistance are enumerated in this section along with
potential counterarguments.
We’ve Always Done It This Way. Past practice has a powerful inertia. Combined
with some measure of success (e.g., a relatively low incidence of LEO fatalities), this can
be a persuasive argument. Often a lack of some experience incorrectly reassures people
that they are doing the right thing. An example of this is air transportation security.
There were thousands of uneventful flights before the tragedies that occurred on
September 11, 2001. Thereafter, a system-wide review found numerous security
146
inadequacies and failures (Kettl, 2004). The mantra that we’ve always done it this way is
insufficient and should invite a discussion about needs and goals. Simply proposing
change because it appears in the latest popular publication is similarly inadequate. While
decisions may not follow a democratic process in the realm of law enforcement
operations and policy, policymakers should collaborate with stakeholders and engage in
dialogue. This will at least create an opportunity for buy-in.
But What If… Many times people will look at a new requirement and take issue
with it because there might be a time when it would actually be better not to have such a
requirement. Wehr, Alpert, and Rojek (2012) describe this (with a direct quote) as the
“fear of the ninja assassin” (p. 25) when reporting why many LEOs routinely choose not
to wear their seat belts. Many LEOs fear that they might be ambushed and unable to take
cover or return fire if they are wearing a seat belt in a car.
There are exceptional circumstances and rarities that happen. Policy should be
designed for frequent, normal, and likely circumstances. If there are specific times when
it should not apply, those can be exempted. LEOs have been involved in tens of
thousands of traffic collisions. In just the 15 years analyzed in this research, 587 LEOs
died in automobile traffic collisions and approximately half of them were not wearing
seat belts. Although the exact number is not known, it is clear that few LEOs are
ambushed. Even fewer (if any) have been ambushed while driving. It just makes good
sense to operate based on what is most likely to happen as opposed to what might just
happen.
Local Control—The Miracle of Federalism. The reality is that the 10th
Amendment to the Constitution granted rights to states, not cities, counties, or other
147
subdivisions of government. The concept of local control took hold in this country many
years ago and has been clearly evidenced in the areas of education and law enforcement.
Many police chiefs and sheriffs will assert that it is their right to decide what is best for
their departments and communities. Arguing this point may be futile; rather, it may be
better to assert that since law enforcement has long desired recognition as a profession
(McClellan & Gustafson, 2012), it should be treated like one and uniformly regulated at
the state level (Gerber & Teske, 2000). Issues of broad concern and impact should be
broadly addressed.
Politics of Labor and Management. Matters relating to work hours, overtime,
and accountability (which might include discipline) are likely to create interest and
concern among labor leaders and representatives of management. The key in this respect
is to consider what the point of any policy change is and identify who is served by it.
Each of the proposed policy actions here is aimed at enhancing safety for LEOs. This
should be a common goal for everyone—labor and management leaders alike.
In terms of work hours and overtime, it is important to recognize operational
needs of law enforcement agencies and the monetary needs of LEOs. With regard to
accountability, it is important to avoid simply assigning blame, liability, or punishment.
The result may include staffing challenges or reduced overtime earnings, and there may
be times when someone gets blamed, held liable, or punished. These challenges of
operations and accountability are common in the professional workplace. What is more
important is that there will be lives saved as a result.
In fairness, this kind of policy change is a challenging undertaking, especially at
the state-level. In a smaller setting such as a municipality, the change effort can benefit
148
from the personal relationships of the stakeholders. Trust and mutual regard are difficult
to develop among stakeholders in larger groups where many people are not acquainted
with each other. Still, the degree to which a policy supports safety or specifies sanctions
is a matter of design, not a foregone conclusion of the policy process. One option could
be to specify liability or punishment in only the most egregious of cases where due
regard is wholly ignored. In terms of work hours and overtime, it may be an option to
specify required rest instead of limited work. The key is for all stakeholders to work
toward the common goal of LEO safety and find areas of agreement.
Economics of Programmatic Changes. Some policy changes require funding.
New standards or training requirements create new costs. Other policy changes require
little more than the time it takes to develop them (although it is a given that there should
be some form of training to support any policy change). When funding is limited, as it
tends to be for most states and law enforcement agencies, these issues must be prioritized.
Little cost is involved in terms of training or policy implementation to mandate that
officers wear seat belts or observe some form of speed limit(s). Enhanced standards or
training curricula may need to be strategically planned over time and phased in as
funding allows. Quite simply, though, a lack of funding should not in and of itself serve
as a barrier to policy action.
Examples for Review and Adaptation
The previous sections provided recommended policies and a review of likely
impediments to policy action. This section identifies states that have implemented
policies similar to those proposed. These states serve as examples and starting points for
149
creating new policies and adapting existing policies to better support LEO traffic safety
and reduce fatalities.
New York Work Hours Limits. State law limits LEO secondary employment to
20 hours per week. According to the survey respondent, except in an emergency
situation, LEO work shifts are limited to a maximum of 16 hours (including overtime)
and LEOs have to be given a minimum of 8 hours time off between shifts.
Connecticut and Minnesota Statewide Pursuit Policies. State law establishes
pursuit policies in Connecticut and Minnesota. In Connecticut, the policy is uniform
throughout the state. In Minnesota, the POST Board reviews and approves policies for
conformance with the state’s model policy and has oversight with the ability to take
action in terms of licensing for failure to comply with standards.
Mandatory Seat Belt States. Many states require everyone to wear seat belts
(according to survey respondents from AZ, DE, FL, IA, IN, KS, MA, MD, ME, MN,
MO, ND, NV, OR, SD, TN, TX, WV, and WY) without a LEO exception.
Talking/Text Prohibition States. Many states prohibit all drivers from texting
and/or talking on a cellular telephone. Among survey respondents, Massachusetts, New
Hampshire, and Vermont prohibit texting. Delaware, Oregon, South Dakota, and
Wyoming prohibit talking and texting.
Summary
This chapter outlined empirical findings that indicate specific methods for
reducing LEO traffic fatalities. It also provided practical policy recommendations for
putting research findings into law enforcement practice. It identified likely challenges
and areas of resistance to recommended policy changes. Finally, it highlighted several
150
example states where representative policies are already in place. The next chapter
outlines a process model for implementing policy changes given the complexities of these
four components—empirical evidence, policy recommendations, resistance, and
functioning examples. The chapter concludes with an agenda for future research.
151
CHAPTER VII
PATHWAY TO IMPLEMENTATION & FUTURE RESEARCH
This chapter provides ideas about where to go from here—what to do and how to
go about doing it. It begins with a brief review of the problem and interventions to
address the problem. It then discusses the gap between research and practice and clarifies
the nature of policy solutions. Next, it works through the process of implementation and
describes steps for successful policy implementation. The chapter then summarizes
recommendations and proposes a future research agenda in terms of needs and
opportunities. Finally, it concludes with key thoughts about the issue of law enforcement
officer (LEO) traffic fatalities.
Research Implementation Gap
Research findings presented in this study suggest that persistent rates of LEO
traffic fatalities are not just a technical problem, but a problem of politics, economics,
training, human performance, and industry/organizational culture. The findings
presented also suggest that the problem persists not due to a lack of knowledge about
effective interventions, but because known interventions and strategies are not
implemented.
The interventions and strategies found and recommended in this research consist
of (a) LEO speed limits, (b) standardization of emergency and pursuit driving, (c)
mandatory seat belt requirements, (d) work hours limits to manage fatigue, (e) limits on
driving distractions, (f) required basic and refresher training, and (a) LEO accountability
systems. These recommendations comprise what are alternatively known as evidence-
152
based practices (EBPs), which are described as “skills, techniques, and strategies that can
be used by a practitioner” (Fixsen, Naoom, Blase, Friedman, & Wallace, 2005, p. 26).
Unfortunately, modern history is rife with examples of the gap between good
theory and good practice. Popular media accounts frequently detail stories about
problematic practices that would benefit from research that has already been conducted in
diverse disciplines of scholarship ranging from nutrition to architecture to chemistry. For
example, a recent headline in The Herald read, “Sunscreen Recalled After Users Catch
Fire” (Associated Press, 2012). A brief review of this article indicates that the sunscreen
in question has highly flammable ingredients that, when sprayed on a person and exposed
to an ignition source, will likely catch on fire. There was no new research needed to
solve the problem in this case. The company simply did not utilize the basic chemical
research that already existed. While this is a particularly shocking example of direct
cause and effect, it clearly illustrates the theme of research findings not making it into
practice.
Detailing the Gap and Defining Policy
It may be worthwhile to highlight key points related to the gap in practice and
policy recommendations as they are conceptualized and used here. First, when referring
to the research implementation gap, I am not claiming a total absence of research in
practice. Rather, I mean to specify a general absence of the research findings in practice.
As described in Chapter 6, there are examples of states that have policies in place similar
to those recommended. These are still the exception rather than the rule. The policies
have not become normative throughout law enforcement practice. On a related point,
policy adoption is not the same as policy implementation (Pressman & Wildavsky, 1979).
153
There are many examples of policies that are never enforced and lines in budgets that
never get used. After a policy has been selected (i.e., a position adopted), then it must be
implemented (i.e., put into practice) (deLeon, 1988).
Second, when referring to policy, I use the term in the broad public policy sense
that specifies standards, requirements, or prohibitions. Pressman and Wildavsky (1979)
suggest it is useful “to talk about policy as a hypothesis containing initial conditions and
predicted consequences” (p. xx). When used in this way, policy can take many forms.
At a minimum, at the state-level, policy can take the form of (a) statutes (i.e., legislated
law), (b) rules (i.e., administrative law), or (c) case law (i.e., judicial precedence). For
the purposes of reducing LEO traffic fatalities, rules designed collaboratively through the
rulemaking process may be the most appropriate and accessible to law enforcement
stakeholders and policymakers.
Theory to Practice
Research and evidence-based practices are rarely sufficient to ensure concrete
changes to policy and practice; political and operational environments matter (deLeon,
1988). If this study simply stopped after detailing findings in Chapter 5, it would be
unlikely to spawn state-level policy change. There are many important factors in a
successful research-to-practice equation. At a minimum, there must be alignment
between the findings and the perceived problem and a willingness to attempt change on
the part of policymakers or stakeholders (and ideally both). An example best illustrates
this point.
In the mid-1960s, as part of President Johnson’s War on Poverty initiative and the
Great Society legislation, $23,000,000 was appropriated for the Economic Development
154
Administration to create jobs and reduce the unemployment rate through public projects
in Oakland, CA. The effort was an abject failure of implementation. Only $3,000,000 of
the allocation was spent and it was not used to create jobs for the unemployed. Pressman
and Wildavsky (1979) detailed this history in their book Implementation: How Great
Expectations in Washington Are Dashed in Oakland; Or, Why It's Amazing that Federal
Programs Work at All, This Being a Saga of the Economic Development Administration
as Told by Two Sympathetic Observers Who Seek to Build Morals on a Foundation of
Ruined Hopes. Their analysis revealed many failures. Notably, neither the City of
Oakland nor Oakland employers who were supposed to utilize the federal money to hire
the unemployed of inner-city Oakland, was a key stakeholder. Additionally, the key
program personnel left during implementation and the will to see the program through to
success apparently left with them.
Law enforcement has its own examples. The President’s Commission on Law
Enforcement and Administration of Justice (1967) called for (a) LEOs to have
baccalaureate degrees and (b) a form of national standardization (i.e., selection and
training standards). Forty-five years later there are still no states that require LEOs to
have a four-year degree and LEO selection and training are far from standardized
nationally (McClellan & Gustafson, 2012). A more recent example exists with the
ongoing debate and varied practices around police line-ups. For approximately half a
decade there has been scholarly research evidence that simultaneous line-ups are linked
with higher rates of incorrect identification of an innocent person as a suspect in a crime
than with sequential line-ups (Steblay, 2011). In this example there were some
methodological issues with early findings, but even so, many law enforcement agencies
155
have ignored the issue entirely, even since better studies have been conducted (Steblay,
2011). One reason for this resistance may likely be because the problem has not been
broadly perceived by law enforcement stakeholders. Researchers are answering
questions that many policymakers are not asking and this presents a real obstacle in the
theory-to-practice conversion process.
Policy Implementation and Implementation Science
Policy implementation has always been a feature of government—since the
Chinese built the Great Wall, the Romans created their empire, or the Colonies formed a
militia. In terms of modern scholarship, policy implementation has been a recognized
topic for study since at least the 1950s (deLeon, 1988). Even so, as deLeon and deLeon
(2002) note, “practical recognition has come and gone like an elusive spirit” (p. 467).
Apparently, in some cases, there is often more interest in adopting a policy position than
actually carrying one through to practice. deLeon (1988) describes this result as “more
of a policy veneer than a viable policy application” (p. 54). But, given the current
complexities and interest in education and healthcare reform, the challenge of
implementation is again rightly in vogue as a matter of serious concern.
It is worth noting that there are countless ideas—theories, models, frameworks—
about policy implementation and different scholars feel strongly about the merits of
various concepts (deLeon & deLeon, 2002). At a basic level, one issue that is
consistently questioned is that of top-down versus a bottom-up approach to
implementation (deLeon & deLeon, 2002; Sabatier, 1986). In the top-down approach,
policy is implemented via edict or direct action from a policymaker who may be the
advocate of the policy. In the bottom-up approach, policy is implemented by
156
stakeholders or line-personnel who are the doers of the organization and likely most in
touch with the implementation. Basically, the former model is more hierarchical and the
latter more democratic and discursive (deLeon & deLeon, 2002). What may be most
commonly advocated now is a mix of these approaches in what has been termed the
science of implementation (Fixsen, Naoom, Blase, Friedman, & Wallace, 2005).
Examples of Relevant Frameworks
Several academic disciplines (e.g., medicine, psychology, business, education)
have begun developing alternative implementation theories and frameworks. As such,
implementation science (also known as dissemination and implementation research) has
emerged as a sub-discipline useful for implementing EBPs. Implementation frameworks
guide policymakers and stakeholders in selection of appropriate EBPs, determination of
requirements for implementation, assessment of feasibility, and planning, adoption, and
realization (i.e., implementation) of EBPs.
Different implementation frameworks rely on varying assumptions and
definitions. While some researchers have attempted to reconcile these differences, new
models are still being generated. What appears to be common among the various
frameworks is a design intended to increase the probability that EBPs are adopted and
maintained. A key feature of this design is attention to how EBPs should be implemented
rather than simply which EBPs should be implemented.
Two example frameworks are presented here in brief to give a sense of the
diversity of options available to guide implementation. Succinct summaries of strengths
and weakness of these examples are noted. The next section provides a more robust
157
discussion of a third, nominally preferred framework, highlighted here for its ease of use
and broad applicability to the state-level systems at issue.
Exploration, Preparation, Implementation, Sustainment Model. Aarons,
Hurlburt, and Horwitz (2011) designed the Exploration, Preparation, Implementation,
Sustainment (EPIS) Model for the implementation of EBPs. The EPIS utilizes stages of
implementation in conjunction with consideration for the relationship between inner and
outer organizational contexts. EPIS specifies implementation considerations across
stages and notes intermediate transition points that are essential to the success of an
implementation initiative. This framework recognizes issues and activities of
stakeholders and emphasizes individual characteristics in addition to inter-organizational
dynamics. Strengths of EPIS include its emphasis on the interplay between
organizational contexts and its recognition of give-and-take dynamics. With regard to
weaknesses, EPIS depends on purveyors and practitioners to consider so many factors
that it may be burdensome for practitioners with limited resources.
Consolidated Framework for Implementation Research. Damschroder, Aron,
Keith, Kirsh, Alexander, and Lowery (2009) created the Consolidated Framework for
Implementation Research (CFIR). The CFIR offers a classification system to facilitate
implementation theory development in different domains. As such, it may be better
suited to academicians, although the authors claim it can provide value to practitioners as
well. The CFIR is somewhat complex. It specifies five major areas (e.g., intervention
characteristics, stakeholder characteristics) and then develops as many as a dozen
subareas under each major area to hone the framework for a given organizational context.
The strength of this model is that is recognizes interpersonal and organizational details
158
that impact implementation. At the same time, this strength also presents a downside in
that there is a great deal of complexity and detail involved in making it useful.
National Implementation Research Network Model
The National Implementation Research Network Model (NIRNM), developed by
Fixsen, Naoom, Blase, Friedman, and Wallace (2005), is one among many options, but
may be an ideal fit for the state-level policy recommendations proffered in this research.
These authors have worked to fuse the top-down and bottom-up approaches by bringing
purveyors together with local sites to implement systemic EBP interventions. Purveyors
refers to “an individual or group of individuals representing a program or practice who
actively work to implement that practice or program with fidelity and good effect”
(Fixsen, Naoom, Blase, Friedman, & Wallace, 2005, p. 14). These authors also make
reference to sites, which are the actual locations where EBPs are enacted.
The policy recommendations (i.e., EBPs) put forward in this research are intended
to be state-level initiatives. However, the actual implementation of the EBPs will occur
at police, sheriff, and state patrol agencies within the state, i.e., sites. In this respect,
stakeholders and state-level policymakers like POST agencies function as purveyors and
individual local law enforcement agencies serve as the sites. This is one reason the
NIRNM appears to be a good or natural fit. The emphasis on purveyors as facilitating
partners provides a clear advantage for law enforcement agencies and related
organizations embedded in a larger (state-level) network of policymakers and
stakeholders.
The NIRNM (Fixsen, Naoom, Blase, Friedman, & Wallace, 2005) integrates
research from numerous disciplines to provide tangible means for practitioners to select
159
suitable EBPs and implement them effectively. The authors refer to core intervention
components to describe foundational aspects of an intervention and stress that these
components need to be specifically attended to for successful implementation of EBPs.
The NIRNM has been used in healthcare and education reform efforts, but not
specifically for law enforcement or criminal justice applications. According to Fixsen,
Naoom, Blase, Friedman, and Wallace (2005), however, the approach is appropriate to
any institutional dynamic given that informed purveyors are involved in the
implementation process and key stakeholders have buy-in.
The NIRNM recognizes four phases of implementation (a) exploration and
adoption, (b) installation, (c) initial implementation, and (d) full implementation (Fixsen,
Naoom, Blase, Friedman, & Wallace, 2005). These phases encompass the essential
activities involved in the planning and execution of EBPs and may occur simultaneously
or repeatedly depending on the specific application. As such, implementation drivers are
an important aspect of the NIRNM and provide the implementation framework and
support mechanisms for the change process. These drivers are divided into three primary
content areas: (a) competency drivers, (b) organization drivers, and (c) leadership. These
drivers have been illustrated in a graphic form by the National Implementation Research
Network (NIRN) (2012) as depicted in Figure VII.1.
