Skip navigation

Houston City Controller Taser Report 2008

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
CITY OF HOUSTON
Annise D. Parker
City Controller
Steve Schoonover
City Auditor

CONDUCTED ENERGY
DEVICE PROGRAM
PERFORMANCE AUDIT

Mir&ox
Report No. 2009-09

Rodriguez, P.c.
Certified Public Accountants

CED PROGRAM – PERFORMANCE AUDIT
TABLE OF CONTENTS
Page
EXECUTIVE SUMMARY
Background
Objectives and Scope
Overall Conclusion
Assessment
Recommendations
Acknowledgements
Audit Team Contact Information

1
1
2
3
15
19
20

PART I
Mir•Fox & Rodriguez, P.C. Detailed Background and Audit Methodology
Detailed Background
Audit Methodology

21
27

PART II
Mir•Fox & Rodriguez, P.C. Observations, Recommendations, and Corrective Actions
HPD CED Focus Groups
CED Inventory Count
Digital Power Magazine
Training Cartridges
Unrecorded Inventory
Scanner
Spark Test

29
50
52
52
53
53
54

PART III
University of Houston Center for Public Policy
A Statistical Analysis of the Use of Conducted Energy Devices
by the Houston Police Department
Summary
Background
Study Components
Data
Methods of Analysis
Results
Data Management
Conclusion
References

55
59
63
67
77
79
107
109
110

CED PROGRAM – PERFORMANCE AUDIT
TABLE OF CONTENTS

Page
PART IV
Sam Houston State University College of Criminal Justice
A Qualitative Analysis of the Use of Conducted Energy Devices by the Houston Police
Department
Summary
Background
Study Components and Method of Analysis
Results of Compliance Review of Houston Police Department Incident Reports
Houston Police Department Incident Report Review
Houston Police Department Training Academy - CED Training
Comments on UH CCP Team Statistical Analysis of the
Use of CEDS
Recommendations

111
114
118
120
132
139
143
147

PART V
Prototype, Fusion & Modeling, LLC
Graphical Deployment Analysis
Introduction
Background
Recommendation

150
150
156

CEDView Operating Manual
Main Window
Menus
Panel Menu
GIS Panel
Query Panel
Timeline Panel
Info Panel
Example

158
161
163
164
165
169
171
172

PART VI
Views of Responsible Officials

Mi'&'bX
P.c.

Rodriguez,

174

CED PROGRAM – PERFORMANCE AUDIT
EXECUTIVE SUMMARY
BACKGROUND
On November 3, 2004, the City of Houston (City) contracted to purchase 4,227 Conducted
Energy Devices (CEDs) with the related consumables, including CED cartridges and other
supplies for $4.68 million. The City’s contract for the Taser X-26 model CEDs has a five-year
“no questions asked” replacement warranty.
A CED is the technically correct name for the device commonly referred to by the brand name
Taser. In the probe mode this device propels probes that are attached to wires with a 21 foot
range that conducts sufficient energy to cause neuromuscular incapacitation (i.e. uncontrolled
muscle contractions that override an individual’s voluntary motor response). In the drive stun
mode a CED with no cartridge or a spent cartridge is placed in direct contact with the body and
is discharged. The drive stun mode is generally the secondary option used by the officer. A
cycle is defined as each time the officer pulls the trigger on the CED. The duration of a CED
electrical discharge is a minimum of five seconds per cycle. CEDs have a computerized
memory that tracks deployment data which can be downloaded for later review.
In conjunction with the purchase of the CEDs, HPD developed a CED Program that included
CED policies, procedures, and related training for the patrol officers. HPD’s policies state that
CEDs are classified as intermediate weapons 1 and are not a substitute for lethal force.
HPD’s CED Program has the following objectives:
•
•
•
•

Assist officers in securing and controlling combative individuals,
Reduce injuries to officers and suspects,
Reduce financial impact of civil liability in use-of-force calls for service/incidents, and
In limited situations, provide an alternative to deadly force.

OBJECTIVES AND SCOPE
Objectives of our HPD CED Program Performance Audit were as follows:
•
•
•
•
•
•

1

Determine compliance with procurement laws, ordinances, the City’s Policies and
Procedures, HPD Standard Operating Procedures, and General Orders (GOs).
Determine to what extent the HPD CED Program objectives were being met.
Determine if HPD was effectively managing the CED Program.
Determine compliance with CED Program Policies and Procedures.
Determine the HPD CED practices and compare them to the Police Executive
Research Forum’s (PERF) National Guidelines for CEDs.
Perform a statistical analysis to analyze the frequency and variables of HPD CED
calls for service/incidents for any notable patterns, aberrations, and/or adverse
trends which may warrant further study.

Intermediate weapons are defined by HPD as:
•
•
•
•

Baton
Oleoresin Capsicum (OC) spray (commonly known as pepper spray)
Soft impact weapon (i.e. Beanbag Shotgun)
CED (i.e. stun gun or Taser)

Mi'&'bX
Rodriguez, P.c.

1

The scope of the engagement was the period beginning January 1, 2000 through June 30, 2007
(the Scope period) with the CED Program being initiated in December 2004. The audit did not
examine the issue of whether exposure to a CED deployment (or multiple deployments) has
medical implications.