160
Figure VII.1 Implementation Drivers (NIRN, 2012)
According to Fixsen, Naoom, Blase, Friedman, and Wallace (2005), drivers of
competency include (a) selection (i.e., choosing the right people), (b) training (i.e.,
ensuring requisite knowledge, skills, and abilities), and (c) coaching (i.e., providing
support and feedback to personnel). Organization drivers include (a) decision support
data system (e.g., information to facilitate decision-making and assess EBPs), (b)
facilitative administration (e.g., alignment of procedures, resources, and structures to
support EBPs), and (c) systems intervention (e.g., strategies to inform and influence
external factors). These drivers are supported by leadership drivers—the primary
guidance for successful implementation of EBPs. This approach recognizes that both
adaptive and technical aspects of leadership are needed to facilitate change. All of these
drivers are aimed at continuous performance assessment to ensure consistency with EBPs
and timely identification of needed modifications.
Implementation teams also play a key role in guiding the implementation process
through each phase (Fixsen, Naoom, Blase, Friedman, & Wallace, 2005). Teams can
161
function at multiple levels, but must communicate effectively to ensure that changes are
scalable and systemic. It is best if “members of the implementation team represent each
organization or system whose cooperation is required to successfully install and operate”
the EBPs (p. 97). In the law enforcement settings described here the team would need to
include executives, trainers, and labor leaders, at a minimum.
Summary of Recommendations and Actions
Implementing EBPs is a tremendous undertaking at any level, but especially when
it involves multiple organizations as any state-level approach would. Given the
complexities of multiple-organization initiatives, it will likely be beneficial for
policymakers and stakeholders to plan for implementation even in advance of policy
adoption. The cliché start with the end in mind is apropos. To clarify, the next sections
offer simple breakdowns of five suggested steps for policymakers and practitioners who
read this research and want to move forward with the EBPs (i.e., policy
recommendations) detailed in Chapter 6.
Assemble Stakeholders and Select an Implementation Framework. Purveyor
teams should include a cross-section of stakeholders including labor (i.e., union) leaders,
management (e.g., chiefs and sheriffs), and state-level policymakers (e.g., POST
director). Given the specifics of the state, the purveyors (i.e., coalition of interested
participants) should choose an implementation framework (e.g., NIRNM) that meets their
specific needs. The primary components and stages of this framework should remain
consistent while purveyors are responsive to local (i.e., department) needs and practices
associated with any necessary modifications to both intervention components and
162
implementation practices that do not fundamentally alter the core goals or EBPs. If
possible, it may prove worthwhile to retain an implementation consultant.
Staff a Support Team. Purveyor teams should plan to be available to provide
technical assistance to law enforcement agencies during the implementation process.
Teams should be staffed with personnel skilled in specific EPBs (e.g., driver training,
scheduling) and related activities (e.g., local policy development).
Develop Templates and Example Resources. Law enforcement personnel will
not embrace a complex and time-consuming implementation process without practical
aids. Proponents of the policy initiative must continually balance an interest in EBP
consistency and implementation framework continuity with the likelihood that law
enforcement personnel will not adopt procedures that are vague and require significant
effort. Models (e.g., policies, training programs) should be clear and concise to make
them uncomplicated to implement. The easier a proposal is to implement, the more likely
it will happen. In few cases is it necessary for virtually any initiative to start from
scratch. Templates and examples from other agencies or states should be made available
as a resource for adaptation.
Develop and Maintain A Stakeholder Network. Implementation proponents and
law enforcement site (i.e., agency) partners should work collaboratively. In order to be
successful, the process has to be maintained and supported until the EBPs become
common practices—standard operating procedures as opposed to a new initiative.
Creating regular opportunities for discussion of the implementation process will allow for
refinement, troubleshooting, and a stronger, more engaged network.
163
Support Continuous Improvement. While this process may start solely to
address LEO traffic fatalities, the framework can be useful to all kinds of law
enforcement practices and should be taken advantage of as a means to disseminate
information and EBPs and to strengthen coordination and networking among law
enforcement agencies.
Future Research Agenda
Many questions about LEO traffic fatalities and collisions remain and there is a
lack of data. That is, there are not enough fatalities to clearly differentiate many state-
level differences and there is no nationally organized database of non-fatal LEO traffic
collisions. Research should be conducted on injuries, in addition to fatalities. This
would provide opportunity for better analyses through improved sample sizes. These
data may best be developed through cooperative arrangements with NIOSH and state
workers compensation systems. Any on-the-job injury should be reported to a
centralized state agency for workers compensation purposes and it should be possible to
filter by employer (e.g., police department) and mechanism of injury (e.g., traffic
collision).
Next, there is a dearth of information about training effects on collision
involvement. This research made extensive references to basic training and in-service
training and training hours in general. In a practical and political sense it would be useful
to both learning researchers and law enforcement practitioners to better understand how
different learning processes and inputs (in terms of type, quality, and duration of training)
impact outcomes (e.g., collision involvement). A study of this nature could be designed
around just a small sample of training sites and then a training typology could be
164
developed that might allow for more generalized comparisons (e.g., among agencies,
states).
Finally, a thorough study of types and impacts of state-level LEO regulation is
long overdue. Currently available sources were cited in this research and the survey data
produced through this study add to the available knowledge. However, a systematic
study of selection and training standards, as well as ongoing operational regulatory
standards by state would allow for more thoughtful analysis of the state of LEO
professionalization and the potential impacts of increased selection, training, education,
and oversight standards.
Summary and Conclusion
This chapter has provided detailed pathways to implementing policy change and
recommendations for how interested stakeholders or policymakers might begin the
process. It has also noted areas of need and opportunity for future research on LEO
traffic collisions. This is the culmination of this study and its recommendations for
practitioners and scholars.
It is clear that vehicles and traffic have become an integral part of society. They
have also become the most dangerous threat to on duty LEOs. This study shows,
however, that this is not an inevitable reality for LEOs. There are practical means for
individuals, supervisors, agency leaders, associations, and state and national law
enforcement leaders to address this challenge. While I have consistently advocated for a
state-level approach, there is nothing preventing an individual LEO from wearing a seat
belt, slowing down, getting more rest, and maintaining focus when driving. Similar
165
opportunities exist for supervisors, agency leaders, and others who choose to embrace
these recommendations. My hope is that those who read this will do so.
166
REFERENCES
Aarons, G. A., Hurlburt, M., & Horwitz, S. M. (2011). Advancing a conceptual model of evidence-based practice implementation in public service sectors. Administration and Policy in Mental Health, 38(1), 4-23.
Abdel-Aty, M. (2003) Investigating the relationship between cellular phone use and traffic safety. ITE Journal, 73(10), 38-42.
Adler, N. J., & Hansen, H. (2012). Daring to care: Scholarship that supports the courage of our convictions. Journal of Management Inquiry, 21(2), 128-139.
Alqadi, N. (2012, August). Changing the culture around vehicle safety. Community Policing Dispatch, 5(8) [e-newsletter]. Retrieved September 8, 2012 from http://cops.usdoj.gov/html/dispatch/08-2012/Culture-of-Vehicle-Safety.asp
Alpert, G. P., & Smith, W. C. (2008, June). Police pursuits after Scott v Harris: Far from ideal? Ideas in American Policing (No. 10). Washington, DC: Police Foundation.
Alpert, G. P., Rojek, J., & Porter, L. E. (2012). Measuring the impact of organisational culture and climate on police officers’ decisions and behaviour [Working paper]. Nathan, Australia: ARC Centre of Excellence in Policing and Security
Aristotle. (1943). On man in the universe: Metaphysics, parts of animals, politics, poetics (Louise Ropes Loomis, Ed.). Rosyln, New York: Walter J. Black. (Original work published circa 350 BCE)
Associated Press. (2012, October 19). Sunscreen recalled after users catch fire. The Herald. Accessed October 20, 2012 from http://www.heraldnet.com/article/20121019/BIZ/710199835/1005/BIZ
Barker, J. R. (1993). Tightening the iron cage: Concertive control in self-managing teams. Administrative Science Quarterly, 38, 408-437.
Batiste, J. R., Wagers, M. L., & Ashton, R. J. (2011, July). Preventing traffic-related line-of-duty deaths. The Police Chief, 78, 52–55.
Bayley, D. H., & Shearing, C. D. (2000). New structure of policing: Description, conceptualization, and research agenda—Final report. Washington, DC: National Institute of Justice.
167
Bener, A., Lajunen, T., Özkan, T., & Haigney, D. (2006). The effect of mobile phone use on driving style and driving skills. International Journal of Crashworthiness, 11, 459-465.
Bolman, L. G., & Deal, T. E. (1997). Reframing organizations: Artistry, choice, and leadership (2nd ed.). San Francisco: Jossey-Bass.
Bureau of Labor Statistics. (2011). National census of fatal occupational injuries in 2010 (preliminary results) (News release (USDL-11-1247)). Washington, DC: Author.
California Commission on Peace Officer Standards and Training (2010, May). POST works with IACP to improve officer highway safety (Administrative Progress Report). Sacramento, CA: Author.
California Commission on Peace Officer Standards and Training (2012). State-level differences in law enforcement officer traffic fatalities [survey]. Sacramento, CA: Author.
Carbon Motors. (2011). Carbon Motors Corporation [Brochure]. Retrieved September 22, 2012 from http://www.carbonmotors.com/machine/brochure
Carte, G. E. (1969). Evaluation of professionalism within the criminal justice system. Criminologica, 7, 39-50.
Castro, C. (2008). Human factors of visual and cognitive performance in driving. Boca Raton, FL: CRC Press.
Christie, R. (2001). The effectiveness of driver training as a road safety measure: A review of the literature. Report no: 01/03. Noble Park, Victoria: Royal Automobile Club of Victoria (RACV) Ltd.
Clarke, C., & Zak, M. J. (1999). Fatalities to law enforcement officers and firefighters, 1992-97. Compensation and Working Conditions, 4, 3-7.
Conser, J. A., Paynich, R., & Gingerich, T. E. (2011). Law enforcement in the United States. Burlington, MA: Jones & Bartlett.
Cox, O. (2004, July). From the director: 2003 traffic safety data. The Police Chief, 71. Retrieved August 2, 2012 from http://www.policechiefmagazine.org/magazine/index.cfm?fuseaction=display_arch&article_id=342&issue_id=72004
168
Cutcliffe, C. M. (1966). [Review of the book Unsafe at any Speed by Ralph Nader]. Washington & Lee Law Review, 23(2), 445-447.
Damschroder, L., Aron, D., Keith, R., Kirsh, S., Alexander, J., & Lowery, J. (2009). Fostering implementation of health services research findings into practice: A consolidated framework for advancing implementation science. Implementation Science, 4, 50.
Dawson, D., & Reid, K. (1997). Fatigue, alcohol and performance impairment. Nature, 388, 235.
DeAngelo, G., & Hansenyz, B. (2010). Life and death in the fast lane: Police enforcement and roadway safety. Manuscript submitted for publication.
deLeon, P. (1988). Advice and consent: The development of the policy sciences. New York: Russell Sage Foundation.
deLeon, P., & deLeon, L. (2002). What ever happened to policy implementation? An alternative approach. Journal of Public Administration Research and Theory, 12(4), 467-492.
De Smet, A. (Ed.). (2008). Transportation accident analysis and prevention. New York: Nova Science.
Derksen, S., & Keselman, H. J. (1992). Backward, forward and stepwise automated subset selection algorithms: Frequency of obtaining authentic and noise variables. British Journal of Mathematical and Statistical Psychology, 45(2), 265–282.
Dewey, J. (1927). The public and its problems. New York: H. Holt and Co.
Diggle, P., Heagerty, P. J., Liang, K., & Zeger, S.L. (2002). Analysis of longitudinal data (2nd ed.). Oxford, UK: Oxford University Press.
Dorn, L. (Ed.). (2005). Driver behaviour and training (Vol. 2). Burlington, VT: Ashgate.
Dorn, L. (Ed.). (2010) Driver behaviour and training (Vol. 4). Burlington, VT: Ashgate.
Dorn, L., & Barker, D. (2005). The effects of driver training on simulated driving performance. Accident Analysis & Prevention, 37(1), 63–69.
85-mph speed limit ok'd on Texas road. (2012, September 7). USA Today, p. 3A.
169
Federal Bureau of Investigation. (1995a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (1996a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (1997a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (1998a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (1999a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2000a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2001a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2002a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2003a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2004a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2005a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2006a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2007a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (2008a). Crime in the United States. Washington, DC: Author.
170
Federal Bureau of Investigation. (2009a). Crime in the United States. Washington, DC: Author.
Federal Bureau of Investigation. (1995b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (1996b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (1997b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (1998b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (1999b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2000b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2001b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2002b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2003b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2004b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2005b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2006b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2007b). Law enforcement officers killed and assaulted. Washington, DC: Author.
171
Federal Bureau of Investigation. (2008b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Bureau of Investigation. (2009b). Law enforcement officers killed and assaulted. Washington, DC: Author.
Federal Highway Administration. (1995). Highway Statistics 1995. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs95/
Federal Highway Administration. (1996). Highway Statistics 1996. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs96/
Federal Highway Administration. (1997). Highway Statistics 1997. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs97/
Federal Highway Administration. (1998). Highway Statistics 1998. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs98/
Federal Highway Administration. (1999). Highway Statistics1999. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs99/
Federal Highway Administration. (2000). Highway Statistics 2000. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs00/
Federal Highway Administration. (2001). Highway Statistics 2001. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs01/
Federal Highway Administration. (2002). Highway Statistics 2002. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs02/
Federal Highway Administration. (2003). Highway Statistics 2003. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs03/
Federal Highway Administration. (2004a). Highway Statistics 2004. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs04/
Federal Highway Administration. (2004b). Maximum daytime posted speed limits - 2002. Retrieved February 9, 2012 from http://www.fhwa.dot.gov/policy/ohpi/speeds.htm
Federal Highway Administration. (2005). Highway Statistics 2005. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs05/
172
Federal Highway Administration. (2006). Highway Statistics 2006. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs06/
Federal Highway Administration. (2007). Highway Statistics 2007. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs07/
Federal Highway Administration. (2008). Highway Statistics 2008. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs08/
Federal Highway Administration. (2009). Highway Statistics 2009. Retrieved September 13, 2012 from http://www.fhwa.dot.gov/policy/ohim/hs09/
Fell, J. C. (1975). A motor vehicle accident casual system: The human element. Human Factors and Ergonomics Society Annual Meeting Proceedings, 19, 42-48.
Fell, J., & Tharp, K. J. (1969). Multi-disciplinary investigations to determine automobile accident causation: Report No. 2 - Accident case analysis and data processing procedures. Detroit, MI: Cornell Aeronautical Laboratory, Inc. (CAL)/Automobile Manufacturer's Association, Inc.
Fixsen, D. L., Naoom, S. F., Blase, K. A., Friedman, R. M. & Wallace, F. (2005). Implementation research: A synthesis of the literature (FMHI Publication #231). Tampa, FL: University of South Florida, Louis de la Parte Florida Mental Health Institute, The National Implementation Research Network.
Fischer, F. (2003). Reframing public policy: Discursive politics and deliberative practices. New York: Oxford University Press.
Frankfort-Nachmias, C. & Nachmias, D. (2008). Research methods in the social sciences (7th ed.). New York: Worth.
Fredrickson, G. M. (1995). From exceptionalism to variability: Recent developments in cross-national comparative history. Journal of American History, 82(2), 587-604.
Friedman, L. S., Barach, P., & Richter, E. D. (2007). Raised speed limits, case fatality and road deaths: A six year follow-up using ARIMA models. Injury Prevention, 13(3), 156–161.
Friedman, L. S., Hedeker, D., & Richter, E. D. (2009). Long-term effects of repealing the national maximum speed limit in the United States. American Journal of Public Health, 99, 1626-1621.
173
Frost, R. (1917). North of Boston. New York: Henry Holt and Company.
Garber, N., J., Gadiraju, R. (1992). Impact of differential speed limits on the speed of traffic and the rate of accidents. Transportation Research Record, 1375, 44-52.
Gerber, B. J., & Teske, P. E. (2000). Regulatory policymaking in the American states: A review of theories and evidence. Political Research Quarterly, 53, 849-886.
Gerring, J. (2007). Case study research: Principles and practices. New York: Cambridge University Press.
Giddens, A. (1991). Modernity and self-identity: Self and society in the late modern age. Stanford, CA: Stanford University Press.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: strategies for qualitative research. Chicago: Aldine.
Gustafson, B. G. (Ed.). (2009). Driver training study: Volume I. Sacramento, CA: Commission on Peace Officer Standards and Training.
Gustafson, B. G., & Cappitelli, P. A. (2010, March). SAFE driving—The role of the chief. The Police Chief, 77, 38-41.
Gustafson, B. G., & Rice, T. (2010, November). What’s driving fatal law enforcement collisions? A state-level analysis 2000-2009. Poster session presented at the American Society of Criminology Annual Meeting, San Francisco, CA.
Harcourt, B. E., & Ludwig, J. (2006). Broken windows: New evidence from New York City and a five-city social experiment. The University of Chicago Law Review, 73, 271-320.
Hayter, A. J. (2006). Probability and statistics for engineers and scientists (3rd ed.). Pacific Grove, CA: Duxbury.
Hedlund, J. (2007). Improving traffic safety culture in the United States: The journey forward. Washington, DC: AAA Foundation for Traffic Safety.
Hoyert, D. L. (2012, March). 75 years of mortality in the United States, 1935-2010 (NCHS Data Brief No. 88). Washington, DC: National Center for Health Statistics.
174
Hutson, H. R., Rice, P. L. Jr., Chana, J. K., Kyriacou, D. N., Chang, Y., & Miller, R. M. (2007). A review of police pursuit fatalities in the United States from 1982-2004. Prehospital Emergency Care, 11, 278–283.
International Association of Chiefs of Police (Producer). (2010). Is today your day? [Training Video]. Alexandria, VA: Author.
International Association of Chiefs of Police. (2011). Preventing traffic-related line-of-duty deaths. Alexandria, VA: Author.
International Association of Directors of Law Enforcement Standards and Training. (2005). Sourcebook. Albion, MI: Author.
James, S. M., & Vila, B. (2012). Driven to distraction. Journal of California Law Enforcement, 46(2), 14-18.
Janeksela, G. M. (1981). Building bridges between the academic and practitioner communities in criminal justice. American Journal of Criminal Justice, 6, 96-101.
Jonsson, P. (2008, November 12). This police cruiser scans license plates, sniffs out 'dirty' bombs. The Christian Science Monitor. Retrieved September 22, 2012 from http://www.csmonitor.com/USA/2008/1112/p01s03-usgn.html
Kettl, D. F. (2004). System under stress: Homeland security and American politics. Washington, DC: CQ Press.
Klinger, D. (2004). Into the kill zone: A cop’s eye view of deadly force. San Francisco: Jossey-Bass.
Lasswell, H. D. (1951). The policy orientation. In Sandra Braman (Ed.) Communication researches and policy-making (pp. 85-104). Cambridge, MA: MIT Press.
Levitt, S. D. (2005). Freakonomics: A rogue economist explores the hidden side of everything. New York: HarperCollins.
Lindlof, T. R., & Taylor, B. B. (2002). Qualitative communication research methods (2nd ed.). Thousand Oaks, CA: Sage.
Loukopoulos, L. D., Dismukes, R. K., & Barshi, I. (2009). The multitasking myth: Handling complexity in real-world operations. Burlington, VT: Ashgate.
175
Los Angeles Police Department. (2012). Officers killed in the line of duty – Wylie, James P. Retrieved August 25, 2012 from http://lapdonline.org/officers_killed_in_the_line_of_duty/content_basic_view/29550
MacGregor, D. (2009). The safety race: Transitions to the fourth age of the automobile. In J. Conley and A. T. McLaren (Eds.) Car troubles: Critical studies of automobility and auto-mobility (pp. 95-110). Farnham, UK: Ashgate.
Magers, J., & Klein, L. (2002). Police basic training: A comparative study of states' standards in the United States. Illinois Law Enforcement Executive Forum, 2(2), 103–113.
Maryland Police & Correctional Fallen Officers Memorial. (2012). The fallen – Alonzo B Bishop. Retrieved August 25, 2012, from http://www.mdfallenofficers.org/the-fallen.html?officer=466
Mashaw, J. L., & Harfst, D. L. (1987). Regulation and legal culture: The case of motor vehicle safety. Yale Journal on Regulation, 4, 257-316.
Mayo, L. (2006, August). College education and policing. The Police Chief, 73, 20-38.
Mccartta, A. T., Hellingaa, L. A., & Bratimana, K. A. (2006). Cell Phones and Driving: Review of Research. Traffic Injury Prevention, 7(2), 89-106.
McClellan, S. E., & Gustafson, B. G. (2012). Communicating law enforcement professionalization: Social construction of standards. Policing: An International Journal of Police Strategies & Management, 35(1), 104-123.
McNabb, D. E. (2002). Research methods in public administration and nonprofit management: Quantitative and qualitative approaches. Aramonk, NY: M. E. Sharpe.
Miller, G. J. (1992). Managerial dilemmas: The political economy of hierarchy. Cambridge, UK: Cambridge University Press.
Müller, A., & Stajic, D. (2011). Safer roads through vehicle safety systems. Journal of Society for Transportation and Traffic Studies, 2(2), 41-55.
Nader, R. (1965). Unsafe at any speed: The designed-in dangers of the American automobile. New York: Grossman.
National Highway Traffic Safety Administration. (1985). National Traffic and Motor Vehicle Safety Act of 1966: Legislative history (Vol. 1). Washington, DC: Author.
176
National Highway Traffic Safety Administration. (1997). Traffic safety facts 1996: A compilation of motor vehicle crash data from the Fatality Analysis Reporting System and the General Estimates System (DOT HS 808 649). Washington, DC: Author.
National Highway Traffic Safety Administration. (2006). This is NHTSA: People saving people (DOT HS 810 552). Washington, DC: Author.
National Highway Traffic Safety Administration. (2008). FARS analytic reference Guide: 1975 to 2007 (DOT HS 810 937). Washington, DC: Author.
National Highway Traffic Safety Administration. (2010a). An analysis of the significant decline in motor vehicle traffic fatalities in 2008 (DOT HS 811 346). Washington, DC: Author.
National Highway Traffic Safety Administration. (2010b). Characteristics of law enforcement officers’ fatalities in motor vehicle crashes (DOT HS 811 411). Washington, DC: Author.
National Highway Traffic Safety Administration. (2012). Fatality Analysis Reporting System (FARS) Encyclopedia [Web-based database]. Available at http://www-fars.nhtsa.dot.gov/Main/index.aspx
National Implementation Research Network. (2012). Implementation drivers. Retrieved October 15, 2012, from http://nirn.fpg.unc.edu/learn-implementation/implementation-drivers
National Institute for Occupational Safety and Health. (2009, April). National occupational research agenda (NORA): National public safety agenda for occupational safety and health research and practice in the US public safety sub sector. Washington, DC: Author.
National Institute for Occupational Safety and Health. (2012). Fatality Assessment and Control (FACE) Program. Retrieved August 1, 2012 from http://www.cdc.gov/niosh/face/default.html
National Law Enforcement Officers Memorial Fund. (2009, December). Law enforcement officer deaths: Preliminary 2009: A tale of two trends: Overall fatalities fall, fatal shootings on the rise (Research Bulletin). Washington, DC: Author.
National Law Enforcement Officers Memorial Fund. (2012a). Alonzo B. Bishop. Retrieved August 25, 2012 from http://www.nleomf.org/officers/search/search-results/alonzo-b-bishop.html
177
National Law Enforcement Officers Memorial Fund. (2012b). J. P. Wylie. Retrieved August 25, 2012 from http://www.nleomf.org/officers/search/search-results/j-p-wylie.html
National Law Enforcement Officers Memorial Fund. (2012c). [Web-based database]. Available at http://www.nleomf.org/officers/search/
National Research Council. (1966). Accidental death and disability: The neglected disease of modern society. Washington, DC: Author.
National Transportation Safety Board. (n.d.). History of the National Transportation Safety Board. Retrieved August 26, 2012 from http://www.ntsb.gov/about/history.html
Officer Down Memorial Page. (2012a). Police Officer Alonzo B. Bishop. Retrieved August 25, 2012 from http://www.odmp.org/officer/1879-police-officer-alonzo-b-bishop
Officer Down Memorial Page. (2012b). Policeman James P. Wylie. Retrieved August 25, 2012 from http://www.odmp.org/officer/14565-policeman-james-p-wylie
Officer Down Memorial Page. (2012c). [Web-based database]. Available at http://www.odmp.org/search
Pant, P. D., Adhami, J. A., & Niehaus, J. C. (1992). Effects of the 65-mph speed limit on traffic accidents in Ohio. Transportation Research Record, 1375, 53-60.
President’s Commission on Law Enforcement and Administration of Justice (1967). The challenge of crime in a free society. Washington, DC: Government Printing Office.
Pressman, J. L., & Wildavsky, A. B. (1979). Implementation: How great expectations in Washington are dashed in Oakland; Or, why it's amazing that federal programs work at all, this being a saga of the Economic Development Administration as told by two sympathetic observers who seek to build morals on a foundation of ruined hopes (2nd ed.). Berkeley, CA: University of California Press.
Rabe-Hesketh, S., & Skrondal, A. (2008). Multilevel and longitudinal modeling using Stata (2nd ed.). College Station, TX: Stata Press.
Reaves, B. A. (2011). Census of state and local law enforcement agencies, 2008 (NCJ 233982). Washington, DC: Bureau of Justice Statistics.
178
Rice, T., & Gustafson, B. (2012) Trends and characteristics of traffic collisions involving California law enforcement officers. Journal of California Law Enforcement, 46(2), 19-24.
Richens, J., Imrie, J., & Copas, A. (2000). Condoms and seat belts: The parallels and the lessons. The Lancet, 355, 400–403.
Richmond, CA – Official Website. (2012). Memorial to Bradley Moody. Retrieved August 2, 2012, from http://www.ci.richmond.ca.us/index.aspx?nid=1601
Rivara, F. P., & Mack, C. D. (2004). Motor vehicle crash deaths related to police pursuits in the United States. Injury Prevention, 10, 93–95.
Rosen, C. (2008). The myth of multitasking. The New Atlantis, 20, 105-110.
Sabatier, P. A. (1986). Top-down and bottom-up approaches to implementation research: A critical analysis and suggested synthesis. Journal of Public Policy, 6(1), 21-48.
Salminen, S., & Lähdeniemi, E. (2002). Risk factors in work-related traffic. Transportation Research Part F: Traffic Psychology and Behaviour, 5, 77-86.
Schneider, W. (1985). Training high-performance skills: Fallacies and guidelines. Human Factors, 27(3), 285-300.
Schultz, D. P., Hudak, E., & Alpert, G. P. (2009, April). Emergency driving and pursuits. FBI Law Enforcement Bulletin, 78, 1-6.
Schultz, D. P., Hudak, E., & Alpert, G. P., (2010, March). Evidence-based decisions on police pursuits: The officer’s perspective. FBI Law Enforcement Bulletin, 79(3), 1-7.
Scott, M. S. (2009). Progress in American policing? Reviewing the national reviews. Law & Social Inquiry, 34, 171-185.
Sewell, G. (1998). The discipline of teams: The control of team-based industrial work through electronic and peer surveillance. Administrative Science Quarterly, 43(2), 397-428.
Shepsle, K. A., & Bonchek, M. S. (1997). Analyzing politics: Rationality, behavior, and institutions. New York: W. W. Norton.
179
Sorensen, D. W. M. (1994). Motor vehicle accidents. In Travis H. and M. Gottredson (Eds.) The generality of deviance (pp. 113-130). New Brunswick, NJ: Transaction Publishers.
Steblay, N. K. (2011). What we know now: The Evanston Illinois field lineups. Law and Human Behavior, 35(1), 1-12.
Stewart, R. B. (1983). Regulation in a liberal state: The role of non-commodity values. Yale Law Journal, 92(8), 1537-1590.
Strauss, A. L., (1987). Qualitative analysis for social scientists. Cambridge, UK: Cambridge University Press.
Strauss, A. L., & Corbin, J. (1998). The basics of qualitative analysis: Grounded theory procedures and techniques (2nd ed.). Thousand Oaks, CA: Sage.
Taylor, F. W. (1911). The principles of scientific management. New York: Harper & Brothers.
Teske, P. E. (2004). Regulation in the states. Washington, DC: Brookings Institution Press.
US Census Bureau. (2009). Population density – 2009. Retrieved October 5, 2010 from http://www.census.gov/popest/gallery/maps/popdens-2009.html
US Census Bureau. (2012) Quick facts. Retrieved September 22, 2012 from http://quickfacts.census.gov/qfd/states/00000.html
Van de Ven, A. H. (2007). Engaged scholarship: A guide for organizational and social research. New York: Oxford University Press.
Vila, B. (2000). Tired cops: The importance of managing police fatigue. Washington, DC: Police Executive Research Forum.
Vila, B. (2006). Impact of long work hours on police officers and the communities they serve. American Journal of Industrial Medicine, 49(11), 972-980.
Vila, B. (2009). Sleep deprivation: What does it mean for public safety officers?. NIJ Journal, 262, 26-31.
180
Vila, B., & Gustafson, B. G. (2011, Spring). The ongoing crisis: Officer-involved collisions—Why they happen and what can be done. California Peace Officer, 12-13.
Vogel, R. E., & Gamache, J. M. (1981). Competency-based education: An alternative for criminal justice education?. American Journal of Criminal Justice, 6, 81-95.
Wagenaar, A. C., Maybee, R. G., & Sullivan, K. P. (1988). Mandatory seat belt laws in eight states: A time-series evaluation. Journal of Safety Research, 19, 51-70.
Wehr, K., Alpert, G., & Rojek, J. (2012). “The fear of the ninja assassin:” Understanding the role of agency culture in injurious and fatal on-duty vehicle Collisions. Journal of California Law Enforcement, 46(2), 25-31.
Weimer, D. L. & Vining, A. R. (2005). Policy analysis: Concepts and practice (4th ed.). Upper Saddle River, NJ: Pearson Education.
Weiss, N. A. (2008). Introductory statistics (8th ed.). Boston: Pearson
Wetmore, J. M. (2009). Implementing restraint: Automobile safety and the US debate over technological and social fixes. In J. Conley and A. T. McLaren (Eds.) Car troubles: Critical studies of automobility and auto-mobility (pp. 111-124). Farnham, UK: Ashgate.
Whitelegg, J. (1983). Road safety: Defeat, complicity and the bankruptcy of science. Accident Analysis & Prevention, 15(2), 153-160.
Williamson, A. M., Feyer, A-M., Mattick, R. P., Friswell, R., & Finlay-Brown, S. (2001). Developing measures of fatigue using an alcohol comparison to validate the effects of fatigue on performance. Accident Analysis & Prevention, 33(3), 313-326.
Yauch, C. A., & Steudel, H. J. (2003). Complementary use of qualitative and quantitative cultural assessment methods. Organizational Research Methods, 6, 465-481.
Zlatoper, T. J. (1991). Determinants of motor vehicle deaths in the United States: A cross-sectional analysis. Accident Analysis & Prevention, 23, 431-436.
181
APPENDIX A
CalPOST (2012) Survey Instrument Summary
This appendix provides a summary of the questions/content of the CalPOST (2012) State-Level Differences in Law Enforcement Officer Traffic Fatalities survey that was distributed to the US population of POST directors and state highway patrol chiefs/colonels/commissioners via SurveyMonkey. Where possible, information is present verbatim as it was in the online version of the survey (color images are presented in grayscale here). Purpose and Audience – Why is California POST conducting this survey?
182
Respondent Information: You, Your Agency, Your State... – This survey is being distributed to state peace officer standards and training (POST) agencies (or equivalents) and state traffic/highway patrol/police agencies (or equivalents). 1. Please fill in the contact information below to identify who is responding to this
survey, what your title/position is, and what agency you work for. •
2. Please confirm if you are responding as a standards and training agency or as a
highway enforcement agency: • Standards & Training (i.e., POST) • Highway Enforcement (i.e., State Patrol) • Both (Standards/Training AND Highway Enforcement) • Other (please explain below)
Background on the State-Level Differences Study – Reviewing this page will provide you with background information that will contextualize the information this survey aims to gather. This study builds on preliminary research into state-level differences in law enforcement officer (LEO) traffic-related fatalities conducted by Bryon Gustafson (CalPOST & University of Colorado Denver) and Tom Rice (University of California, Berkeley). The research foundation is the 10-year average of LEO traffic fatalities by state (2000-2009), which is summarized in the two images at the bottom of this page. For more detail, you can view the published poster “What’s Driving Fatal Law Enforcement Collisions? A State-Level Analysis: 2000-2009” which includes high-resolution versions of the images below. The goal of this study is to determine what accounts for different law enforcement traffic fatality rates by state. More specifically, the study aims to identify which of the explanations for different collision rates are amendable via policy and practice adaptation. To clarify what this study is asking, consider that there are several levels at which traffic fatality rates can be impacted: First, individual-level factors can contribute to involvement in a fatal collision. An example of this would be an officer who is exhausted and falls asleep while driving and fatally crashes. Such a crash would be the result of the individual officer's condition. Second, there are agency-level factors that can contribute. An example of this might be an agency that requires officers to work mandatory overtime on patrol. Officers working for such an agency are likely to drive significantly more and might also be more fatigued. In this example, the increased exposure (more hours and miles on the road) and potential fatigue would be an agency-level factor that could contribute to a higher fatality rate.
183
Last (for our purposes), there are state-level factors that can contribute. An example of a state-level factor would be a mandatory seatbelt law for officers. Another example would be required driver training. In these examples, the required use of safety equipment and mandated training (statewide) would be state-level factors (that would be expected to contribute to lower fatality rates). With this background in mind, please consider the state-level factors in your state as you answer the questions on the following pages. For examples of some state-level factor differences and to see the officer traffic fatality rate in your state (from 2000-2009), reference the map and chart below (or review the “What’s Driving Fatal Law Enforcement Collisions? A State-Level Analysis: 2000-2009” poster publication, which will open in a new tab/window).
The rates shown on this map are the annual average (statistical mean) number of LEOs killed in traffic per 100,000 LEOs. The "per 100,000" rate is the norm in traffic fatality reporting. Actual numbers of LEOs killed in traffic are reported in the chart below. State-Level Differences 2000-2009
184
MPH: Maximum Daytime Speed Limit on Federal Highway (Federal Highway Administration: http://www.fhwa.dot.gov/policy/ohpi/speeds.htm) DT: State-mandated Minimum Hours of Basic Driver Training (Int'l Assoc. of Directors of Law Enforcement Standards and Training Sourcebook, 2005) PD: Mean (2000-2009) Population Density - Persons Per Square Mile (US Census Bureau) OpR: Mean (2000-2009) Officers Per 1,000 Residents - Officers/Residents*1,000 (US Census Bureau & FBI Crime in the United States) GTFR: Mean (2000-2009) General Population Traffic Fatality Rate - Fatalities/Population*100,000 (National Highway Traffic Safety Administration Fatality Analysis Reporting System) # 00-09: Total Number (count) of Law Enforcement Officers Killed in Traffic 2000-2009 (inclusive) (FBI Crime in the United States, LEOKA, & Officer Down Memorial Page) OTFR: Mean (2000-2009) Officer Traffic Fatality Rate - Officer Fatalities/Officer Population*100,000 (FBI Crime in the United States, LEOKA, & Officer Down Memorial Page)
185
186
Training-Related Questions – What impact do statewide training standards (minimums) have on law enforcement officer (LEO) traffic fatalities? Generally, the industry-assumed proposition is that additional training will result in reduced negative outcomes (e.g., traffic collisions). The minimum and average "hours...established for law enforcement basic training" from the International Association of Directors of Law Enforcement Standards and Training (IADLEST) Sourcebook (2000) are listed for each state below. Please review the information for your state (and any states you normally reference for comparison purposes).
The "minimum number of hours...for entry-level law enforcement driver training" from the IADLEST Sourcebook (2005) are listed for each state below. Please review the information for your state (and any states you normally reference for comparison purposes).
187
3. Does your state mandate in-service (refresher) driver training for LEOs?
• Yes • No
4. If your state does mandate in-service driver training:
a. How often? • N/A (describe below) • Yearly • Every 2 years • Every 3 years • Every 4 years • Every 5 years • Other (describe below)
b. How much?
• N/A (describe below) • 1 hour
188
• 2 hours • 3 hours • 4 hours • 5 hours • 6 hours • 7 hours • 8 hours • 9 hours • 10 hours • 11 hours • 12 hours • Other (describe below)
c. What kind?