OVERALL CONCLUSION
The CED has been a very effective intermediate weapon and has been widely accepted by
most of the HPD patrol officers who were issued a CED. Additionally, HPD met its CED
program objectives related to assisting officers in securing and controlling combative individuals.
The chances of being subjected to a CED deployment are negligible. The CED is not deployed
in over 99% of the calls for service/incidents. On 53 of 1,284 occasions the officers used the
CED as an alternative to deadly force even though they were not required to do so by HPD
policies and State law.
HPD has been effectively managing the CED Program and was generally in compliance with
HPD and the City’s Policies and Procedures, as well as the related procurement laws. HPD’s
CED practices were used to develop the PERF National Guidelines. However, HPD plans to
revise its CED training to reflect certain CED situations encountered in the field. Also, HPD
needs to provide additional training to officers who were reluctant to deploy their CEDs because
of the related reporting requirements.
Over the course of conducting the statistical analysis, the statistical team found that the
available data contained some patterns and/or aberrations related to gender, race, and
geography. The lack of adequate data on suspect and officer characteristics (e.g., officer and
suspect physical size, suspect criminal history) resulted in crucial potential explanatory factors
being absent from the statistical analysis. Consequently, it is important to exercise considerable
caution when making any inferences from the CED statistical analysis. With this caveat in mind,
the results from the CED analysis suggest that certain combinations of officer and suspect
characteristics resulted in an increased probability of CED deployment. Depending on the race
of the officer and of the suspect, it was possible to see significant increases and/or decreases in
the rate of CED deployment. 2 However, some of these differences disappeared or changed
when looking at certain results at the City Council District level because there are more data
items at the City Council District Level. Rather than making inaccurate inferences by using
smaller subsets such as HPD Divisions, City Council Districts have the advantage of
representing distinct regions of the City that are drawn with the approval of the U.S. Department
of Justice (due to aspects of the Voting Rights Act); being mutually exclusive (for example,
events can occur in one, and only one Council District); and being independent of any HPD or
researcher decisions (Council Districts cannot be altered to affect the outcome of the analysis).
In short, the use of City Council Districts as a control allowed the results to be analyzed more
objectively.
Furthermore, HPD Management indicated that the City Legal Department would not provide the
necessary data to assess HPD’s objective related to reducing the financial impact of civil
liabilities in use-of-force calls for service/incidents.

2

See PART III - TABLE 7a THROUGH 7c CED DEPLOYMENTS BY CITY COUNCIL DISTRICT for details.

Mi'&'bX
Rodriguez, P.c.

2

The ambiguity in the statistical findings extended to the injuries analysis as well. For example,
during the Scope period, there was inadequate data available to make a determination related
to suspects’ injuries to compare with the CED cases. On the other hand, for officer injuries,
where data existed, we found that the reduction in injuries to officers began prior to the
introduction of the CED and related CED policies. This downward trend of the reduction in
injuries to officers continued throughout the Scope period. With the passage of time (and
more data) it will be possible to make a determination as to whether there are further reductions
in officer injuries consistent with the timing of the CED policy.
Despite cooperation from HPD in providing the data, antiquated data systems, and a general
lack of administrative staffing complicated the completion of the overall analysis. It should be
noted that HPD is in the process of addressing these data management and processing issues
by transitioning to a new database system, which should make data retrieval much easier.
However for this particular study, data acquisition and processing problems associated with
some data made it impossible to adequately control for important contextual variables in the
statistical analysis.

ASSESSMENT
The assessment was performed by the Audit Team that consisted of experts from Mir•Fox &
Rodriguez, P.C. (MFR); the University of Houston, Center for Public Policy (UH CPP); Rice
University; University of San Francisco; Sam Houston State University, College of Criminal
Justice (SHSU CCJ); and Prototype, Fusion & Modeling, LLC (PFM). MFR led the Audit Team
and provided the performance audit, accounting, information system (IS), and internal control
expertise. UH CPP provided statistical expertise and coordinated with mathematical experts
from Rice University and the University of San Francisco. UH CCP also coordinated with
Geospatial and IS experts from PFM. SHSU CCJ provided criminology and mathematical
expertise.
The Audit Team performed various tasks including:
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Analyzed HPD CED GOs and related Policies and Procedures,
Reviewed data management processes,
Analyzed CED deployments,
Reviewed Geo-Coordinates,
Reviewed CED related inventory,
Determined the effectiveness of a CED in controlling a suspect,
Analyzed the number of CED cycles used when deployed,
Analyzed the impact of alcohol and/or drugs consumed by suspects who were
subsequently subjected to a CED deployment,
Reviewed injuries to officers,
Reviewed HPD Training Academy (Academy) CED related training,
Analyzed suspect complaints related to CED deployment,
Determined who initiated the call for service as well as the nature of the call related to
CED deployment,
Analyzed the use of a CED instead of deadly force,
Performed statistical analysis of call for service/incident reports for the Scope period,
and
Conducted HPD Officer Focus Group Sessions.

Mi'&'bX
Rodriguez, P.c.

3

Based upon the results of performing these tasks, the Audit Team’s assessment is summarized
as follows:
HPD CED GOs and Policies and Procedures
HPD’s CED policies were effective in accomplishing the CED program objectives. The majority
of the calls for service/incident reports reviewed 3 indicated that the officer used the CED in
compliance with HPD policies.
The Audit Team was able to review all 1,284 of the hard copy CED incident reports that were
recorded. During the audit we noted that an Executive Assistant Police Chief was assigned to
oversee the CED Program and he was reviewing each CED incident report. However,
documented evidence of his review was not present in the early stages of the CED program.
According to HPD Management, other incident reports are not subjected to the same review
and signature process.
Data Management Processes
The process for collecting the data raised some important issues that affected the statistical
analysis for this audit. The Audit Team found HPD personnel to be fully cooperative in all
requests for information. However, there were complications in acquiring and assembling the
data as the data collection process was underway.
For example, HPD fielded approximately 2.8 million calls for service that resulted in
approximately 1.4 million service/incident reports which were recorded in multiple databases
during the Scope period. Approximately 48% of the 1.4 million electronic police calls for
service/incident reports did not contain suspect information (e.g., the incident was reported after
the suspects had long left the scene of the incident, no suspect was involved in the incident,
and/or no information on the suspect was collected). The addition of key explanatory variables
(suspect race/ethnicity, Uniform Crime Report (UCR) code, zip code of incident location, City
Council District of incident location) resulted in the exclusion of approximately 110,000 cases
while the lack of officer data for an incident led to the exclusion of approximately 50,000
additional cases. This left the Audit Team with a final analysis population of approximately
570,000 merged records (the Analysis Population).
The original electronic data was of poor quality, incomplete, inconsistent, and retrieval was
difficult. The physical size (weight, height) of the suspect was often not recorded in either the
electronic or hardcopy reports and, if it was, we noted that the majority of the suspects were 175
pounds. Included in the approximately 730,000 calls for service/incident reports were 1,284
incidents where a CED was deployed. Only 951 (75%) of the 1,284 CED deployments could be
statistically analyzed primarily because of the data merging challenges. In summary, the Audit
Team reviewed all 1,284 of the hardcopy CED calls for service/incident reports. They did not
include all of the 1,284 reports in the statistical analysis because CED incidents that were lost
during the electronic data merges would have biased the results of the analysis.