• N/A (describe below) • Computer-based • Table-top • Simulator (LEDS) • Classroom • Behind-the-wheel • Mix (please describe) • Other (please describe)
5. Please indicate your response to the following statements by selecting whether you:
SD (Strongly Disagree), Disagree, N (are Neural), Agree, SA (Strongly Agree) • Additional "basic training" (hours) contributes to lower LEO traffic fatality rates. • Additional "entry-level driver training" (hours) contributes to lower LEO traffic
fatality rates. • Additional "in-service (refresher) driver training" (hours) contributes to lower
LEO traffic fatality rates. State Regulation Questions – What impact do statewide regulations have on LEO traffic fatalities? Regulatory standards literature supports the proposition that more stringent regulatory standards (e.g., laws, regulations, law enforcement requirements) will result in lower collision fatality rates. 6. Does your state have a "mandatory seatbelt law" that requires on-duty LEOs to wear a
seatbelt? • No • Yes
7. Does your state have a "cellphone law" that prohibits on-duty LEOs from talking on a
cellphone and/or texting while driving? • No • Yes
189
o If yes, what year did the law go into effect? (a web-link to the law would also suffice):
8. Does your state have any "speed limits" (i.e., laws, regulations, or policies) that
restrict the overall speeds on-duty LEOs can drive when using lights and sirens? (This question assumes that LEOs are required—legally/technically—to drive the speed limit when not using lights and sirens. Please comment below if this is not the case in your state.) • No • Yes
9. Does your state have any "pursuit laws" (i.e., laws, regulations, or policies) that
specify what LEOs can, cannot, or must do--or when they can or cannot engage--in pursuits? • No • Yes
o If yes, please indicate the year the pursuit law took effect and what it requires/prohibits (a web-link to the law would suffice):
10. Does your state have any "shift length/overtime or secondary employment laws" (i.e.,
laws, regulations, or policies) that specify or otherwise limit how much LEOs can work? • No • Yes
o If yes, please indicate the year the law took effect and what it requires/prohibits (a web-link to the law would suffice):
11. Does your state have a "mandatory move over law" that requires motorists to slow
and/or move over for emergency vehicles stopped on the side of the road? • No • Yes
o If yes, what year did the law go into effect? (a web-link to the law would also suffice):
12. Please indicate your response to the following statements by selecting whether you:
SD (Strongly Disagree), Disagree, N (are Neural), Agree, SA (Strongly Agree) • "Mandatory seatbelt laws" (for on-duty LEOs) contribute to lower LEO traffic
fatality rates. • "Cellphone laws" (prohibition for on-duty driving LEOs) contribute to lower LEO
traffic fatality rates. • "Speed limits" (for on-duty LEOs) contribute to lower LEO traffic fatality rates. • "Pursuit laws" contribute to lower LEO traffic fatality rates. • "Shift length/overtime or secondary employment laws" contribute to lower LEO
traffic fatality rates. • "Mandatory move over laws" contribute to lower LEO traffic fatality rates.
190
Budget/Spending Questions – What impact does state-level budget/spending have on officer traffic fatalities? Economic policy literature supports the proposition that states that invest more on law enforcement and highway safety will have lower collision fatality rates. 13. What is your agency's budget for the current ('11-'12) fiscal year?
• 14. Please indicate your response to the following statements by selecting whether you:
SD (Strongly Disagree), Disagree, N (are Neural), Agree, SA (Strongly Agree) • My state has a strong commitment to highway safety. • My state sufficiently funds highway/traffic enforcement. • My state sufficiently funds LEO driver training.
Miscellaneous Questions – What else effects the officer traffic fatality rate? The chart below shows the traffic fatality rate (2000-2009) for "law enforcement officers" and the "general public" by state. Please review the information for your state (and any states you normally reference for comparison purposes).
191
15. Please indicate your response to the following statements by selecting whether you
would reply: No, D (Doubtful), Maybe, P (Probably), Yes • The LEO traffic fatality rate in my state is a matter of chance. • The LEO and General Public traffic fatality rates in my state are related. • There are other state-level factors that contribute to LEO traffic fatality rates.
16. Why do you think a relationship does or does not exist between LEO traffic fatalities
and public traffic fatalities in your state? •
17. Are there state-level factors not listed in this survey that you think contribute to lower
(or higher) LEO traffic fatality rates (in your state or elsewhere)? If so, please describe these factors. If not, please feel free to share any other insights you think may be relevant, over-looked, or under-studied. NOTE - This study has only considered "policy relevant" factors (meaning that it has not considered factors beyond policy control like weather or geography). •
192
APPENDIX B
Gustafson Dataset
Variable & Description Source ST State N/A YR Year S Maximum MPH Speed Limit FHA, 2004b DT Minimum Hours of Driver Training IADLEST, 2005 MO Population of Male Officers FBI, 1995a-2009a FO Population of Female Officers LEOS Population of LEOs (MO+FO) STPOP State Population AREA Area in Square Miles US Census Bureau, 2009 A LEO Automobile Deaths FBI, 1995b-2009b M LEO Motorcycle Deaths ODMP, 2012c P LEO Pedestrian Deaths NLEOMF, 2012c LD LEO Traffic Deaths (A+M+P) LI LEO-Involved Traffic Deaths NHTSA, 2012 TTF Total Traffic Fatalities RM Rural Mileage FHA, 1995-2009 UM Urban Mileage HREC Highway Receipts ($1,000s) HSES Highway Safety/Enforcement Spending ($1,000s)
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
AL 95 70 0 8453 699 9152 4292000 50645 2 2 0 4 6 1114 8925 1887 1030837 61566 AL 96 70 0 8638 717 9355 4271000 50645 2 1 0 3 5 1146 8936 1917 1040527 72301 AL 97 70 0 8946 747 9693 4319000 50645 0 0 0 0 3 1192 8943 1919 1005516 64643 AL 98 70 0 9213 741 9954 4351000 50645 0 0 1 1 4 1071 8937 1927 1013509 64534 AL 99 70 0 9332 774 10106 4356000 50645 2 0 0 2 4 1138 8934 1935 1149923 75683 AL 00 70 0 9396 765 10161 4405485 50645 0 0 0 0 3 996 8947 1945 1262239 81907 AL 01 70 0 9370 753 10123 4451163 50645 2 0 0 2 2 991 8959 1940 1501426 90320 AL 02 70 0 8664 687 9351 3930746 50645 2 0 0 2 2 1038 8949 1944 1641691 89798 AL 03 70 0 9410 740 10150 4350590 50645 0 0 0 0 3 1004 8959 1934 1679528 103400 AL 04 70 0 8804 668 9472 4162261 50645 2 0 1 3 6 1154 9589 1989 1460330 107082 AL 05 70 0 8058 689 8747 3984466 50645 2 0 0 2 4 1131 8859 2096 1534188 66948 AL 06 70 0 9580 767 10347 4459997 50645 1 0 0 1 5 1208 8795 2183 1761284 112481 AL 07 70 0 9919 752 10671 4533379 50645 1 1 1 3 9 1110 8774 2163 1795835 121788 AL 08 70 0 9988 787 10775 4437374 50645 2 0 0 2 5 966 8751 2188 1915157 124638 AL 09 70 0 10438 807 11245 4587633 50645 1 0 0 1 1 848 8751 2188 1799268 147659 AK 95 65 0 1015 80 1095 604000 570627 0 0 0 0 0 87 5049 485 438198 22339 AK 96 65 0 1029 81 1110 606000 570627 0 0 0 0 0 81 5158 486 453482 22538 AK 97 65 0 1045 90 1135 608000 570627 1 0 0 1 0 77 5170 460 435426 22466 AK 98 65 0 983 91 1074 614000 570627 0 0 0 0 0 70 5148 459 403786 22549 AK 99 65 0 1030 100 1130 619000 570627 0 0 0 0 0 79 5112 453 415566 22744 AK 00 65 0 1016 100 1116 624094 570627 0 0 0 0 0 106 5059 451 501359 23158 AK 01 65 0 1061 104 1165 634787 570627 0 0 0 0 0 89 5225 452 481651 23866 AK 02 65 0 1052 109 1161 642955 570627 1 0 0 1 0 89 5106 451 540984 24692 AK 03 65 0 1027 108 1135 648020 570627 0 0 0 0 1 98 5059 565 697046 26268 AK 04 65 0 1089 124 1213 655435 570627 0 0 0 0 0 101 5061 574 596460 27395 AK 05 65 0 1058 116 1174 662033 570627 0 0 0 0 0 72 5083 575 599300 30017 AK 06 65 0 1097 123 1220 669380 570627 0 0 0 0 0 74 5084 590 645954 34349 AK 07 65 0 1088 116 1204 681180 570627 0 0 0 0 0 84 5059 592 652846 35856 AK 08 65 0 1144 116 1260 686293 570627 0 0 0 0 0 62 5061 589 797188 38167 AK 09 65 0 1156 106 1262 696273 570627 0 0 0 0 1 64 4765 641 935475 37597 AZ 95 65 0 7790 823 8613 4163000 113594 1 0 0 1 7 1035 5482 655 1191475 59904 AZ 96 65 0 8130 879 9009 4417000 113594 0 0 0 0 1 994 5469 674 1642548 75638 AZ 97 65 0 8396 918 9314 4545000 113594 0 0 0 0 8 951 5489 663 1406124 59525 AZ 98 65 0 8772 996 9768 4659000 113594 1 0 1 2 3 980 5840 768 1365700 53561 AZ 99 65 0 9196 1072 10268 4758000 113594 0 0 0 0 2 1024 5825 782 1818345 57637 AZ 00 75 28 9330 1104 10434 5122137 113594 2 0 0 2 5 1036 5822 787 2138935 113955 AZ 01 75 28 9671 1139 10810 5297623 113594 1 0 0 1 5 1051 5834 818 2308960 123120 AZ 02 75 28 9784 1180 10964 5443984 113594 1 1 0 2 7 1132 5967 817 2568338 113276 AZ 03 75 28 9830 1173 11003 5489285 113594 0 0 0 0 4 1118 5869 917 2428770 149816 AZ 04 75 28 10120 1197 11317 5627161 113594 1 1 0 2 0 1150 5876 938 2564249 134780 AZ 05 75 28 10364 1208 11572 5815860 113594 1 0 1 2 7 1177 5844 958 2612536 151152 AZ 06 75 28 10674 1258 11932 6153379 113594 2 2 0 4 3 1288 5838 975 2625810 159982 AZ 07 75 28 11041 1313 12354 6054541 113594 1 0 0 1 0 1066 5814 971 4787754 78232 AZ 08 75 28 11520 1380 12900 6382545 113594 1 0 0 1 1 937 5773 982 3023799 126331 AZ 09 75 28 11596 1375 12971 6482281 113594 0 0 1 1 5 807 5773 982 2879417 196566
193
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
AR 95 70
4306 354 4660 2479000 52035 0 0 0 0 1 631 15028 1228 713855 38763 AR 96 70
4530 392 4922 2505000 52035 0 0 0 0 1 615 15038 1253 757302 37916
AR 97 70
4654 401 5055 2510000 52035 3 0 0 3 2 660 14999 1307 813398 39179 AR 98 70
4844 444 5288 2538000 52035 6 1 0 7 2 625 14988 1333 773753 42740
AR 99 70
4377 486 4863 2546000 52035 0 0 0 0 1 604 15023 1344 781194 39564 AR 00 70
4406 497 4903 2673400 52035 0 0 1 1 0 652 15028 1346 1037247 36214
AR 01 70
4517 513 5030 2692090 52035 1 0 0 1 0 611 15024 1346 820533 37225 AR 02 70
4520 544 5064 2710079 52035 0 1 0 1 0 640 15030 1350 1180663 49914
AR 03 70
4596 529 5125 2702925 52035 1 0 0 1 0 640 15028 1355 1284982 55168 AR 04 70
4690 544 5234 2685784 52035 0 0 0 0 0 704 15039 1381 1072610 52032
AR 05 70
4871 564 5435 2779961 52035 1 0 1 2 1 648 15070 1374 1072746 51402 AR 06 70
4862 543 5405 2809079 52035 1 0 0 1 1 665 15044 1388 1128757 58830
AR 07 70
5006 580 5586 2832708 52035 0 0 0 0 0 650 15025 1413 1096969 62156 AR 08 70
5560 504 6064 2850876 52035 1 0 0 1 0 600 15013 1419 1084919 40021
AR 09 70
5682 518 6200 2888639 52035 0 0 0 0 1 585 15013 1419 1073797 98227 CA 95 70 24 54857 6711 61568 28185000 155769 1 0 2 3 20 4192 11383 3830 5805441 683646 CA 96 70 24 57189 7049 64238 28898000 155769 3 1 2 6 12 3989 11383 3829 6484062 793450 CA 97 70 24 58156 7420 65576 28520000 155769 3 1 1 5 5 3688 11396 3810 6613729 758371 CA 98 70 24 59518 7799 67317 29057000 155769 6 0 2 8 5 3494 11400 3806 7150626 864014 CA 99 70 24 62880 8619 71499 28741000 155769 1 1 0 2 8 3559 11398 3806 6643373 870082 CA 00 70 24 60897 8216 69113 29079921 155769 3 1 2 6 10 3753 11385 3796 7351668 938494 CA 01 70 24 62319 8621 70940 30099815 155769 2 1 1 4 13 3956 11421 3779 8660529 976279 CA 02 70 24 65011 9163 74174 30685929 155769 4 1 0 5 17 4088 11440 3843 7802188 1040803 CA 03 70 24 65526 9385 74911 30948866 155769 4 3 1 8 13 4224 11415 3810 7663992 1141513 CA 04 70 24 64511 9353 73864 31273858 155769 2 4 0 6 18 4120 11381 3828 8914456 1362248 CA 05 70 24 63540 9313 72853 31138193 155769 4 2 2 8 10 4329 11091 4124 9517205 1456251 CA 06 70 24 65804 9679 75483 31660392 155769 4 1 3 8 10 4236 10822 4413 12982120 1304420 CA 07 70 24 68521 9995 78516 31683834 155769 5 0 0 5 7 3974 10829 4440 13402278 1437103 CA 08 70 24 70594 10692 81286 31804738 155769 5 1 3 9 11 3434 10811 4393 15871564 1623277 CA 09 70 24 69749 10572 80321 31832381 155769 2 1 0 3 8 3081 10808 4384 23381301 1039052 CO 95 75 40 7579 925 8504 3353000 103643 0 0 0 0 5 645 8070 1025 932046 55135 CO 96 75 40 7936 1069 9005 3771000 103643 0 0 1 1 0 617 8070 1019 914440 63814 CO 97 75 40 8774 1259 10033 3868000 103643 1 0 0 1 1 613 8061 1020 932169 78140 CO 98 75 40 8804 1305 10109 3966000 103643 0 0 0 0 0 628 8061 1023 1278180 73441 CO 99 75 40 8449 1119 9568 3998000 103643 0 0 1 1 1 626 8059 1012 1400358 79882 CO 00 75 40 8246 1135 9381 3981192 103643 0 1 0 1 4 681 8050 1037 1958473 87717 CO 01 75 40 8990 1260 10250 4345118 103643 1 0 0 1 3 741 8062 1033 1994004 84539 CO 02 75 40 9388 1316 10704 4403008 103643 0 0 0 0 0 743 8062 1038 1893343 111067 CO 03 75 40 9254 1337 10591 4393483 103643 0 0 0 0 3 642 7693 1421 1880266 113052 CO 04 75 40 9206 1322 10528 4443898 103643 1 0 0 1 2 665 7694 1421 2259249 121924 CO 05 75 40 9245 1338 10583 4397035 103643 0 0 0 0 2 606 7694 1411 1812956 120943 CO 06 75 40 9833 1465 11298 4752554 103643 0 0 0 0 3 535 7694 1415 2049984 133190 CO 07 75 40 9939 1448 11387 4807287 103643 1 0 1 2 2 554 7693 1398 2463400 131368 CO 08 75 40 10117 1491 11608 4869988 103643 0 0 0 0 2 548 7702 1399 2069686 140060 CO 09 75 40 10286 1499 11785 5018161 103643 0 0 0 0 0 465 7702 1399 2162347 136748 CT 95 65 40 6869 532 7401 2773000 4843 0 0 0 0 0 317 1903 1812 1222053 41061 CT 96 65 40 6923 534 7457 