3

One of the difficulties in conducting any research and/or analysis based on incident reports is that the researcher
does not have the capability of independently assessing the accuracy of what is reflected in the reports.

Mi'&'bX
Rodriguez, P.c.

4

The current structure for data management is organizationally deficient and under-staffed. This
combination of factors may produce inefficiencies in data transmission, increases in
measurement and coding errors, and an overall inability to create a template for connecting
disparate pieces of information to support overall HPD Management processes. The
implications are even more severe, however, if there are efforts to increase situational
awareness for HPD officers that require data in real time.
HPD was fully aware of these data management challenges, but had insufficient resources to
resolve this issue. We understand that HPD is firmly committed and is actively engaged in
identifying a vendor to replace its present online offense report writing system with a more
efficient and effective version or model than what is currently in place.
CED Deployments
As of 2006, the City’s population had the following racial/ethnic distribution: Latino (41.9%),
Anglo (27.6%), African American (24.7%), and Other Groups (5.8%).
The Audit Team compared the race/ethnicity of suspects in the Total Service/Incident
Reports Analysis Population to the race/ethnicity of the suspects noted by HPD in the CED
Service/Incident Reports. The results of our comparison are as follows:

Suspect’s
Race/Ethnicity

Total Service/Incident Reports
Analysis Population

African
American
Latino
Anglo
Other Groups

46.0 %
28.2 %
24.4 %
1.4 %

CED
Service/Incident
Reports
66.9 %
23.5 %
9.0 %
0.6%

Difference
20.9 %
-4.7 %
-15.4 %
-0.8 %

Based on the above analysis, African American suspects were involved in a proportionally
greater number of total Service/Incident Reports analyzed as well as CED service/incident
reports. In addition, the proportion of CED Service/Incident Reports was 20.9% more than the
total service/incident reports. The Latino, Anglo, and Other Group suspects were involved in
proportionally less service/incident reports.
According to HPD, the Department-wide officer demographics during the scope period were as
follows:
Average for the Period January 1, 2005 through June 30, 2007
Female
Male Classified
Classified
Total Classified
% of Total

Officer’s
Race/Ethnicity
Anglo
Latino
African American
Other Groups
Total

Mi'&'bX
Rodriguez, P.c.

1,268
529
451
95
2,343

126
49
98
4
277

5

1,394
578
549
99
2,620

53.2 %
22.0 %
21.0 %
3.8 %
100.0 %

The Audit Team compared the race/ethnicity of officers in the Total Service/Incident
Reports Analysis Population to the race/ethnicity of the officers noted by HPD in the CED
Service/Incident Reports. The results of our comparison are as follows:
Officer’s
Total Service/Incident Reports
CED
Race/Ethnicity
Analysis Population
Service/Incident Reports Difference
25.1 %
17.3 %
-7.8 %
African American
24.3 %
27.9 %
3.6 %
Latino
46.2 %
52.3 %
6.1 %
Anglo
4.4 %
2.5 %
-1.9 %
Other Groups
Based on the above analysis, the positive values in the Difference column indicate the officer
racial/ethnic group was involved in a proportion of CED events that was larger than the
proportion represented by it for all HPD incidents in the Analysis Population. Negative values
indicate the officer racial/ethnic group was involved in a proportion of CED events that was
smaller than the proportion represented by it for all HPD incidents in the Analysis Population.
Our analysis further indicated that 803 officers deployed their CED for a total of 1,417
deployments during the Scope period that resulted in 1,284 service/incident reports summarized
as follows:
Number of Times an
Officer Deployed a
CED
1
2
3
4
5
6
7
8
9
10
11
12
13
Total

Total Number of
Officers
492
156
90
26
23
7
1
4
2
0
0
1
1
803

Total Number of
Deployments
492
312
270
104
115
42
7
32
18
0
0
12
13
1,417

In the table above it should be noted that more than one officer may have deployed their
CED related to an individual service/incident report.

Mi'&'bX
Rodriguez, P.c.

6

Geo-Coordinates
Within the Analysis Population, only 1,020 (79%) of the 1,284 CED incident reports could be
analyzed primarily because of incomplete electronic data. Furthermore, the geographic
information was missing from 30 of the 1,020 electronic CED call for service/incident reports.
In addition, there was contradictory geographic information appearing in about 1,000 of the
valid electronic call for service/incident reports in the Analysis Population. Among the errors
that we noted, the key map address and zip code listed for a single call for service/incident
did not coincide geographically. In some cases, the locations were up to 20 miles apart.
According to HPD Management, the reporting address and the incident address may not be
the same.
CED Inventory
HPD requires each patrol officer to carry a CED with two cartridges on their Sam Browne
(tool belt). The CED and related cartridges each have a unique bar code serial number that
HPD maintains in a database at the Academy. MFR randomly selected 100 officers and
recorded the serial numbers on their CED and cartridges. MFR traced all 100 of the CED
serial numbers to the database records without exception. The database containing the
inventory of cartridge serial numbers was not current. However, MFR traced the serial
numbers for 110 of the 173 cartridges to the inventory records. We understand that the
officer in charge of CEDs at the Academy plans to trace the remaining serial numbers to the
inventory records once the records have been updated.
MFR noted that 26 of the 100 officers selected were not carrying their second CED
cartridges on their tool belts as required by GO # 400-26. In addition, one of the 26 officers
did not have either cartridge, meaning that their CED could not be deployed in the probe
mode.
MFR further noted that the Digital Power Magazines (DPMs), commonly known as batteries,
were being replaced on the CED devices at a high frequency. The lithium battery in the
DPM does not appear to have the 10 year shelf life as claimed by the manufacturer.
As a partial test of the DPM, officers were required to spark test their CED at roll call. We
noted that neither the roll call supervisors nor the officers attending roll call were in
compliance with HPD spark test procedures. To obtain a replacement DPM, HPD requires
the officers to go through a process that takes the officer away from his/her duties.
Effectiveness of CED in Controlling a Suspect
During our audit of the call for service/incident reports that resulted in CED deployments, the
device was reported as being effective in controlling the subject in 77% of the calls. It
should be noted that this does not mean that the device was effective in controlling
the suspect’s behavior with initial deployment or only one trigger pull or cycle of the
CED. In some cases, the officer deployed the device in probe mode and then resorted to
using the CED in the drive stun mode. In other cases, the officer used multiple cycles in the
probe mode to gain control of the suspect.

Mi'&'bX
Rodriguez, P.c.

7

The reasons given in the call for service/incident reports that the CED was not effective in
securing and controlling the suspect are summarized as follows:

Reason for CED being ineffective
The suspect appeared to be either under the
influence of alcohol and/or drugs or experiencing
some form of mental health crisis
Suspect’s clothing (i.e. was too loose and/or too thick
for the probes to penetrate)
One or both of the probes missed the suspect
Probes hit the subject who subsequently pulled them
out
CED was defective and did not fire the probes
Total Percentage

Percent of total
number of
calls for
service/incidents

11 %
4%
4%
2%
2%
23 %

Number of CED Cycles Used When Deployed
When the trigger is initially pulled to deploy the probes which attach to a suspect, a five
second pulse of electrical energy is transmitted into the suspect. This is known as the initial
or first cycle. Should the suspect not comply with the officer’s directions when the initial
cycle is administered, the officer can continue to pull the trigger and administer additional
pulses provided the probes are still attached to the suspect. The second and all subsequent
trigger pulls transmit a minimum five second electrical pulse each time.
During our audit of the call for service/incident reports related to the deployment of the
CEDs, we noted the following number of cycles being used:

Number of Cycles
One cycle (the initial cycle)
Two cycles
Three cycles
Four cycles
Five cycles
Six through ten cycles
Greater than ten cycles
Total Percentage

Mi'&'bX
Rodriguez, P.c.

8

Percent of
the total
reports
reviewed
38 %
31 %
15 %
5%
3%
5%
3%
100%

HPD has a policy that requires the officer to reevaluate after each cycle; however, the policy
does not limit the number of cycles that should be deployed. While there are legitimate
reasons for not limiting an officer’s discretion to use multiple cycles, it is suggested that
incidents in which the CED is cycled more than five times be closely reviewed for
compliance with applicable GOs.
Impact of Alcohol and/or Drugs Consumed by Suspects
In 15% of the calls for service/incidents, the call for service/incident reports indicated that
the subject appeared to be under the influence of alcohol. In 27% of the calls for
service/incidents, the reports stated that the subject appeared to be under the influence of
drugs and/or possessing drugs.
Officers did not note in their written call for service/incident reports whether the subjects
appeared to be under the influence of drugs and/or alcohol in 58% of the calls. The Audit
Team did not attempt to determine whether the officer, if asked, would have said that drugs
and/or alcohol might have been a contributing factor to the behavior which resulted in the
need for the deployment of the CED. However, it is highly likely that a significant
percentage of these subjects were under the influence of drugs and/or alcohol given the
behavior that was described by the officer in the call for service/incident report.
Injuries to Officers
Since June 2004, the estimated total number of Workers’ Compensation claims by the
officers decreased by an accumulated approximate 20%. This began in May 2003. Both
the decreases in claims and claim amounts began prior to the incorporation of the CED
program at HPD. This downward trend continued through the Scope period. With the
passage of time it may be possible to determine if there was yet further reduction in the
number of Workers’ Compensation claims consistent with the CED policy.
Year
Number of Claims Total Claim Amount
2000
255
$ 1,494,341
2001
278
1,668,954
2002
276
1,095,361
2003
283
2,233,479
2004
271
1,563,661
2005
258
1,152,195
2006
233
738,028
January through June 2007
117
$
105,900

Mi'&'bX
Rodriguez, P.c.

9

HPD Training Academy-CED Training
A SHSU CCJ member of the Audit Team attended the actual training received by cadets at
the Academy. The students and instructors were observed during the classroom and
hands-on portions of the training by the same SHSU CCJ team member who had read and
analyzed the CED service/incident reports. During the observation, the team member paid
particular attention to compliance with CED policy and whether any concerns noted during
the analysis of the reports might be attributable to or influenced by the training provided.
There were numerous issues involved in the training of cadets on the use of the CED.
There was a lack of emphasis placed on the potential danger to a suspect when numerous
cycles are used after a CED was deployed, particularly on a suspect who might be at risk for
excited delirium. According to the HPD GOs, excited delirium is a state of extreme mental
and physiological excitement, characterized by extreme agitation, hyperthermia, hostility,
exceptional strength, or endurance without fatigue. Additional concerns center around the
lack of training which more closely resembles actual situations such as a moving target,
heavy clothing, cuffing a suspect who has been subjected to a CED deployment, and
transitioning to another type of use-of-force if multiple cycles prove ineffective. Also, while
each cadet deployed the CED, a demonstration of true proficiency under field conditions
was not observed. There was no requirement for every cadet to experience a CED
deployment on themselves; however, the CED was deployed on volunteers from the
cadet class.
Suspect Complaints Related to CED Deployments
There were 55 complaints filed pertaining to 59 officers. The disposition of complaints is
summarized as follows:
CED Complaints 4
No Disposition
Exonerated
Information
Never Formalized
Not Sustained
Open Case
Sustained
Unfounded
Total CED Complaints

4

Total
12
13
1
2
9
4
3
11
55

No Disposition – CED activity was not the focus of the complaint and the investigation found CED usage to be
proper and appropriate.
Exonerated: Incident occurred, but was lawful and proper.
Information: No evidence to prove that an incident even occurred.
Never Formalized: Complainant refused to make a formal written statement or if a written statement was made,
refused to swear or affirm that the statement was true (notarized).
Not Sustained: insufficient evidence to either prove or disprove justification for the incident.
Open Case: Investigation is on-going.
Sustained: Evidence is sufficient to prove the allegation.
Unfounded: Allegation is false or not factual.