2772000 4843 0 0 0 0 1 310 1903 1812 1413230 43961 CT 97 65 40 6992 550 7542 2769000 4843 1 0 0 1 4 339 1904 1813 1231110 43802 CT 98 65 40 7108 591 7699 2813000 4843 0 0 0 0 0 329 1903 1813 1372129 44299 CT 99 65 40 6701 582 7283 3127000 4843 0 0 0 0 0 301 1901 1816 1194190 50221 CT 00 65 40 7132 629 7761 3405565 4843 1 0 0 1 4 341 1901 1816 1269463 76797 CT 01 65 40 7280 664 7944 3425074 4843 0 0 0 0 3 318 1901 1816 1276524 81662 CT 02 65 40 7122 666 7788 3374179 4843 0 0 0 0 0 325 1291 2426 1889845 7904 CT 03 65 40 7183 675 7858 3483372 4843 1 0 0 1 2 298 1288 2430 1778716 117822 CT 04 65 40 7201 697 7898 3503604 4843 0 0 0 0 1 291 1288 2430 1764726 125558 CT 05 65 40 7138 692 7830 3510297 4843 0 0 0 0 0 274 1288 2427 1387281 126451 CT 06 65 40 7166 709 7875 3504809 4843 0 0 1 1 1 301 1288 2427 1368560 11813 CT 07 65 40 7262 711 7973 3502309 4843 0 0 0 0 1 277 1288 2428 1197021 14173 CT 08 65 40 7845 787 8632 3501252 4843 1 0 0 1 1 264 1288 2428 1255688 8642 CT 09 65 40 7830 792 8622 3518288 4843 0 0 0 0 2 223 1289 2430 1889183 14724 DE 95 65 24 1468 141 1609 544000 1949 0 0 0 0 0 121 3461 1500 429296 30208 DE 96 65 24 1438 147 1585 550000 1949 1 0 0 1 0 116 3485 1560 466438 32062 DE 97 65 24 1533 174 1707 550000 1949 0 0 0 0 0 143 3488 1559 420678 34670 DE 98 65 24 1592 176 1768 561000 1949 0 0 0 0 0 115 3494 1561 696993 36448 DE 99 65 24 1681 237 1918 572000 1949 0 0 0 0 0 100 3516 1545 463461 36742 DE 00 65 24 1892 247 2139 597530 1949 0 0 0 0 1 123 3539 1557 723178 42112 DE 01 65 24 1851 251 2102 795253 1949 0 0 1 1 0 136 3550 1573 679428 40990 DE 02 65 24 1948 258 2206 807385 1949 0 0 0 0 0 124 3563 1586 723396 38826 DE 03 65 24 1938 273 2211 817491 1949 0 0 0 0 1 142 3587 1594 1181790 44954 DE 04 65 24 1983 280 2263 830364 1949 1 0 0 1 0 134 3014 2190 666803 49381 DE 05 65 24 2020 287 2307 843524 1949 0 0 0 0 0 134 3038 2205 1067557 53594 DE 06 65 24 1950 297 2247 853476 1949 0 0 0 0 0 148 3047 2228 808666 65836 DE 07 65 24 1952 284 2236 853364 1949 0 0 0 0 0 117 3063 2245 929725 67800 DE 08 65 24 1998 298 2296 871838 1949 0 0 0 0 0 121 3076 2251 658936 74424 DE 09 65 24 1986 310 2296 884765 1949 0 0 0 0 0 116 3082 2256 889646 78641 DC 95 55 0 2826 865 3691 554000 61 0 0 0 0 0 58 0 1037 139758 9516 DC 96 55 0 2738 894 3632 543000 61 0 0 1 1 0 62 0 1371 162951 9681 DC 97 55 0 3268 955 4223 529000 61 0 0 0 0 0 60 0 1381 150878 9295 DC 98 55 0 3205 927 4132 523000 61 0 0 0 0 2 54 0 1373 259399 2327 DC 99 55 0 2637 867 3504 519000 61 0 0 0 0 0 41 0 1372 242134 0 DC 00 55 0 2767 880 3647 572059 61 0 0 0 0 1 48 0 1370 244216 0 DC 01 55 0 3049 920 3969 571822 61 0 0 0 0 0 68 0 1429 406288 0 DC 02 55 0 3084 939 4023 570898 61 0 0 1 1 0 47 0 1427 335670 0 DC 03 55 0 2995 914 3909 563384 61 0 0 0 0 1 67 0 1428 363099 0 DC 04 55 0 3195 969 4164 553523 61 0 0 0 0 1 43 0 1393 354727 0 DC 05 55 0 3210 957 4167 550521 61 0 0 0 0 0 48 0 1393 311485 0 DC 06 55 0 3233 963 4196 581530 61 0 0 0 0 0 37 0 1392 301203 0 DC 07 55 0 3320 982 4302 588292 61 1 0 0 1 1 44 0 1390 633975 0 DC 08 55 0 3455 1000 4455 591833 61 0 0 0 0 0 34 0 1390 268460 0 DC 09 55 0 3492 981 4473 599657 61 0 0 0 0 0 29 0 1391 457523 0
194
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
FL 95 70 48 26671 3500 30171 12268000 53616 1 0 1 2 6 2805 6977 4945 3229611 161385 FL 96 70 48 29346 3920 33266 13772000 53616 0 0 1 1 6 2753 7018 4906 3527847 158353 FL 97 70 48 32378 4441 36819 14017000 53616 5 0 0 5 3 2785 7007 4920 3761370 161586 FL 98 70 48 32589 4516 37105 14362000 53616 5 2 1 8 4 2825 6999 4942 4559688 178967 FL 99 70 48 30842 4308 35150 13520000 53616 5 2 0 7 6 2920 7013 4938 3966678 177124 FL 00 70 48 34005 5028 39033 15519361 53616 1 0 0 1 19 2999 6997 4963 4081420 201661 FL 01 70 48 34482 5108 39590 15916614 53616 3 0 2 5 11 3012 7056 4994 4485840 218165 FL 02 70 48 36158 5353 41511 16401547 53616 2 0 0 2 19 3136 7060 4999 4844042 224352 FL 03 70 48 36278 5859 42137 16516256 53616 1 1 2 4 14 3169 7058 4995 7086647 279648 FL 04 70 48 37869 6168 44037 17395608 53616 7 0 1 8 11 3244 5988 6059 5995828 337914 FL 05 70 48 36447 5995 42442 17661970 53616 1 0 0 1 4 3543 5969 6071 7991055 334773 FL 06 70 48 39192 6531 45723 18002234 53616 1 0 1 2 7 3374 5969 6101 7804531 388174 FL 07 70 48 38409 6251 44660 18251243 53616 3 0 4 7 6 3214 5966 6096 9294988 440166 FL 08 70 48 38114 6362 44476 18297732 53616 4 0 0 4 10 2978 5968 6116 8392721 405405 FL 09 70 48 37894 6224 44118 17648382 53616 1 0 0 1 16 2558 5968 6116 6198950 420253 GA 95 70 24 16226 2187 18413 6326000 57507 0 0 0 0 6 1488 14833 2945 1372068 79995 GA 96 70 24 17140 2322 19462 6493000 57507 2 0 0 2 8 1573 14843 2949 1457968 83742 GA 97 70 24 16599 2598 19197 6378000 57507 1 0 0 1 6 1577 14843 2954 1569007 201190 GA 98 70 24 18353 2911 21264 7565000 57507 2 0 0 2 1 1568 14849 2974 1782979 94116 GA 99 70 24 15840 2239 18079 6191000 57507 2 0 0 2 4 1508 14859 2979 1769962 97228 GA 00 70 24 17858 2608 20466 7246786 57507 3 0 1 4 9 1541 14859 2979 1852170 58508 GA 01 70 24 18433 2661 21094 7692407 57507 0 0 0 0 3 1647 14861 3021 1617121 99105 GA 02 70 24 16336 2417 18753 7226657 57507 1 0 1 2 4 1524 14861 3003 1929521 98794 GA 03 70 24 18993 2992 21985 7959233 57507 4 1 0 5 11 1603 14860 3012 1992015 293316 GA 04 70 24 18377 2893 21270 8125492 57507 1 1 1 3 3 1634 13967 3974 2276872 140342 GA 05 70 24 17924 2872 20796 8209878 57507 3 0 0 3 7 1729 13950 3979 1831045 144696 GA 06 70 24 18995 3167 22162 8663117 57507 3 0 0 3 3 1693 13925 3986 2395342 146660 GA 07 70 24 17594 3391 20985 7708266 57507 2 0 0 2 8 1641 13928 3987 3607779 141642 GA 08 70 24 20514 3775 24289 8931253 57507 2 0 1 3 5 1493 14011 3985 3402038 222557 GA 09 70 24 21107 3921 25028 9154201 57507 1 1 0 2 4 1284 13995 3989 3615074 196855 HI 95 60 0 2331 186 2517 1187000 6423 0 0 0 0 0 130 779 340 477347 8077 HI 96 60 0 2388 203 2591 1184000 6423 0 0 0 0 0 148 765 295 424335 5985 HI 97 60 0 2422 223 2645 1146000 6423 1 0 0 1 1 131 717 261 351626 5508 HI 98 60 0 2431 157 2588 1193000 6423 0 0 0 0 0 120 716 259 255036 5351 HI 99 60 0 2481 246 2727 1185000 6423 0 0 1 1 0 98 723 259 345297 5965 HI 00 60 0 2615 249 2864 1211537 6423 0 0 0 0 0 132 682 263 226138 5326 HI 01 60 0 2495 263 2758 1224398 6423 0 0 1 1 1 140 682 263 355448 3837 HI 02 60 0 2530 269 2799 1244898 6423 0 0 0 0 0 119 672 261 349711 6131 HI 03 60 0 2553 263 2816 1257608 6423 0 1 0 1 2 133 675 261 359948 4837 HI 04 60 0 2447 265 2712 1262840 6423 1 0 0 1 2 142 675 266 286934 7695 HI 05 60 0 2547 275 2822 1275194 6423 0 0 0 0 0 140 567 363 546369 8218 HI 06 60 0 2617 279 2896 1285498 6423 0 1 0 1 0 161 568 361 313203 6715 HI 07 60 0 2596 285 2881 1283388 6423 0 0 0 0 0 138 569 370 269461 7000 HI 08 60 0 2687 310 2997 1288198 6423 0 0 0 0 0 107 580 367 403363 7359 HI 09 60 0 2684 306 2990 1295178 6423 0 0 0 0 0 109 580 366 532415 9390 ID 95 75 36 1938 120 2058 1161000 82643 1 0 0 1 1 262 4698 272 357133 18421 ID 96 75 36 1966 142 2108 1181000 82643 0 0 0 0 1 258 4684 267 391189 19964 ID 97 75 36 2063 131 2194 1209000 82643 0 0 0 0 0 259 4683 280 438841 19215 ID 98 75 36 2112 135 2247 1229000 82643 0 0 0 0 0 265 4665 296 410786 18503 ID 99 75 36 2167 160 2327 1252000 82643 0 0 0 0 1 278 4660 297 473902 19746 ID 00 75 36 2194 132 2326 1271895 82643 0 0 0 0 0 276 4661 294 504630 21877 ID 01 75 36 2246 136 2382 1305535 82643 0 0 0 0 0 259 4660 294 442119 23030 ID 02 75 36 2223 135 2358 1311796 82643 0 0 0 0 0 264 4661 293 533384 21567 ID 03 75 36 2281 144 2425 1348560 82643 1 0 0 1 3 293 4654 300 555684 22371 ID 04 75 36 2289 155 2444 1384946 82643 0 0 0 0 2 260 4623 328 572704 34471 ID 05 75 36 2362 161 2523 1426766 82643 0 0 0 0 2 275 4628 329 600327 29803 ID 06 75 36 2422 172 2594 1464850 82643 0 0 0 0 0 267 4632 327 831511 31198 ID 07 75 36 2484 174 2658 1496944 82643 0 0 0 0 0 252 4636 322 797811 40475 ID 08 75 36 2512 174 2686 1519310 82643 0 0 0 0 0 232 4632 325 802706 42481 ID 09 75 36 2511 179 2690 1543324 82643 2 0 0 2 4 226 4604 341 869170 37988 IL 95 65 0 28855 3601 32456 11796000 55519 1 0 0 1 1 1586 12113 4709 3135784 127158 IL 96 65 0 28705 3681 32386 11797000 55519 0 0 0 0 6 1477 12059 4701 3123175 122152 IL 97 65 0 29232 3867 33099 11843000 55519 1 0 2 3 6 1397 11942 4502 2900028 129046 IL 98 65 0 29633 4109 33742 12023000 55519 0 0 0 0 4 1393 11928 4495 3411514 136186 IL 99 65 0 29599 4290 33889 12078000 55519 0 0 0 0 3 1456 11876 4476 3041122 137971 IL 00 65 0 31553 5091 36644 12380526 55519 1 0 0 1 3 1418 11811 4487 3860474 143915 IL 01 65 0 31261 5193 36454 12398330 55519 0 0 0 0 9 1414 11718 4530 3753447 157064 IL 02 65 0 31109 5280 36389 12542030 55519 1 0 0 1 8 1420 11667 4521 4193329 164224 IL 03 65 0 31176 5397 36573 12590504 55519 3 0 0 3 8 1454 11620 4542 4095027 227340 IL 04 65 0 30889 5543 36432 12645893 55519 3 0 0 3 8 1356 11474 4650 4143963 212687 IL 05 65 0 30964 5572 36536 12720852 55519 2 0 0 2 6 1361 11323 4780 5149960 216402 IL 06 65 0 31478 5751 37229 12793900 55519 6 0 0 6 8 1254 10992 5091 5556523 231920 IL 07 65 0 31524 5778 37302 12813230 55519 1 0 0 1 3 1249 10959 5096 5895475 226899 IL 08 65 0 31496 5752 37248 12846505 55519 0 0 0 0 1 1043 10936 5104 5963993 214678 IL 09 65 0 30663 5574 36237 12675815 55519 0 0 0 0 1 911 10914 5110 4988828 257528 IN 95 65 40 8359 639 8998 5386000 35823 0 0 0 0 5 960 9569 1743 1428518 54719 IN 96 65 40 8496 676 9172 5285000 35823 1 0 0 1 5 984 9553 1741 1425531 59965 IN 97 65 40 9067 706 9773 5545000 35823 0 0 2 2 1 935 9559 1743 1596133 54818 IN 98 65 40 9970 811 10781 5813000 35823 1 0 1 2 6 982 9564 1735 1869973 54329 IN 99 65 40 9338 731 10069 5651000 35823 2 0 1 3 2 1020 9553 1669 1675527 54118 IN 00 65 40 9341 822 10163 5582518 35823 1 0 1 2 4 886 9554 1663 1978156 62928 IN 01 65 40 9609 758 10367 5897947 35823 3 0 0 3 6 909 9539 1654 3223021 60756 IN 02 65 40 9945 797 10742 6049242 35823 1 0 0 1 2 792 9531 1654 1942600 137211 IN 03 65 40 10008 798 10806 6122366 35823 0 0 0 0 2 833 9534 1652 2452528 203115 IN 04 65 40 9963 806 10769 5991832 35823 1 0 0 1 4 947 9534 1652 2614404 217923 IN 05 65 40 9839 778 10617 6043427 35823 3 0 0 3 4 938 9531 1651 2099441 121202 IN 06 65 40 9449 748 10197 5570912 35823 0 0 2 2 2 899 9531 1652 2142033 121252 IN 07 65 40 9944 766 10710 5961949 35823 1 0 1 2 1 898 9358 1831 3737847 2293 IN 08 65 40 9690 770 10460 5623508 35823 1 0 1 2 1 814 8942 2273 3068215 3259 IN 09 65 40 10326 907 11233 5995956 35823 2 0 0 2 3 693 8841 2334 3068214 9901
195
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
IA 95 65 20 4415 266 4681 2837000 55858 0 0 0 0 0 527 8822 879 1073287 77643 IA 96 65 20 4336 369 4705 2850000 55858 0 0 0 0 0 465 8797 870 1126981 81980 IA 97 65 20 4514 291 4805 2850000 55858 0 0 0 0 0 468 8789 882 1161914 72304 IA 98 65 20 4644 320 4964 2859000 55858 0 0 0 0 0 449 8799 883 1115936 78276 IA 99 65 20 4733 326 5059 2869000 55858 0 0 0 0 1 490 8834 882 1274354 82249 IA 00 65 20 4713 330 5043 2917550 55858 0 0 0 0 0 445 9209 1035 1410210 96266 IA 01 65 20 4732 332 5064 2915711 55858 0 0 0 0 0 446 8837 891 1377477 61783 IA 02 65 20 4709 344 5053 2916660 55858 0 0 0 0 0 405 9200 1055 1405164 74780 IA 03 65 20 4658 345 5003 2944062 55858 0 0 1 1 0 443 7920 961 1416436 80309 IA 04 65 20 4598 361 4959 2943984 55858 0 0 0 0 0 390 7922 958 1492383 113547 IA 05 65 20 4641 368 5009 2966334 55858 0 0 0 0 0 450 7938 956 1445811 113718 IA 06 65 20 4667 373 5040 2961632 55858 0 0 0 0 0 439 7952 957 1450978 121396 IA 07 65 20 4760 394 5154 2976754 55858 1 0 0 1 0 445 7929 956 1562402 113928 IA 08 65 20 4770 398 5168 2967083 55858 0 0 0 0 0 412 7941 953 1613010 117728 IA 09 65 20 4842 439 5281 2988922 55858 0 0 0 0 0 372 7936 956 1673684 122057 KS 95 70 34 5625 429 6054 2484000 81759 1 0 1 2 0 442 10008 672 907779 39933 KS 96 70 34 5761 436 6197 2492000 81759 0 0 0 0 0 490 9999 659 978363 40520 KS 97 70 34 5888 473 6361 2518000 81759 0 0 0 0 1 482 9800 610 968053 40510 KS 98 70 34 6069 503 6572 2550000 81759 0 0 1 1 1 492 9775 610 1246667 43080 KS 99 70 34 6051 515 6566 2570000 81759 0 0 0 0 2 540 9781 606 1441456 41586 KS 00 70 34 6064 492 6556 2674069 81759 0 0 0 0 0 461 9779 605 1440473 50704 KS 01 70 34 6178 516 6694 2618694 81759 0 0 0 0 0 494 9776 606 1095156 50629 KS 02 70 34 6276 511 6787 2691202 81759 0 0 0 0 0 507 9784 596 1925421 73869 KS 03 70 34 6330 573 6903 2699012 81759 1 0 0 1 1 469 9756 623 1611428 56398 KS 04 70 34 6504 640 7144 2688942 81759 0 0 0 0 0 461 9722 651 1796097 60782 KS 05 70 34 6471 614 7085 2698592 81759 0 0 0 0 0 428 9639 730 1151863 64595 KS 06 70 34 6469 616 7085 2678243 81759 0 0 0 0 0 468 9622 744 1424308 86666 KS 07 70 34 6567 654 7221 2691761 81759 0 0 0 0 0 416 9619 749 1496034 73460 KS 08 70 34 6198 613 6811 2550708 81759 0 0 0 0 0 385 9618 750 1454312 81513 KS 09 70 34 6151 610 6761 2498126 81759 0 0 0 0 1 386 9618 750 1408560 79841 KY 95 65 40 6452 625 7077 3801000 39491 0 0 0 0 0 849 24988 2392 1509132 53993 KY 96 65 40 6223 612 6835 3795000 39491 1 0 0 1 1 842 25011 2389 1274849 53472 KY 97 65 40 6042 556 6598 3491000 39491 0 0 1 1 0 857 25031 2387 1289873 54610 KY 98 65 40 6890 675 7565 3879000 39491 0 0 0 0 2 858 25043 2400 1444379 58435 KY 99 65 40 7000 665 7665 3901000 39491 0 0 0 0 2 814 25074 2404 1452514 53137 KY 00 65 40 6897 634 7531 3990782 39491 1 0 0 1 4 820 25052 2421 1670428 48602 KY 01 65 40 7022 629 7651 