Mi'&'bX
Rodriguez, P.c.

10

The complaint analysis indicates that complaints in which a CED is mentioned have a distinctive
racial propensity as follows:
Officer’s
Race/Ethnicity
Anglo
Latino
African American
Other
Total

Total
Classified
Officers
1,394
578
549
99
2,620

Total
Percentage Complaints by
of Total Race/Ethnicity
53.2 %
27
22.0 %
9
21.0 %
20
3.8 %
3
100.0 %
59

Percent By
Complaints
Difference
7.4 %
45.8 %
6.9 %
15.2 %
-12.9 %
33.9 %
-1.3 %
5.1 %
100.0 %

Based on the above analysis, the positive values in the Difference column indicate the officer
racial/ethnic group was involved in a proportion of CED complaints that was less than the
percentage of Total Classified Officers. Negative values indicate the officer racial/ethnic group
was involved in a proportion of CED complaints that was larger than the corresponding
percentage of the Total Classified Officers.
Initiation and Nature of Calls for Service
Based on our review of the call for service/incident reports, the source of the calls related to
CED deployments was as follows:
Percentage of
Source of Call
Total Calls
Officer dispatched
55 %
Officer initiated contact with suspect
33 %
Traffic related contact
12 %
Total
100 %
During our audit of the call for service/incident reports related to CED deployment we noted that
the nature of the original calls for service could be summarized as follows:

Nature of Call
Drug related (most calls were initiated by the officer)
Disturbance
Assault
Misdemeanors (e.g. driving while intoxicated, criminal
mischief, and theft)
Traffic offense
Crisis intervention for suspects having a mental health issue
Family violence
Criminal trespass
Other felonies
Automobile theft
Total Calls for Service

Percentage of
Total Calls
15 %
15 %
13 %
13 %
11 %
11 %
7%
6%
6%
3%
100 %

The total CED deployments during the Scope period represent 0.47% of the
approximately 273,000 individuals who were incarcerated in the City’s Jail system.

Mi'&'bX
Rodriguez, P.c.

11

Use of a CED Instead of Deadly Force
The Audit Team identified the following inconsistencies related to the CED being an
alternative to deadly force such as:
•
•
•

•

According to the October 20, 2004 Request for Council Action (RCA), the CEDs
were “to be used as an alternative to deadly force”.
One of the objectives of the HPD CED Program was “in limited situations, provide an
alternative to deadly force”.
As part of CED training cadets were shown a videotape. In the video tape the Police
Chief stated, in essence, that after introduction of CEDs in Phoenix that officer
involved shootings went down by one-half and that it (CED use) was an alternative to
deadly force.
However, according to HPD GOs, CEDs are classified as intermediate weapons and
are not a substitute for lethal force.

HPD Management did not believe there were any apparent inconsistencies.
There was no statistical evidence that the introduction of CEDs served as a substitute for
the use of firearms by an officer. The results of the audit do not indicate a reduction in the
number of officer involved shootings since the introduction of the CED. This is not
surprising as the opportunities to use a CED instead of a firearm are very limited. For
example, it would be highly unusual for an officer to rely on a CED if the suspect was armed
with a firearm. With the passage of time, as well as the introduction of the collection of data
for intermediate weapons, it may be possible to determine additional consequences, if any,
of the CED policy.
On 53 of 1,284 occasions the officers used the CED as an alternative to deadly force even
though they were not required to do so by HPD policies and State law. It should be noted
that just because the officer would have been legally justified in using deadly force, it does
not mean that the officer would definitely have chosen this option if not for being equipped
with the CED. However, a review of the reports indicated situations in which other
intermediate weapons were unlikely to have been used due to officer safety concerns: thus,
if the CED had not been available, the use of deadly force would have been more likely.
Statistical Analysis of Call for Service/Incident Reports
With regard to the call for service/incident reports, the results from the CED statistical
analysis suggests that certain combinations of officer and suspect characteristics result in
an increased probability of CED utilization. For example, depending on how the race of the
officer and suspect were paired, it was possible to see either significant increases or
decreases in the rate of CED utilization. The Audit Team noted deviations from these
general demographic patterns when the analysis was performed at the City Council District
level.

Mi'&'bX
Rodriguez, P.c.

12

Numerous statistical and research design challenges had to be overcome. As previously
mentioned, only 52% of the electronic call for service/incident reports for the Scope period
could be statistically analyzed. In addition, only 75% of the CED call for service/incident
reports could be statistically analyzed without compromising the validity of the overall
analysis.
Due to data limitations, only a limited amount of statistical controls were used. However, the
importance of statistical controls cannot be overstated since they affect the strength of the
General Observations noted below. For example, when the Audit Team controlled for City
Council Districts the relationships between officers and suspects disappear or change.
•
•
•
•
•
•
•

•

The likelihood of having a CED involved in a call for service/incident is approximately
0.08%. In other words, for every 10,000 calls for service/incidents 8 involved the use
of a CED.
Among officers, there are no gender differences in the overall likelihood of using a
CED.
African American officers were much less likely to use a CED than Anglo or Latino
officers. Yet when controlling CED incidents by City Council Districts, African
American officers may use a CED at the same rate as their Anglo counterpart.
Latino and Anglo officers were equally likely to use a CED.
African American suspects were much more likely to be involved in a CED incident
than Anglo or Latino suspects. Latino suspects were somewhat more likely to be
involved in a CED incident than Anglo suspects.
Male suspects were much more likely to be involved in a CED incident than female
suspects.
African American officers were much less likely to deploy a CED than Anglo and
Latino officers when a suspect was an African American. African American officers
were equally likely to deploy a CED as Anglo and Latino officers when the suspect
was an Anglo. When the suspect was a Latino, African American officers were
equally likely to deploy a CED as Anglo officers and somewhat less likely to deploy a
CED than Latino officers.
CED incidents were much more common in Council Districts D and H than in all
other Districts.

Mi'&'bX
Rodriguez, P.c.

13

HPD Officer Focus Group Sessions
Six Focus Groups were formed with approximately 25 officers each to discuss CED related
performance and HPD policies. The significant observations made by the officers were as
follows:
•
•
•

•
•
•

•

The majority of the officers claimed that the CED was an effective weapon and
that they wanted HPD to continue using it.
Several officers claimed that the CED is becoming well known because once the
suspect sees the laser dot from the CED on his/her body, the suspect pleads
with the officer not to deploy the CED.
Most officers wanted to retain the CED as an intermediate weapon; however,
HPD needs to reduce the paperwork requirements related to call for
service/incident reporting, cartridge, and DPM replacements to be similar to that
of other intermediate weapons.
Officers stated the need for more training in writing reports related to CED calls
for service/incidents.
Numerous officers appeared to need clarification on the applicable GOs for
CEDs.
Several officers want the CED to be optional for carrying, like the other
intermediate weapons. Many of the officers commented on the weight of their
tool belt, how it affected their ability to enter and exit their patrol cars, as well as
how the tool belt weighed them down during a foot pursuit.
Officers wanted the HPD CED policies to be changed so that they could use their
CED on a suspect who was fleeing, as many of the officer were unsure how to
apply the applicable GOs.

The following comments came to the Audit Team’s attention and are not directly attributable
to CEDs:
•
•
•

Certain officers appeared to have a lack of trust for HPD executive management.
Officers believed that executive management did not adequately support them.
Officers really appreciated the opportunity to be heard. According to certain
officers, until this session no one had asked them for their opinion on work
related matters.

For more details of the issues summarized in this assessment please review the attached
detailed report.

Mi'&'bX
Rodriguez, P.c.

14

RECOMMENDATIONS
The following recommendations are noted in the attached detail report.
POLICY
HPD’s policy on the use of CEDs is well written and in line with the best practices of the law
enforcement profession. The problems with its implementation may be the result of training
and supervision issues as opposed to the wording of the policy. However, we recommend
that HPD consider altering the policy so as to:
•

•
•
•
•

•
•
•
•
•

Require the CED download information to be incorporated into the initial report so
that it can easily be obtained and reviewed. This should be in an electronic format.
The current system does not allow for this recommendation to be implemented; it is
suggested that this option be explored when designing the new system.
Place additional training emphasis on the officer’s evaluation of the situation after
each CED deployment and before subsequent deployments.
Perform a detailed assessment of all incidents in which the CED is used in excess of
five cycles to determine compliance with HPD GOs.
Require medical screening at the jail of any subject against whom a CED is
deployed.
Require immediate medical screening and transportation to a medical facility if a
CED is deployed and the subject appears to be experiencing excited delirium. While
the research that has been done on the physical effects of being shocked by a CED
has not definitely concluded that deploying a CED on a person suffering excited
delirium has an adverse effect, a majority of the studies strongly suggest that this
particular group is the one most likely to be adversely affected. Given the sudden
onset of the effects of excited delirium, it is suggested that a subject should be
transported to the hospital immediately for medical care.
Clarify when it is appropriate to use a CED on a subject who is fleeing from an
officer. The HPD GOs set out the policy clearly, but the focus groups indicated there
may be a lack of understanding in the implementation on the part of some officers.
Provide for tracking of CED deployment and the number of cycles used in the HPD
Early Intervention System (EIS).
Continue to prohibit use of the CED by multiple officers at the same time. The HPD
GOs set out the policy clearly; however, it should be emphasized in roll call training.
Review and revise GO # 400-26 to have an internal review process to ensure that
roll call supervisors are adhering to GO # 400-26. The review should be
documented to evidence that the work had been performed.
Review the policy related to the replacement of CED cartridges as the officers have
such a reluctance in carrying their extra cartridge primarily because of their fear of
losing and/or damaging it as well as the related paperwork.

Mi'&'bX
Rodriguez, P.c.

15

TRAINING
We recognize that the Academy has been called upon to train a large number of new cadets
and to provide in-services training to the entire force. We recommend that the CED training
continue to be reviewed. One of the tools that HPD has used that appeared to be
particularly effective was the use of training bulletins and roll call training to emphasize
certain aspects of the GOs. We recommend that the training be reviewed to:
•
•