4025105 39491 0 0 0 0 1 845 25057 2423 1635590 48947 KY 02 65 40 7109 610 7719 4068895 39491 2 0 0 2 2 915 25060 2424 1839848 48955 KY 03 65 40 7185 585 7770 4075332 39491 1 0 0 1 8 928 24995 2503 1827568 57476 KY 04 65 40 7083 572 7655 4087276 39491 0 0 0 0 1 964 25089 2420 1651837 54496 KY 05 65 40 7317 589 7906 4167174 39491 0 0 0 0 4 985 25087 2423 2322045 61502 KY 06 65 40 7473 555 8028 4106153 39491 1 0 0 1 5 913 25102 2427 2134688 44069 KY 07 65 40 7524 545 8069 4210320 39491 1 0 0 1 2 864 25121 2426 2066238 65349 KY 08 65 40 7275 534 7809 4141804 39491 0 0 0 0 4 826 25142 2433 1993161 65591 KY 09 65 40 7700 548 8248 4239650 39491 0 0 0 0 0 791 25148 2429 2247403 84310 LA 95 70 0 11617 1989 13606 4245000 43203 2 0 0 2 2 894 14643 2011 1239851 85905 LA 96 70 0 11734 2181 13915 4254000 43203 0 1 1 2 2 902 14646 2029 1371303 89563 LA 97 70 0 12127 2390 14517 4108000 43203 1 0 0 1 2 931 14643 2027 1351891 88195 LA 98 70 0 12801 2556 15357 4288000 43203 3 1 0 4 4 926 14657 2029 1428139 94249 LA 99 70 0 13713 2836 16549 4295000 43203 6 0 2 8 0 938 14661 2038 1240903 96420 LA 00 70 0 13776 3098 16874 4386553 43203 2 1 0 3 3 938 14660 2036 1262042 76066 LA 01 70 0 13333 3022 16355 4388774 43203 0 0 0 0 3 952 14665 2040 1274345 82530 LA 02 70 0 13558 3399 16957 4356611 43203 0 0 1 1 2 907 14654 2041 1281414 54737 LA 03 70 0 13403 2922 16325 4306455 43203 3 0 1 4 3 940 14653 2039 1744184 137442 LA 04 70 0 13477 3086 16563 4279321 43203 0 0 1 1 1 904 13857 2838 1453826 158692 LA 05 70 0 11067 2555 13622 3713890 43203 2 0 0 2 1 955 13212 3480 1332234 13808 LA 06 70 0 11983 2796 14779 3792516 43203 0 1 1 2 2 982 13203 3484 1777878 56336 LA 07 70 0 12577 2556 15133 3983159 43203 3 0 0 3 14 985 13202 3481 3430908 57390 LA 08 70 0 13960 3346 17306 4293435 43203 0 0 0 0 5 912 13199 3488 2802867 64885 LA 09 70 0 9215 1925 11140 2771692 43203 1 0 0 1 1 821 13196 3481 2637567 80226 ME 95 65 47 1921 90 2011 1231000 30854 0 0 0 0 0 187 7593 787 379619 33179 ME 96 65 47 1919 99 2018 1237000 30854 1 0 0 1 2 169 7598 786 456979 30387 ME 97 65 47 1938 111 2049 1239000 30854 0 0 0 0 2 192 7595 769 452021 27950 ME 98 65 47 2001 116 2117 1242000 30854 0 0 0 0 2 192 7616 772 500597 23796 ME 99 65 47 2016 122 2138 1250000 30854 1 0 0 1 0 181 7619 773 413718 24185 ME 00 65 47 2046 126 2172 1271363 30854 0 0 0 0 0 169 7632 774 751571 27419 ME 01 65 47 2059 126 2185 1280138 30854 0 0 0 0 0 192 7635 770 430691 57421 ME 02 65 47 2062 133 2195 1291698 30854 0 0 0 0 0 216 7636 773 693060 94649 ME 03 65 47 2082 132 2214 1305728 30854 0 0 0 0 0 207 7636 772 557244 71873 ME 04 65 47 2060 134 2194 1314897 30854 0 0 0 0 0 194 7670 819 741658 75865 ME 05 65 47 2084 129 2213 1320239 30854 0 0 0 0 0 169 7557 992 613036 37703 ME 06 65 47 2076 126 2202 1321574 30854 0 0 0 0 0 188 7556 991 578954 43280 ME 07 65 47 2125 136 2261 1294381 30854 0 0 0 0 0 183 7527 992 643877 44178 ME 08 65 47 2107 138 2245 1315497 30854 0 0 0 0 0 155 7516 992 675576 38118 ME 09 65 47 2117 134 2251 1317341 30854 0 0 0 0 0 159 7511 990 696209 34678 MD 95 65
11777 1609 13386 4917000 9706 1 0 0 1 0 671 3623 1595 1376063 119010
MD 96 65
11771 1653 13424 4962000 9706 0 1 0 1 2 608 3623 1595 1397361 119973 MD 97 65
12003 1737 13740 4878000 9706 0 0 0 0 2 611 3589 1553 1458540 122279
MD 98 65
12366 1808 14174 4972000 9706 2 0 0 2 7 606 3582 1547 1513518 117449 MD 99 65
12292 1840 14132 5009000 9706 0 0 0 0 4 590 3578 1542 1609225 142202
MD 00 65
12474 1930 14404 5129364 9706 5 0 0 5 5 588 3583 1547 1669347 145187 MD 01 65
12638 2012 14650 5211141 9706 0 0 0 0 0 659 3585 1546 1743008 155693
MD 02 65
12763 2064 14827 5291592 9706 3 0 0 3 0 661 3585 1545 1942295 167042 MD 03 65
12740 2048 14788 5338977 9706 2 0 0 2 0 650 3099 2037 1795063 151289
MD 04 65
12817 2080 14897 5556884 9706 1 0 1 2 1 643 3095 2040 1906115 131433 MD 05 65
12885 2047 14932 5599201 9706 0 0 0 0 0 614 3095 2045 1951391 126438
MD 06 65
13006 2042 15048 5613488 9706 0 0 0 0 1 651 3097 2052 2352935 159330 MD 07 65
13396 2069 15465 5439792 9706 0 0 2 2 0 614 3099 2051 2598849 157041
MD 08 65
13538 2081 15619 5456159 9706 1 0 1 2 3 591 3098 2050 3120406 106289 MD 09 65
13584 2080 15664 5519662 9706 0 0 0 0 1 547 3101 2051 2905587 120696
196
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
MA 95 65 40 14481 1021 15502 6012000 7801 2 0 0 2 4 444 1277 1751 2478230 144031 MA 96 65 40 14745 1131 15876 6011000 7801 0 0 0 0 1 417 1266 1738 2521134 154181 MA 97 65 40 14935 1182 16117 6040000 7801 0 0 1 1 1 441 1197 1669 4346076 145594 MA 98 65 40 15227 1180 16407 6101000 7801 0 0 0 0 0 406 1207 1671 2811522 139942 MA 99 65 40 15190 1276 16466 6037000 7801 0 0 1 1 1 414 1210 1651 4091702 194414 MA 00 65 40 15325 1284 16609 6289896 7801 0 0 2 2 0 433 1209 1641 3468038 203788 MA 01 65 40 15398 1306 16704 6324198 7801 0 0 1 1 1 477 1204 1640 3755325 234046 MA 02 65 40 15132 1293 16425 6268238 7801 0 0 0 0 0 459 1204 1640 3830472 193489 MA 03 65 40 14950 1324 16274 6359754 7801 0 0 0 0 1 462 704 2131 3463264 190622 MA 04 65 40 14809 1315 16124 6340152 7801 0 0 0 0 0 476 707 2135 3239069 177057 MA 05 65 40 14940 1346 16286 6213692 7801 1 1 0 2 2 442 709 2140 2742642 189779 MA 06 65 40 15656 1376 17032 6378563 7801 0 0 0 0 2 430 709 2121 2180291 213402 MA 07 65 40 15479 1348 16827 6366891 7801 1 0 0 1 2 417 709 2125 2975946 254172 MA 08 65 40 15250 1359 16609 6281021 7801 0 0 0 0 3 363 708 2126 2766980 246501 MA 09 65 40 12802 1153 13955 6396251 7801 1 0 1 2 2 334 753 2242 2751798 180837 MI 95 70 32 17250 2128 19378 9492000 56538 3 0 0 3 5 1530 7641 1999 1820995 174235 MI 96 70 32 17600 2301 19901 9536000 56538 1 0 0 1 6 1505 7627 1993 1904397 197658 MI 97 70 32 17991 2449 20440 9673000 56538 1 0 0 1 8 1446 7636 1994 2170808 210984 MI 98 70 32 18235 2570 20805 9718000 56538 0 0 0 0 2 1366 7737 2023 2739838 194767 MI 99 70 32 18053 2570 20623 9708000 56538 1 0 1 2 1 1382 7700 2026 2553633 210584 MI 00 70 32 18231 2727 20958 9905097 56538 2 0 2 4 2 1382 7695 2018 2815272 218810 MI 01 70 32 18589 2898 21487 9949888 56538 0 0 1 1 4 1328 7703 2023 3080746 239677 MI 02 70 32 18175 2831 21006 9976197 56538 0 0 0 0 1 1277 7695 2017 2915676 260719 MI 03 70 32 17756 2802 20558 10043587 56538 2 0 0 2 6 1283 7144 2597 2647045 228512 MI 04 70 32 17418 2802 20220 10071760 56538 3 0 0 3 2 1159 7124 2597 3064926 213440 MI 05 70 32 17001 2681 19682 10006482 56538 3 0 0 3 9 1129 7087 2610 3438641 211726 MI 06 70 32 16638 2590 19228 9825520 56538 0 0 0 0 1 1085 7088 2608 3283823 222876 MI 07 70 32 16835 2645 19480 10018999 56538 0 0 0 0 3 1088 7068 2604 3640590 227879 MI 08 70 32 17010 2494 19504 9948437 56538 0 0 0 0 3 980 7056 2597 2839864 238939 MI 09 70 32 16302 2498 18800 9845506 56538 0 0 0 0 1 871 7060 2598 3253154 253184 MN 95 70 40 6279 499 6778 4314000 79628 0 0 0 0 1 597 10848 1176 1284864 68965 MN 96 70 40 6746 631 7377 4518000 79628 2 0 0 2 1 576 10822 1174 1340482 69403 MN 97 70 40 6833 671 7504 4518000 79628 0 0 0 0 2 600 10821 1142 1452076 76281 MN 98 70 40 6968 728 7696 4586000 79628 0 0 0 0 4 650 10816 1144 1454510 77927 MN 99 70 40 7012 787 7799 4599000 79628 1 0 0 1 1 626 10807 1132 1507315 85286 MN 00 70 40 7085 819 7904 4751560 79628 0 0 1 1 1 625 10809 1118 1731026 89268 MN 01 70 40 7154 861 8015 4802774 79628 0 0 0 0 2 568 10829 1129 1815670 89713 MN 02 70 40 7220 884 8104 4859720 79628 2 0 0 2 6 657 10806 1112 1962094 90472 MN 03 70 40 7124 912 8036 4881448 79628 0 0 0 0 1 655 10806 1123 1781312 94300 MN 04 70 40 7234 913 8147 4960204 79628 1 0 0 1 0 567 10709 1126 2055235 91707 MN 05 70 40 7367 941 8308 4992123 79628 0 0 0 0 2 559 10737 1135 2057780 95758 MN 06 70 40 7509 984 8493 5030630 79628 0 0 0 0 1 494 10781 1145 2161259 91757 MN 07 70 40 7676 1038 8714 5065797 79628 0 0 1 1 2 504 10729 1153 2779384 113264 MN 08 70 40 7805 1063 8868 5145905 79628 0 0 0 0 1 456 10539 1355 2361119 109777 MN 09 70 40 7806 1076 8882 5180883 79628 0 0 1 1 1 421 10539 1355 2269757 155139 MS 95 70 0 3390 232 3622 1754000 46923 0 0 1 1 0 868 9735 847 711099 51122 MS 96 70 0 4305 274 4579 2436000 46923 0 0 0 0 0 811 9804 854 725664 47361 MS 97 70 0 4430 357 4787 2343000 46923 1 0 0 1 1 861 9787 849 766118 58315 MS 98 70 0 4913 401 5314 2651000 46923 1 0 0 1 2 948 9798 852 801478 56293 MS 99 70 0 4425 365 4790 2249000 46923 1 0 0 1 0 927 9785 860 1152532 49810 MS 00 70 0 4623 377 5000 2345430 46923 1 0 1 2 1 949 9742 909 926906 61021 MS 01 70 0 4717 377 5094 2380448 46923 1 0 1 2 0 784 9750 911 888316 52986 MS 02 70 0 4862 415 5277 2426944 46923 1 0 0 1 2 885 9775 901 1153147 50770 MS 03 70 0 4443 410 4853 2220219 46923 2 0 0 2 0 872 9583 1329 984909 47632 MS 04 70 0 5071 456 5527 2573607 46923 0 0 1 1 0 900 9560 1327 957424 51475 MS 05 70 0 4718 418 5136 2441159 46923 0 0 0 0 0 931 9565 1330 1151850 57836 MS 06 70 0 5158 466 5624 2507121 46923 0 1 0 1 0 911 9635 1335 1326604 20891 MS 07 70 0 4782 440 5222 2375636 46923 2 0 0 2 0 884 9604 1354 1689990 21034 MS 08 70 0 4648 429 5077 2489444 46923 0 1 0 1 1 783 9563 1408 1370997 23775 MS 09 70 0 4924 493 5417 2675080 46923 2 0 0 2 2 700 9507 1393 1215875 36874 MO 95 70 16 9883 937 10820 5142000 68739 1 0 0 1 2 1109 30652 1716 1308605 99990 MO 96 70 16 10125 973 11098 5245000 68739 1 0 0 1 4 1148 30650 1733 1413364 124500 MO 97 70 16 10391 1023 11414 5306000 68739 1 0 1 2 2 1192 30649 1737 1475671 110850 MO 98 70 16 10698 1115 11813 5430000 68739 0 0 1 1 2 1169 30646 1755 1455383 111701 MO 99 70 16 10449 1114 11563 5178000 68739 1 0 0 1 1 1094 30653 1753 1587419 108570 MO 00 70 16 10100 1067 11167 4961039 68739 0 0 0 0 4 1157 30653 1753 2038239 123437 MO 01 70 16 10692 1133 11825 5497044 68739 3 1 0 4 3 1098 30670 1754 2014095 126523 MO 02 70 16 11960 1242 13202 5604305 68739 1 0 2 3 5 1208 30682 1764 2409619 126823 MO 03 70 16 12119 1256 13375 5683974 68739 2 0 0 2 1 1232 30682 1764 1916074 124997 MO 04 70 16 12184 1266 13450 5732783 68739 1 0 0 1 3 1130 30553 1919 2148679 132253 MO 05 70 16 11671 1248 12919 5607484 68739 2 0 2 4 4 1257 30525 1940 1992521 155863 MO 06 70 16 12265 1319 13584 5790380 68739 0 0 0 0 1 1096 31404 2276 2418823 175235 MO 07 70 16 12740 1382 14122 5835175 68739 3 0 0 3 6 992 30803 2883 4792485 200714 MO 08 70 16 13039 1398 14437 5860645 68739 0 0 0 0 1 960 30999 2678 2870403 193413 MO 09 70 16 13218 1420 14638 5874396 68739 1 0 1 2 4 878 30796 2842 2496750 204395 MT 95 75 40 1358 70 1428 850000 145544 0 0 0 0 1 215 7639 167 345642 26422 MT 96 75 40 1382 75 1457 877000 145544 0 0 0 0 0 200 6606 167 383992 23400 MT 97 75 40 1415 75 1490 880000 145544 0 0 0 0 0 265 7633 158 376264 25104 MT 98 75 40 1422 82 1504 880000 145544 0 0 0 0 0 237 7620 160 385033 23829 MT 99 75 40 1421 87 1508 883000 145544 0 0 0 0 0 220 6541 160 435175 24605 MT 00 75 40 1327 74 1401 899802 145544 0 0 0 0 0 237 6549 160 484248 26495 MT 01 75 40 1432 88 1520 902713 145544 0 0 0 0 0 230 7677 181 478908 28149 MT 02 75 40 1488 93 1581 909453 145544 0 0 0 0 0 269 7694 181 544116 29881 MT 03 75 40 1490 97 1587 916711 145544 1 0 0 1 0 262 7684 197 572426 43179 MT 04 75 40 1524 102 1626 926865 145544 0 0
0 0 229 7682 198 550849 44194
MT 05 75 40 1505 108 1613 933609 145544 0 0 0 0 1 251 10493 297 664953 36916 MT 06 75 40 1567 112 1679 943628 145544 2 0 0 2 2 263 10484 297 560629 39472 MT 07 75 40 1625 109 1734 957619 145544 1 0 0 1 1 277 10488 299 626697 38666 MT 08 75 40 1545 106 1651 931062 145544 1 0 0 1 3 229 10497 299 671050 48290 MT 09 75 40 1678 102 1780 972240 145544 1 0 0 1 2 221 10497 299 663530 49888
197
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
NE 95 75
2732 260 2992 1634000 76825 0 0 0 0 0 254 9620 337 567461 28261 NE 96 75
2803 275 3078 1635000 76825 0 0 0 0 0 293 9640 327 641274 27643
NE 97 75
2805 300 3105 1641000 76825 0 0 0 0 0 302 9647 329 616691 31070 NE 98 75
2859 311 3170 1660000 76825 0 0 0 0 0 315 9650 328 636817 31713
NE 99 75
2862 320 3182 1662000 76825 1 0 0 1 0 295 9645 326 649580 33917 NE 00 75
2929 322 3251 1695317 76825 0 0 0 0 0 276 9644 326 718604 34566
NE 01 75
2971 329 3300 1694176 76825 0 0 0 0 0 246 9664 329 694186 41052 NE 02 75
3043 343 3386 1719618 76825 0 0 0 0 0 307 9653 327 858120 52881
NE 03 75
3022 352 3374 1686860 76825 0 0 0 0 0 293 9650 337 824208 58719 NE 04 75
3070 373 3443 1714366 76825 0 0 0 0 0 254 9602 378 820567 67879
NE 05 75
3017 369 3386 1687588 76825 0 0 0 0 0 276 9592 383 893426 72802 NE 06 75
3078 375 3453 1730689 76825 0 0 0 0 0 269 9550 406 858875 69851
NE 07 75
2988 361 3349 1700074 76825 0 0 0 0 0 256 9547 406 1504043 82167 NE 08 75
3106 385 3491 1770070 76825 0 0 0 0 0 208 9551 406 1360399 81543
NE 09 75
3124 394 3518 1750280 76825 1 0 0 1 0 223 9551 402 1351503 78146 NV 95 75
3345 402 3747 1529000 109782 0 0 0 0 1 313 4578 532 438032 31399
NV 96 75
3493 427 3920 1595000 109782 0 0 0 0 1 348 4578 533 451806 34671 NV 97 75
3678 440 4118 1677000 109782 0 1 0 1 1 347 4630 534 486862 37904
NV 98 75
3477 398 3875 1747000 109782 0 0 0 0 1 361 4894 577 521923 40664 NV 99 75
4030 481 4511 1809000 109782 0 0 0 0 0 350 4900 557 541611 48463
NV 00 75
4303 511 4814 1998257 109782 0 0 0 0 4 323 4901 556 535339 54201 NV 01 75
4082 417 4499 2106074 109782 1 0 0 1 1 314 4900 547 732738 61947
NV 02 75
4397 510 4907 2173491 109782 0 1 0 1 2 381 4899 550 660172 56303 NV 03 75
4220 469 4689 2241154 109782 0 0 0 0 2 368 4901 550 1000154 63157
NV 04 75