•

•

•
•

•
•

Determine whether additional scenarios should be included in the Field Problems
Program or Simulations which specifically involve the use of CEDs in situations
where the CED appears to be less effective.
Continue to emphasize when a suspect’s behavior is actively aggressive so as to
warrant CED use. The call for service/incident reports reflect a lack of understanding
by a small number of officers of the difference between passively resisting and
aggressively resisting and the alternatives available to overcome the resistance.
Continue to emphasize when it is appropriate to use a CED on a fleeing subject.
Part of the training should be to demonstrate, or, at least, explain the risk to the
subject of sustaining injury due to falling after being suddenly incapacitated. Also,
training should emphasize the difficulty of actually hitting a fleeing subject with both
darts.
Demonstrate how a subject has difficulty in complying with orders given by an officer
while being subjected to a CED deployment. For instance, have the volunteers who
are subjected to a CED deployment, attempt to comply with common orders such as
placing their hands behind their backs.
Emphasize how to use the initial incapacitation period as an opportunity to gain
control of the subject. For instance, demonstrate how to assume the proper position
for handcuffing the subject.
Continue to emphasize the risk to subjects of being placed in a position which
impairs respiration after deployment of the CED. During the initial CED training,
instructors should demonstrate handcuffing the subject in a manner which does not
impair respiration and have cadets demonstrate proficiency. While it is understood
that this is emphasized later in the cadet training, consideration should be given to
demonstrating it during the initial training.
Continue to emphasize how to accurately report the circumstances that warrant CED
use including describing the actions of the suspect and any warnings given to the
suspect.
Emphasize actual situations faced by officers in the use of CEDs and incorporate
different levels of use-of-force and how to transition from one type of use-of-force to
another. For instance, add specific scenarios in which the excited delirium may be
present and/or the CED is ineffective in controlling the suspect’s behavior. It is
recommended that the actual calls for service/incident reports and/or complaints be
used as the basis for designing the scenarios.

Mi'&'bX
Rodriguez, P.c.

16

•

•

Continue to emphasize the signs of excited delirium and the proper steps to be taken
to lessen the likelihood of serious injury if a CED is used. The training should
continue to emphasize the potential danger to a suspect if the suspect is
experiencing excited delirium and the need for immediate medical treatment.
Continue to train on the various roles of officers during a call for service/incident
where the CED is deployed. It is suggested that this be part of the scenario based
training and that the various roles be given a designation so that officers are able to
interchange the roles in the field. The current training films should be reviewed with
an emphasis towards moving towards consistency in the message being delivered
by HPD policy and training. It is recommended that the time currently being used to
explain the workings and history of the CED would better be utilized to address some
of the concerns noted in this report.

REPORTS
The reports are generally well written, but there are areas of deficiency. There were
concerns expressed during the Focus Groups related to reporting requirements surrounding
CED use. We recommend that the report form and process be revised so as to:
•
•
•
•

Continue to emphasize that the report should reflect the behavior that warranted the
deployment of the CED.
Provide a drop down menu where appropriate. For example, the menu could include
a place to indicate whether a verbal warning was given before the initial and
subsequent CED use and whether the subject voluntarily complied.
Require the CED download information (a history of spark testing and deployments
since the last download) be incorporated into the initial call for service/incident report
when the new system allows.
Require the new system to have the capability of generating geo-coordinates from
addresses. Not only could geo-coordinates more accurately pin-point the location of
an important scene, the geo-coordinates would also fit into the current scheme HPD
uses for locating an incident. A geo-coordinate pair uniquely maps to one key map
address, one zip code and one City Council District.

DATA MANAGEMENT PROCESSES
Develop a process to enhance the forthcoming modernization in data management by
means of a process audit. The current structure of data management seriously impairs
efficient data processing and data acquisition. The audit would seek to identify, document,
and merge efficient processes, remove impediments to efficient processing, and combine
these methods with the new data processing capabilities now being acquired.

Mi'&'bX
Rodriguez, P.c.

17

DIVERSE PATROL EXPERIMENT
It is clear that a complex set of factors have yet to be investigated. Among these variables
are measures (to be developed) that capture the threat that officers face, the general
context in which the CED incident occurs, as well as the relation between an officer's
productivity, arrest history, and his/her use of CEDs. In order to obtain a more thorough and
complete understanding of the dynamics of these new variables we recommend that it
would be advisable to conduct a series of natural experiments. These natural experiments
would be designed to evaluate, for example, the role of officer and suspect race and
ethnicity in the probability that a CED incident occurs.
CED EQUIPMENT
The CED Equipment was adequately secured and generally accounted for by the Academy.
The Audit Team identified and recommends that HPD consider the following areas for
potential improvement:
•

•
•
•

Consider installing bar code scanners in all police stations to facilitate the recording
and issuance of the CEDs, cartridges, and DPMs. Furthermore, the scanners would
be part of a centralized HPD CED tracking system that would contain the serial
numbers of the CEDs and cartridges issued to each officer. To the extent possible
and practical, the HPD CED tracking system should be incorporated into the new
system being acquired.
Review and assess the DPM failures and the amount of time officers are expending
to get them replaced. Consider a more cost effective and efficient replacement
process.
Record the serial numbers of all CEDs and cartridges upon receipt from the
manufacturer, and subsequent issuance to the officer.
Implement a process to improve the controls over training cartridge inventory. 5

5

Corrective Action: HPD has revised the training cartridge process. Process improvements include, the CED training
cartridges are now a blue color and can be easily identified by HPD.

Mi'&'bX
Rodriguez, P.c.

18

ACKNOWLEDGEMENTS
The HPD executive team was very cooperative during the audit and facilitated our numerous
and often very challenging information requests throughout the engagement. They made their
best efforts to provide complete and accurate data to the extent that it was feasible and
practical. In certain instances, extra hours had to be worked by HPD management and staff to
obtain the available data for the Audit Team. The Audit Team would like to thank HPD for the
opportunity to attend the CED training and to have the experience of deploying a CED. Certain
members of the Audit Team also appreciated the opportunity to participate in HPD’s Ride Along
Program to gain an understanding of a patrol officer’s normal work day.
The Audit Team would like to especially thank those officers who were randomly selected for
the Focus Groups, for being very forthright with their observations and comments. Many of the
officers had to extend their work day or attend on their day-off on short notice to be able to
participate in these very productive sessions.
A special acknowledgement should be given to Executive Assistant Police Chief
Charles A. McClelland and Sergeant George Alderete for the amount of extraordinary effort and
patience they had with the Audit Team combined with their time for facilitating our meetings and
information requests. Furthermore, they should be commended for their responsiveness to
accept and implement certain recommendations during fieldwork. Last but not least we would
like to thank Chief Harold H. Hurtt for his cooperation and support throughout the entire audit.
*****
Details of the assessment are contained in PART I through PART V of this report.