4292 466 4758 2334771 109782 1 0 0 1 2 395 4901 550 875465 78999 NV 05 75
4445 521 4966 2414807 109782 0 0 0 0 5 427 4694 705 926373 81659
NV 06 75
4748 514 5262 2495529 109782 0 0 0 0 9 432 4745 636 1184838 91902 NV 07 75
5718 677 6395 2565382 109782 1 0 0 1 2 373 4744 639 1359940 103932
NV 08 75
5194 534 5728 2600167 109782 1 0 0 1 1 324 4737 642 979623 94487 NV 09 75
5410 574 5984 2643085 109782 2 0 0 2 4 243 4758 627 1091846 115459
NH 95 65 40 1793 102 1895 886000 8952 0 0 0 0 0 118 3585 404 325148 32935 NH 96 65 40 1952 122 2074 946000 8952 0 0 0 0 0 134 3582 403 319098 33508 NH 97 65 40 1967 117 2084 947000 8952 0 0 0 0 0 125 3579 406 353845 33129 NH 98 65 40 2083 122 2205 1007000 8952 0 0 0 0 0 128 3545 407 347695 36603 NH 99 65 40 2178 138 2316 1049000 8952 0 0 0 0 0 140 3593 411 485797 46418 NH 00 65 40 1759 106 1865 879572 8952 0 0 0 0 0 126 3575 410 396279 42759 NH 01 65 40 1892 129 2021 968701 8952 0 0 0 0 0 142 3589 412 413024 52074 NH 02 65 40 1794 123 1917 925055 8952 0 0 0 0 0 127 3580 414 533594 55260 NH 03 65 40 1906 130 2036 980388 8952 0 0 0 0 0 127 3674 441 480499 56258 NH 04 65 40 1857 148 2005 985947 8952 0 0 0 0 0 171 3674 441 450029 54680 NH 05 65 40 2175 173 2348 1124982 8952 0 0 0 0 0 166 3216 759 450029 54680 NH 06 65 40 2228 186 2414 1156338 8952 0 0 0 0 0 127 3220 760 536089 54745 NH 07 65 40 2287 179 2466 1167387 8952 0 0 0 0 0 129 3218 772 740228 61436 NH 08 65 40 2367 199 2566 1156970 8952 0 0 0 0 0 139 3207 764 737335 61565 NH 09 65 40 2349 192 2541 1166104 8952 0 0 0 0 0 110 3207 764 705583 68464 NJ 95 65 0 26239 1458 27697 7620000 7354 2 0 0 2 7 774 869 1564 1776002 184096 NJ 96 65 0 26429 1569 27998 7636000 7354 1 0 0 1 1 814 893 1522 3667419 206814 NJ 97 65 0 27482 1712 29194 7729000 7354 2 0 0 2 4 775 872 1442 2363788 192310 NJ 98 65 0 27725 1795 29520 7849000 7354 0 0 0 0 3 741 875 1446 2388723 248984 NJ 99 65 0 27899 1887 29786 7889000 7354 1 1 0 2 6 726 875 1445 3066407 240103 NJ 00 65 0 28118 2048 30166 8150729 7354 0 0 0 0 3 731 871 1437 5970119 221221 NJ 01 65 0 28322 2128 30450 8233372 7354 1 0 0 1 10 745 871 1440 5018314 190730 NJ 02 65 0 28331 2152 30483 8331239 7354 0 0 1 1 2 771 868 1442 4748124 198614 NJ 03 65 0 28918 2377 31295 8358119 7354 1 1 0 2 2 733 452 1860 5777466 204819 NJ 04 65 0 28835 2478 31313 8433144 7354 1 0 0 1 2 731 455 1863 4115364 309315 NJ 05 65 0 29346 2531 31877 8447923 7354 2 0 1 3 2 748 455 1867 7148940 316094 NJ 06 65 0 26561 2452 29013 8444334 7354 1 0 0 1 5 772 455 1873 6639001 341595 NJ 07 65 0 29733 2674 32407 8405701 7354 1 0 0 1 6 724 455 1873 3226525 356344 NJ 08 65 0 29640 2739 32379 8400257 7354 2 0 0 2 3 590 455 1870 3854305 355058 NJ 09 65 0 29437 2653 32090 8415289 7354 0 0 0 0 1 583 455 1870 5581552 398271 NM 95 75 60 3374 256 3630 1661000 121297 0 0 0 0 2 485 10879 611 512915 40389 NM 96 75 60 2883 230 3113 1352000 121297 0 0 0 0 1 485 10805 601 546466 42352 NM 97 75 60 3261 264 3525 1567000 121297 0 0 1 1 0 484 10808 594 686389 40617 NM 98 75 60 3648 299 3947 1732000 121297 0 0 0 0 1 424 10857 593 518231 62739 NM 99 75 60 3440 299 3739 1590000 121297 0 0 0 0 0 460 10844 591 974423 54836 NM 00 75 60 3619 306 3925 1720053 121297 0 0 0 0 1 432 10827 591 1108855 61133 NM 01 75 60 2823 275 3098 1315400 121297 2 0 0 2 2 464 10826 588 1062379 74939 NM 02 75 60 3779 363 4142 1851009 121297 1 0 0 1 0 449 10815 584 996974 16391 NM 03 75 60 3265 324 3589 1571837 121297 0 0 0 0 0 439 10593 814 703893 12671 NM 04 75 60 3573 371 3944 1758428 121297 0 0 0 0 2 521 11042 967 1943143 34309 NM 05 75 60 3672 364 4036 1809415 121297 0 0 0 0 0 488 11023 968 781168 56636 NM 06 75 60 3106 315 3421 1507980 121297 1 0 0 1 1 484 11024 969 711997 8553 NM 07 75 60 3710 373 4083 1850280 121297 2 2 0 4 1 413 11020 961 752581 8553 NM 08 75 60 3730 390 4120 1805058 121297 1 0 0 1 0 366 10989 959 873424 36435 NM 09 75 60 4048 427 4475 1958665 121297 0 0 1 1 0 361 10991 960 903841 23535 NY 95 65 21 55730 7511 63241 17033000 47126 2 0 1 3 11 1679 11351 4241 5135855 173565 NY 96 65 21 51034 7173 58207 15668000 47126 0 0 2 2 7 1593 11352 4247 4201474 174581 NY 97 65 21 55830 7543 63373 16908000 47126 1 0 1 2 6 1652 10998 4012 5344670 235814 NY 98 65 21 58161 7925 66086 18007000 47126 0 0 2 2 2 1514 11010 4013 5757485 266873 NY 99 65 21 56749 7604 64353 15924000 47126 2 0 0 2 6 1599 11010 4017 5224390 268596 NY 00 65 21 57268 8259 65527 18127223 47126 4 0 0 4 11 1460 11009 4017 5117702 261456 NY 01 65 21 55020 8226 63246 17362904 47126 0 0 0 0 7 1564 11009 4030 5564323 343243 NY 02 65 21 51892 7762 59654 16675972 47126 3 0 0 3 5 1530 11003 4030 6773486 288395 NY 03 65 21 46505 6349 52854 17576779 47126 1 0 1 2 9 1493 11002 4030 6527306 306640 NY 04 65 21 54783 8325 63108 18896255 47126 3 0 1 4 7 1493 11001 4031 6014501 303272 NY 05 65 21 52544 8351 60895 17905562 47126 0 0 1 1 1 1429 11001 4031 10390646 397190 NY 06 65 21 53597 8579 62176 18719867 47126 2 0 0 2 2 1456 10075 5473 5533117 360439 NY 07 65 21 53742 8579 62321 18883746 47126 1 0 0 1 5 1333 9919 5053 7912424 338298 NY 08 65 21 52722 8607 61329 18103583 47126 0 0 1 1 4 1231 9936 5034 7119528 336575 NY 09 65 21 53588 8572 62160 19120958 47126 2 0 0 2 7 1156 9936 5033 8072728 410993
198
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
NC 95 70 40 14407 1637 16044 6967000 48619 1 0 0 1 5 1448 68452 9267 1962649 168850 NC 96 70 40 15297 1732 17029 7308000 48619 1 0 0 1 5 1494 68593 9186 1924314 200471 NC 97 70 40 15759 1803 17562 7415000 48619 1 0 0 1 0 1483 68715 9167 2048822 176245 NC 98 70 40 16099 1884 17983 7542000 48619 2 0 0 2 2 1596 68770 9234 2509987 190774 NC 99 70 40 16602 1970 18572 7526000 48619 3 0 0 3 2 1505 68926 9176 2433617 165406 NC 00 70 40 17252 2078 19330 8041374 48619 2 0 0 2 2 1557 69069 9197 2619172 259306 NC 01 70 40 17436 2153 19589 8162627 48619 4 0 1 5 6 1530 69190 9188 2970413 236463 NC 02 70 40 17510 2181 19691 8313727 48619 1 1 1 3 4 1576 69291 9226 2883140 228881 NC 03 70 40 17969 2221 20190 8310332 48619 4 0 0 4 12 1553 69351 9291 2640069 283596 NC 04 70 40 18414 2355 20769 8533414 48619 1 0 0 1 2 1557 69537 9336 3581258 303936 NC 05 70 40 18656 2417 21073 8586228 48619 1 0 0 1 4 1534 62515 16515 3622847 278077 NC 06 70 40 18917 2502 21419 8851847 48619 0 0 0 0 3 1559 62609 16459 3388175 307827 NC 07 70 40 19098 2405 21503 9036361 48619 4 0 0 4 6 1675 62733 16555 3576305 343367 NC 08 70 40 19488 2563 22051 9189952 48619 1 0 1 2 2 1433 62890 16576 3852853 353417 NC 09 70 40 19851 2455 22306 9260266 48619 3 0 0 3 3 1314 62890 16576 3573805 359568 ND 95 70 0 958 63 1021 639000 69001 0 0 0 0 0 74 7174 204 266826 10816 ND 96 70 0 968 66 1034 640000 69001 0 0 0 0 0 85 7174 204 275575 10317 ND 97 70 0 982 67 1049 622000 69001 0 0 0 0 0 105 7174 205 316344 12063 ND 98 70 0 1011 72 1083 638000 69001 0 0 0 0 0 92 7174 205 311182 11744 ND 99 70 0 1009 78 1087 630000 69001 0 0 0 0 0 119 7174 205 413951 13216 ND 00 70 0 1012 80 1092 637023 69001 0 0 0 0 0 86 7174 205 395485 13166 ND 01 70 0 1017 78 1095 625681 69001 0 0 0 0 0 105 7174 205 357343 14497 ND 02 70 0 1021 83 1104 608703 69001 0 0 0 0 0 97 7175 205 386566 20680 ND 03 70 0 1067 82 1149 625680 69001 0 0 0 0 0 105 7178 205 386794 20698 ND 04 70 0 1097 85 1182 630131 69001 0 0 0 0 0 100 7167 215 383358 23700 ND 05 70 0 1079 95 1174 634904 69001 0 0 0 0 0 123 7167 214 414925 17710 ND 06 70 0 1080 98 1178 632973 69001 0 0 0 0 0 111 7168 217 543150 27044 ND 07 70 0 1090 110 1200 638897 69001 0 0 0 0 0 111 7168 217 467096 18811 ND 08 70 0 1110 114 1224 640830 69001 0 0 0 0 0 104 7168 217 481308 20395 ND 09 70 0 1128 132 1260 646844 69001 0 0 0 0 1 140 7167 217 488383 24002 OH 95 65 60 15112 1436 16548 8884000 40859 1 0 0 1 4 1360 15437 4083 2707031 162035 OH 96 65 60 18428 1998 20426 10595000 40859 0 0 0 0 3 1391 15435 4099 3164532 184704 OH 97 65 60 19079 2205 21284 10559000 40859 1 0 0 1 0 1441 15275 4022 3160312 172270 OH 98 65 60 20140 2327 22467 11152000 40859 1 0 1 2 0 1422 15277 4024 3503543 117799 OH 99 65 60 18625 2128 20753 10066000 40859 0 0 0 0 1 1430 15271 4024 3377774 193176 OH 00 65 60 20493 2400 22893 10992944 40859 2 0 0 2 5 1366 15269 4020 3125999 108509 OH 01 65 60 20216 2532 22748 10680532 40859 2 0 2 4 3 1378 15269 4023 3444961 209058 OH 02 65 60 21099 2565 23664 10878422 40859 0 0 0 0 3 1418 15275 4026 3330804 219930 OH 03 65 60 20339 2524 22863 10734993 40859 1 0 0 1 3 1274 14265 5027 3481120 238473 OH 04 65 60 17636 1953 19589 9782447 40859 1 0 0 1 2 1286 14281 5027 3773536 221799 OH 05 65 60 19889 2383 22272 10491311 40859 0 0 0 0 2 1323 14274 5017 3922994 251049 OH 06 65 60 19063 2226 21289 10476045 40859 3 0 0 3 6 1238 14254 5010 4476316 272084 OH 07 65 60 19591 2299 21890 10072335 40859 1 0 0 1 5 1257 14254 5011 4675735 281334 OH 08 65 60 19541 2309 21850 9715611 40859 1 0 0 1 3 1190 14253 5005 4543735 300320 OH 09 65 60 19818 2292 22110 10223161 40859 0 0 1 1 1 1021 14253 5005 4752972 280582 OK 95 75 24 6140 442 6582 3275000 68594 2 0 0 2 3 669 11556 980 811647 55534 OK 96 75 24 6114 464 6578 3299000 68594 0 0 0 0 1 772 11550 977 890689 53721 OK 97 75 24 6220 499 6719 3317000 68594 1 0 0 1 0 838 11305 961 923554 55772 OK 98 75 24 6340 497 6837 3347000 68594 1 0 1 2 0 755 11314 962 1649247 26134 OK 99 75 24 6325 516 6841 3357000 68594 1 0 0 1 2 741 11314 962 1178871 78347 OK 00 75 24 6435 522 6957 3450654 68594 2 0 0 2 5 650 11314 957 1317186 87844 OK 01 75 24 6555 525 7080 3460097 68594 0 1 0 1 2 682 11311 957 1153533 87332 OK 02 75 24 6590 518 7108 3493714 68594 1 0 0 1 4 739 11312 954 1780379 77293 OK 03 75 24 6534 525 7059 3511532 68594 0 0 0 0 0 671 11078 1186 1267509 79922 OK 04 75 24 6478 519 6997 3523553 68594 0 0 0 0 1 774 11082 1197 1249733 139599 OK 05 75 24 6454 538 6992 3547884 68594 3 0 0 3 4 802 11090 1196 1174878 90635 OK 06 75 24 6513 545 7058 3579212 68594 1 0 0 1 4 765 11090 1196 2070167 132934 OK 07 75 24 6735 548 7283 3617316 68594 0 0 0 0 5 754 11090 1193 1496340 26010 OK 08 75 24 7002 693 7695 3642361 68594 2 1 0 3 6 749 11089 1191 2283657 59031 OK 09 75 24 7250 698 7948 3681857 68594 1 0 0 1 3 738 11089 1191 1848927 107307 OR 95 65 24 4586 418 5004 2567000 95987 0 0 0 0 3 574 6882 733 898252 47881 OR 96 65 24 4797 441 5238 3186000 95987 0 0 0 0 1 526 6851 740 995430 44767 OR 97 65 24 4749 464 5213 3147000 95987 2 0 0 2 3 524 6880 754 963192 49690 OR 98 65 24 4818 464 5282 3208000 95987 0 0 0 0 0 538 6889 764 1012150 48970 OR 99 65 24 5025 525 5550 3292000 95987 1 0 0 1 1 414 6855 757 1007122 49137 OR 00 65 24 5051 533 5584 3394066 95987 0 0 0 0 0 451 6865 732 1023632 58528 OR 01 65 24 5048 536 5584 3446968 95987 0 0 2 2 0 488 6859 732 1087767 55408 OR 02 65 24 5066 551 5617 3492816 95987 1 0 0 1 1 436 6855 720 1289524 57180 OR 03 65 24 4946 528 5474 3529899 95987 0 0 0 0 0 512 6813 732 1183794 52722 OR 04 65 24 4423 497 4920 3107327 95987 0 0 0 0 0 456 6690 863 1198718 48362 OR 05 65 24 4726 536 5262 3321523 95987 0 0 0 0 0 488 6680 854 1708105 49014 OR 06 65 24 4173 383 4556 2928366 95987 0 0 0 0 0 477 6679 852 1714715 57187 OR 07 65 24 5349 623 5972 3712636 95987 1 0 0 1 2 455 6683 852 2377932 73482 OR 08 65 24 5276 614 5890 3752587 95987 0 0 0 0 0 416 6689 851 1436210 85361 OR 09 65 24 5447 588 6035 3790072 95987 0 0 0 0 0 377 6693 853 1571950 94779 PA 95 65 40 19235 2004 21239 8383000 44743 1 0 0 1 4 1480 32739 8030 3446430 290341 PA 96 65 40 18546 2121 20667 7830000 44743 1 0 0 1 0 1469 32772 8030 3019428 291558 PA 97 65 40 19308 2275 21583 8307000 44743 1 1 0 2 8 1557 32388 7856 3415629 310020 PA 98 65 40 20506 2532 23038 8803000 44743 1 0 0 1 5 1481 32302 7860 3982893 336207 PA 99 65 40 20191 2560 22751 8506000 44743 0 0 1 1 0 1549 32243 7858 4776456 338818 PA 00 65 40 20124 2639 22763 8524494 44743 1 0 1 2 3 1520 32207 7843 4026523 378681 PA 01 65 40 18705 2564 21269 7309014 44743 0 0 1 1 1 1532 32102 7833 4684296 344587 PA 02 65 40 21015 2698 23713 8590601 44743 1 0 0 1 2 1614 32071 7835 5379569 362871 PA 03 65 40 20308 2652 22960 8054446 44743 1 0 0 1 4 1577 29558 10334 4971466 461140 PA 04 65 40 20109 2647 22756 8097970 44743 2 1 0 3 7 1490 28880 11011 5065285 423189 PA 05 65 40 20451 2613 23064 8268941 44743 0 0 0 0 3 1616 28865 11024 5056466 440981 PA 06 65 40 20785 2629 23414 8387382 44743 0 0 1 1 4 1525 28827 11015 5779626 483814 PA 07 65 40 22245 2696 24941 9052251 44743 0 0 0 0 6 1491 28853 11020 6631191 571161 PA 08 65 40 22290 2698 24988 9356098 44743 2 1 2 5 4 1468 28847 11013 6188175 574906 PA 09 65 40 22382 2628 25010 9523147 44743 0 0 1 1 0 1256 28847 11013 6724227 574136
199
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
RI 95 65 40 2217 125 2342 979000 1034 0 0 0 0 0 69 322 810 292454 10373 RI 96 65 40 2217 128 2345 979000 1034 0 0 0 0 0 69 322 810 301570 10586 RI 97 65 40 2242 135 2377 976000 1034 0 0 0 0 0 75 325 883 232202 9253 RI 98 65 40 2255 136 2391 982000 1034 0 0 0 0 0 74 325 893 339506 11241 RI 99 65 40 2252 139 2391 985000 1034 0 0 0 0 0 88 328 902 320431 11948 RI 00 65 40 2304 146 2450 1041849 1034 0 0 0 0 0 80 267 902 267353 13281 RI 01 65 40 2286 153 2439 1052814 1034 0 0 0 0 0 81 375 740 382411 15616 RI 02 65 40 2324 161 2485 1063557 1034 0 0 0 0 0 84 375 740 387384 17669 RI 03 65 40 2337 169 2506 1069852 1034 0 0 0 0 1 104 316 785 302716 16993 RI 04 65 40 2308 165 2473 1074295 1034 0 0 0 0 0 83 316 787 387977 17583 RI 05 65 40 2372 179 2551 1069926 1034 0 0 0 0 0 87 316 785 411965 19071 RI 06 65 40 2380 181 2561 1067610 1034 0 0 0 0 0 81 316 787 625915 19046 RI 07 65 40 2392 186 2578 1057832 1034 0 0 0 0 0 69 316 788 484768 33601 RI 08 65 40 2403 180 2583 1050788 1034 0 0 0 0 0 65 315 793 439757 23398 RI 09 65 40 2370 184 2554 1053209 1034 0 0 0 0 0 83 316 793 379193 27347 SC 95 70
7516 686 8202 3671000 30058 1 0 0 1 2 881 34593 6909 682535 54235
SC 96 70
7468 711 8179 3696000 30058 2 0 0 2 2 930 34584 6916 714229 61842 SC 97 70
7852 795 8647 3756000 30058 0 0 0 0 0 903 34609 6916 760079 64872
SC 98 70
8207 922 9129 3834000 30058 2 0 0 2 0 1002 34613 6921 776301 70989 SC 99 70
8091 938 9029 3774000 30058 2 0 0 2 0 1065 34595 6922 1007385 82295
SC 00 70
8527 949 9476 3498701 30058 2 0 0 2 5 1065 34614 6916 872060 95570 SC 01 70
8333 1014 9347 3965611 30058 0 0 0 0 3 1060 34527 6949 1275610 110834
SC 02 70
7791 996 8787 3735856 30058 2 0 3 5 3 1053 34545 6953 1014652 131528 SC 03 70
7876 998 8874 3634757 30058 2 0 0 2 2 969 34524 6951 982824 114275
SC 04 70
9411 1156 10567 4188882 30058 0 0 0 0 1 1046 34588 6944 1184536 111695 SC 05 70
9832 1208 11040 4185868 30058 2 0 1 3 2 1093 30191 11202 1588966 100164
SC 06 70
9343 1254 10597 4275140 30058 0 0 1 1 0 1037 30241 11188 1396807 89946 SC 07 70
9739 1295 11034 4406102 30058 1 0 0 1 1 1066 30256 11181 1186743 94216
SC 08 70
10061 1288 11349 4479234 30058 1 0 0 1 0 920 30264 11164 1430676 99531 SC 09 70
8449 1112 9561 3779301 30058 1 1 0 2 2 894 30268 11154 1178824 94268
SD 95 75 42.5 1041 46 1087 728000 75813 0 0 0 0 0 158 7625 188 293532 13386 SD 96 75 42.5 1047 48 1095 716000 75813 0 0 0 0 0 175 7624 190 289135 12780 SD 97 75 42.5 1074 50 1124 721000 75813 0 0 0 0 1 148 7623 191 349462 20796 SD 98 75 42.5 1104 53 1157 733000 75813 0 0 0 0 0 165 7612 187 356146 11094 SD 99 75 42.5 1143 56 1199 730000 75813 0 0 0 0 1 150 7605 188 371942 16551 SD 00 75 42.5 1133 62 1195 752792 75813 1 0 0 1 0 173 7605 188 411768 16419 SD 01 75 42.5 1181 67 1248 756600 75813 0 0 0 0 0 171 7651 189 432598 26977 SD 02 75 42.5 1198 69 1267 747844 75813 0 0 0 0 0 180 7653 188 496349 18766 SD 03 75 42.5 1218 76 1294 761048 75813 0 0 1 1 0 203 7652 189 454799 31584 SD 04 75 42.5 1278 84 1362 761402 75813 0 0 0 0 0 197 7606 244 485106 20045 SD 05 75 42.5 1310 85 1395 774216 75813 0 0 0 0 1 186 7634 237 449252 18867 SD 06 75 42.5 1146 85 1231 775859 75813 0 0 0 0 0 191 7633 210 464009 24972 SD 07 75 42.5 1323 97 1420 792578 75813 0 0 0 0 0 146 7613 229 452587 26313 SD 08 75 42.5 1347 85 1432 800171 75813 0 0 0 0 0 119 7608 228 477793 30023 SD 09 75 42.5 1396 88 1484 809838 75813 0 0 0 0 0 131 7608 228 446059 35078 TN 95 70 40 8819 810 9629 4731000 41234 1 0 0 1 6 1259 11329 2390 1244488 65814 TN 96 70 40 9844 968 10812 4960000 41234 2 0 0 2 4 1239 11348 2391 1297313 68673 TN 97 70 40 10570 998 11568 4857000 41234 2 0 0 2 3 1225 11329 2423 1371907 69688 TN 98 70 40 11306 1142 12448 5166000 41234 0 0 0 0 6 1216 11328 2432 1492952 75995 TN 99 70 40 11382 1201 12583 4926000 41234 4 0 0 4 3 1302 11359 2452 1475245 70411 TN 00 70 40 12400 1300 13700 5238529 41234 2 2 1 5 7 1307 11350 2441 1439811 36898 TN 01 70 40 13334 1452 14786 5607075 41234 3 0 2 5 4 1251 11356 2436 1534021 81251 TN 02 70 40 13629 1545 15174 5787364 41234 2 0 0 2 6 1177 11339 2457 1589318 93458 TN 03 70 40 13845 1564 15409 5840183 41234 1 0 1 2 8 1193 10945 2848 1634944 90417 TN 04 70 40 13946 1639 15585 5898401 41234 3 0 1 4 7 1288 10859 2949 1641829 13571 TN 05 70 40 13680 1481 15161 5958488 41234 0 1 1 2 4 1270 10849 2969 1824220 108551 TN 06 70 40 13587 1397 14984 6034250 41234 2 0 0 2 9 1287 10834 3002 1787812 36880 TN 07 70 40 13898 1448 15346 6150558 41234 0 0 0 0 2 1210 10868 3020 1742081 36880 TN 08 70 40 14016 1513 15529 6209570 41234 1 0 0 1 4 1035 10861 3021 1866303 28900 TN 09 70 40 14385 1505 15890 6293243 41234 1 0 0 1 2 989 10855 3016 1860157 36560 TX 95 75 0 38285 4078 42363 18719000 261230 4 0 2 6 9 3183 67988 10395 3765214 245333 TX 96 75 0 39074 4146 43220 19123000 261230 3 0 0 3 10 3742 68093 10434 4599517 343368 TX 97 75 0 39221 4244 43465 19439000 261230 1 2 1 4 11 3513 68298 10653 4396423 309656 TX 98 75 0 41118 4551 45669 19757000 261230 2 0 1 3 12 3586 68468 10618 4410860 367809 TX 99 75 0 39787 4333 44120 19503000 261230 2 0 0 2 5 3522 68595 10569 4516860 388528 TX 00 75 0 41434 4616 46050 20622676 261230 2 1 0 3 8 3779 68587 10673 5652251 401870 TX 01 75 0 42246 4778 47024 21037517 261230 5 3 2 10 11 3736 68687 10661 5410308 406452 TX 02 75 0 42806 4904 47710 21670261 261230 6 1 1 8 4 3823 68451 11043 6202358 438215 TX 03 75 0 42682 4916 47598 21886408 261230 4 1 1 6 4 3821 68371 11120 7007585 462826 TX 04 75 0 44001 5118 49119 22478824 261230 8 0 2 10 7 3583 68441 11182 6697083 543764 TX 05 75 0 44446 5236 49682 22763951 261230 3 0 1 4 5 3504 68439 11209 9111098 583895 TX 06 75 0 44192 5278 49470 23183189 261230 4 0 1 5 5 3475 68417 11432 8692574 654809 TX 07 75 0 45425 5552 50977 23413736 261230 5 2 2 9 15 3363 68357 11618 13480107 671705 TX 08 75 0 46909 5766 52675 24229899 261230 5 2 2 9 11 3382 66419 13648 19661220 767562 TX 09 75 0 48579 6278 54857 24590665 261230 2 0 0 2 2 3071 66419 13648 9671400 777356 UT 95 75 40 3565 314 3879 1937000 82198 1 0 0 1 0 325 5031 761 486161 25304 UT 96 75 40 3296 231 3527 1982000 82198 0 0 0 0 0 321 5034 762 547030 25860 UT 97 75 40 3513 268 3781 1998000 82198 0 0 0 0 0 366 5035 767 1239865 30523 UT 98 75 40 3776 311 4087 2081000 82198 1 0 0 1 1 350 5071 772 943111 32012 UT 99 75 40 4381 487 4868 2120000 82198 0 0 0 0 0 360 5067 771 869845 35201 UT 00 75 40 4255 386 4641 2224607 82198 0 0 1 1 0 373 5066 768 922769 45521 UT 01 75 40 4225 389 4614 2269233 82198 1 0 0 1 0 291 5058 765 885814 31681 UT 02 75 40 4260 376 4636 2315689 82198 1 0 0 1 0 328 5088 765 1120382 41115 UT 03 75 40 4249 380 4629 2321175 82198 1 0 0 1 0 309 5083 770 959617 36845 UT 04 75 40 4193 332 4525 2373842 82198 0 0 0 0 0 296 4801 1056 1775199 38076 UT 05 75 40 4268 373 4641 2468942 82198 0 0 0 0 0 282 4797 1060 944870 38396 UT 06 75 40 4172 323 4495 2549402 82198 0 0 0 0 0 287 4772 1076 1091088 39910 UT 07 75 40 4315 315 4630 2644187 82198 0 0 0 0 2 299 4778 1052 1462086 43504 UT 08 75 40 4362 344 4706 2735255 82198 1 0 0 1 2 275 4781 1059 1474838 45601 UT 09 75 40 4430 355 4785 2783798 82198 0 0 0 0 0 244 4782 1058 1814805 55702
200
ST YR S DT MO FO LEOS STPOP AREA A M P LD LI TTF RM UM HREC HSES
VT 95 65
844 62 906 348000 9217 0 0 0 0 0 106 2451 176 200559 23614 VT 96 65
793 51 844 353000 9217 0 0 0 0 0 88 2452 176 183244 24197
VT 97 65
760 54 814 320000 9217 0 0 0 0 0 96 2455 176 230903 25352 VT 98 65
830 54 884 320000 9217 0 0 0 0 0 104 2454 177 208447 25557
VT 99 65
838 64 902 345000 9217 0 0 0 0 0 90 2454 177 254560 26928 VT 00 65
828 65 893 357836 9217 0 0 0 0 0 76 2454 177 272088 47877
VT 01 65
864 65 929 362470 9217 0 0 0 0 0 92 2452 177 305845 19800 VT 02 65
888 68 956 364545 9217 0 0 0 0 0 78 2454 177 329811 42168
VT 03 65
908 72 980 364662 9217 1 0 1 2 1 69 2454 177 315288 46130 VT 04 65
988 77 1065 371894 9217 0 0 0 0 0 98 2454 180 311353 39776
VT 05 65
1066 92 1158 371547 9217 0 0 0 0 2 73 2455 179 303804 45061 VT 06 65
1072 91 1163 372588 9217 0 0 0 0 0 87 2455 179 329145 45785
VT 07 65
1019 88 1107 342945 9217 0 0 0 0 1 66 2455 179 393657 46318 VT 08 65
874 81 955 271869 9217 0 0 0 0 0 73 2452 179 390582 77620
VT 09 65
976 96 1072 303744 9217 0 0 0 0 0 74 2451 178 366931 57939 VA 95 65
12888 1356 14244 6605000 39493 0 0 1 1 3 900 48624 8438 2178313 105285
VA 96 65
13164 1419 14583 6672000 39493 1 0 0 1 0 877 48625 8502 2228077 112219 VA 97 65
13350 1458 14808 6734000 39493 0 0 0 0 0 984 48662 8570 2382184 111278
VA 98 65
13761 1502 15263 6788000 39493 1 1 1 3 0 935 48702 8647 2440481 105430 VA 99 65
14118 1561 15679 6873000 39493 2 0 0 2 0 878 48987 8751 2833302 116001
VA 00 65
14385 1651 16036 7078515 39493 0 0 0 0 0 929 49197 8656 2751585 112447 VA 01 65
14668 1753 16421 7186387 39493 0 1 0 1 1 935 49274 7668 3058650 118384
VA 02 65
14787 1765 16552 7292028 39493 1 1 0 2 2 914 49358 7723 3231710 128576 VA 03 65
14893 1868 16761 7381636 39493 0 0 0 0 1 943 47933 9390 3641578 143476
VA 04 65
15120 1891 17011 7456600 39493 0 0 0 0 1 925 47996 9519 3035586 142939 VA 05 65
15406 1927 17333 7566489 39493 2 0 0 2 0 947 48150 9710 3470356 146386
VA 06 65
15656 2016 17672 7641879 39493 1 1 1 3 8 963 47960 9520 3511979 169197 VA 07 65
15987 2094 18081 7710724 39493 2 0 0 2 1 1027 47788 9940 3507970 178642
VA 08 65
16265 2193 18458 7767656 39493 0 0 0 0 3 824 47628 10291 3784462 210243 VA 09 65
16168 2155 18323 7880881 39493 3 0 0 3 3 757 47723 10380 3099443 191353
WA 95 70
7953 714 8667 5276000 66456 0 0 0 0 0 653 5954 1081 1873164 125446 WA 96 70
7941 733 8674 5292000 66456 0 0 0 0 1 712 5932 1104 1817475 121626
WA 97 70
8183 790 8973 5474000 66456 0 0 0 0 0 674 5933 1110 1935871 125531 WA 98 70
8308 835 9143 5545000 66456 0 0 0 0 0 662 5934 1114 1765821 133597
WA 99 70
8523 913 9436 5597000 66456 0 1 0 1 1 637 5933 1113 1859009 126667 WA 00 70
8633 947 9580 5713225 66456 0 0 0 0 1 631 5933 1113 1698258 132082
WA 01 70
8846 947 9793 5968232 66456 1 0 0 1 2 649 5934 1114 2131913 157609 WA 02 70
8910 958 9868 6064698 66456 0 0 0 0 5 658 5934 1118 2155435 150122
WA 03 70
8954 947 9901 6125732 66456 1 0 0 1 1 600 5978 1070 2251390 162161 WA 04 70
8870 955 9825 6197043 66456 0 1 0 1 2 563 5727 1318 2612263 136233
WA 05 70
9152 978 10130 6278257 66456 0 0 0 0 2 647 5734 1311 2578776 140925 WA 06 70
9259 1001 10260 6385798 66456 1 0 0 1 2 630 5731 1310 2836313 163673
WA 07 70
9394 973 10367 6459529 66456 1 0 0 1 1 568 5731 1311 3305180 169639 WA 08 70
9605 1000 10605 6538812 66456 0 0 0 0 0 521 5732 1310 3988204 174235
WA 09 70
9612 981 10593 6642851 66456 1 0 0 1 0 492 5737 1325 3486787 190781 WV 95 70
2764 106 2870 1518000 24038 0 0 0 0 0 376 30661 1353 786341 17778
WV 96 70
2824 113 2937 1518000 24038 0 0 0 0 0 348 30780 1369 861062 19385 WV 97 70
2868 109 2977 1522000 24038 0 0 0 0 0 381 30850 1357 945698 23163
WV 98 70
2923 107 3030 1520000 24038 0 0 0 0 2 354 31276 1391 890675 17805 WV 99 70
3077 112 3189 1781000 24038 1 0 0 1 1 395 31754 1424 1089541 21430
WV 00 70
3027 101 3128 1802649 24038 0 0 0 0 1 411 32479 1452 1108290 70289 WV 01 70
2992 100 3092 1792619 24038 1 0 0 1 1 376 32520 1455 1183586 74222
WV 02 70
2936 92 3028 1790599 24038 0 0 0 0 0 439 32556 1458 1268403 78300 WV 03 70
3071 103 3174 1803483 24038 0 0 0 0 1 394 32444 1455 1098943 41565
WV 04 70
3080 97 3177 1805840 24038 0 0 0 0 0 411 32514 1456 1039179 31977 WV 05 70
3122 98 3220 1800849 24038 0 0 0 0 1 374 31517 2471 1457507 34154
WV 06 70
3225 108 3333 1802755 24038 0 0 0 0 0 410 31118 2971 1085171 34163 WV 07 70
3189 110 3299 1799090 24038 1 0 0 1 1 431 31154 3065 1971302 34481
WV 08 70
3276 112 3388 1814873 24038 0 0 0 0 1 380 31295 3073 1167332 38142 WV 09 70
3362 114 3476 1810824 24038 0 0 0 0 0 356 31448 3062 1371144 38549
WI 95 65 40 10371 1280 11651 5038000 54158 0 0 0 0 0 745 10365 1451 1256047 58058 WI 96 65 40 10380 1305 11685 5012000 54158 0 0 0 0 3 761 10383 1445 1397335 61351 WI 97 65 40 10465 1343 11808 5052000 54158 0 0 1 1 1 725 10385 1452 1378325 61480 WI 98 65 40 10780 1432 12212 5224000 54158 3 0 0 3 3 714 10318 1440 1315570 57033 WI 99 65 40 10746 1528 12274 5165000 54158 0 0 0 0 0 745 10300 1453 1704561 57472 WI 00 65 40 10723 1600 12323 5180621 54158 0 0 0 0 3 799 10310 1442 1612018 61123 WI 01 65 40 10377 1594 11971 5153905 54158 0 0 0 0 0 763 10311 1442 1781753 73119 WI 02 65 40 9781 1566 11347 4849982 54158 0 0 0 0 0 803 10318 1437 2208996 68856 WI 03 65 40 10735 1766 12501 5268938 54158 1 0 0 1 1 848 10001 1773 2024030 72823 WI 04 65 40 11001 1838 12839 5504848 54158 0 1 0 1 0 792 9692 2120 1990433 75695 WI 05 65 40 11223 1847 13070 5535323 54158 0 0 0 0 0 815 9754 2027 2304773 73032 WI 06 65 40 11192 1846 13038 5553640 54158 1 0 0 1 2 724 9743 2029 2128850 80087 WI 07 65 40 11175 1795 12970 5598633 54158 0 0 0 0 1 756 9738 2031 2975175 88528 WI 08 65 40 11230 1808 13038 5607356 54158 0 0 1 1 2 605 9731 2041 2187975 88414 WI 09 65 40 11326 1794 13120 5648330 54158 1 0 0 1 0 561 9731 2041 2648591 106177 WY 95 75 17.5 1084 59 1143 479000 97092 0 0 0 0 0 170 6388 393 283705 14486 WY 96 75 17.5 1154 73 1227 474000 97092 0 0 0 0 0 143 6387 393 285250 15579 WY 97 75 17.5 1093 57 1150 478000 97092 0 0 0 0 1 137 6387 397 294398 16583 WY 98 75 17.5 1097 71 1168 480000 97092 1 0 0 1 1 154 6389 400 325889 13101 WY 99 75 17.5 1119 83 1202 477000 97092 0 0 0 0 1 189 6346 405 393043 12977 WY 00 75 17.5 1112 80 1192 493272 97092 0 0 0 0 0 152 6353 405 385358 18896 WY 01 75 17.5 1158 81 1239 492634 97092 0 0 0 0 0 186 6353 406 359600 21635 WY 02 75 17.5 1153 86 1239 496899 97092 0 0 0 0 0 176 6352 405 419855 23425 WY 03 75 17.5 1212 104 1316 497319 97092 0 0 0 0 0 165 6350 409 458817 24696 WY 04 75 17.5 1176 103 1279 504265 97092 0 0 0 0 0 164 6347 407 450939 27784 WY 05 75 17.5 1194 111 1305 505737 97092 0 0 0 0 0 170 6347 411 438251 29217 WY 06 75 17.5 1203 110 1313 511433 97092 0 1 0 1 0 195 6336 416 423823 30504 WY 07 75 17.5 1280 124 1404 519988 97092 0 0 0 0 0 150 6336 416 596256 32062 WY 08 75 17.5 1285 107 1392 528864 97092 0 0 0 0 0 159 6332 411 584929 38440 WY 09 75 17.5 1331 115 1446 540376 97092 0 0 0 0 0 134 6322 412 611469 43084