Mi'&'bX
Rodriguez, P.c.

19

AUDIT TEAM CONTACT INFORMATION
SAM HOUSTON STATE UNIVERSITY
COLLEGE OF CRIMINAL JUSTICE TEAM

UNIVERSITY OF HOUSTON, CENTER
FOR PUBLIC POLICY TEAM

Margo L. Frasier J.D.
Senior Associate
MGT of America, Inc.
502 East 11th Street
Austin, Texas 78701
(512) 476-4697
mfrasier@mgtamer.com

Jim Granato, Ph.D.
Director, UH Center for Public Policy
University of Houston
104 Heyne Building
Houston, Texas 77204-5021
(713) 743-3887
(713) 743-3978 fax
www.uh.edu/cpp

Brian Laton Ph.D.
Jennifer Schulenberg, Ph.D.
Criminal Justice Center
Sam Houston State University
Huntsville, Texas 77341
brian.laton@shsu.edu
jennifer.schulenderg@shsu.com

Renee Cross
Associate Director, UH Center for
Public Policy
Lecturer, Department of Political Science
University of Houston
104 Heyne Building
Houston, Texas 77204-5021
(713) 743-3972
(713) 743-3978 fax
www.uh.edu/cpp/cross.htm
www.uh.edu/cpp

PROTOTYPE, FUSION &
MODELING, LLC
Terrence J. Mayes, CTO
Prototype, Fusion & Modeling, LLC
Software & Support
c/o UH Center for Public Policy
University of Houston
104 Heyne Building
Houston, Texas 77204-5021
mayes@pfmtx.dnsalias.com

Mark P. Jones, Ph.D.
Professor
Department of Political Science
Rice University
Houston, Texas 77251-1892
(713) 348-2107
mpjones@rice.edu

Matthew G. Soltis, COO
Prototype, Fusion & Modeling, LLC
Company Information
c/o UH Center for Public Policy
University of Houston
104 Heyne Building
Houston, Texas 77204-5021
soltis@pfmtx.dnsalias.com

William Reed
Associate Professor
Political Science
Rice University
www.koldekrig.com
M.C. Sunny Wong, Ph.D.
Assistant Professor of Economics
Department of San Francisco
2130 Fulton Street
San Francisco, CA 97117
415 422-6194
mwongll@usfca,edu
http://www.usfca.edu/fac-staff/mwong11

Mi'&'bX
Rodriguez, P.c.

Tom C. Duncavage, CEO
Prototype, Fusion & Modeling, LLC
Business Inquiries
c/o UH Center for Public Policy
University of Houston
104 Heyne Building
Houston, Texas 77204-5021
duncavage@pfmtx.dnsalias.com

20

Certified Public Accountants

PART I
DETAILED BACKGROUND AND AUDIT METHODOLOGY

PART I

Mirst°x
Rodriguez, P.C.

DETAILED BACKGROUND AND AUDIT METHODOLOGY
DETAILED BACKGROUND
The CED is a relatively new electronic intermediate weapon where technology is often
misunderstood. The purpose of this background section is to clarify certain common
misunderstandings and to provide a basis for the detailed observations and comments in this
report. This background section will describe:
•
•
•

CED procurement
CED program objectives and select policies
The manufacturer’s comments on:
1. Who has purchased CEDs?
2. Legal Liabilities and Risks
3. CED Technology

CED Procurement
On November 3, 2004, the City Council awarded a $4.68 million, five year contract (Number
56426) (the Contract) for the purchase of Taser X26 model CEDs, related consumables, and
supplies to G.T. Distributors, Inc. (Contractor) and Taser International, Inc. (Manufacturer).
HPD issued CED policies subsequent to the purchase of the CEDs that classified CEDs as
intermediate weapons.
Intermediate weapons are defined by HPD in GO# 600-17 as:
•
•
•
•

Baton
Oleoresin Capsicum (OC) spray
Soft impact weapon (i.e. Beanbag Shotgun)
CED (i.e. stun gun or Taser)

The HPD CED policy further states that CEDs are not a substitute for lethal force.
The following purchases for CEDs were made under the Contract:

No. of
CEDs

Total
Amount

11/15/2004 PG00010006809

3,700

$3,408,388

3/31/2006 PG00010017886

395

437,380

6/5/2006 PG00010008078
Totals

132
4,227

125,832
$3,971,600

Date

Mir&'ox
Rodrih'uez, P.C.
~""""",Ac<>Do>n_

P.O.
No.

21

According to the Contract, the CEDs shall be free from defects in workmanship and materials
for a period of one year from the date of the Director’s letter accepting the Taser Devices
supplied to the City. Additionally, the Contractor and Manufacturer shall provide a “No
Questions Asked” Extended Warranty on each Taser Device purchased by the City which shall
begin the day following the expiration of the one year free warranty. The Contractor and
Manufacturer shall provide each Taser Device purchased by the City, with the Warranty XDPM
module programmed to reflect the five year warranty.
CED Program Objectives and Select Policies
In conjunction with the purchase of the CEDs, HPD developed a CED Program that included
CED policies, procedures, and related training for the patrol officers. The CED Program has the
following objectives:
•
•
•
•

Assist officers in securing and controlling combative individuals.
Reduce injuries to officers and suspects.
Reduce financial impact of civil liability in use-of-force calls for service/incidents.
In limited situations, provide an alternative to deadly force.

HPD used a series of circulars and directives to provide the officers with HPD Policies and
Procedures to meet the CED program objectives.
During the Scope period HPD amended existing and/or issued various circulars and directives
to include the policies and procedures related to the CEDs. To reduce the confusion to the
officers and based on the results of a CED conference hosted by HPD, the CED circulars and
directives were consolidated by HPD into GO # 400-26 that was issued on March 26, 2007.
The GO clearly states when an officer is to use a CED as follows:
“CEDs are authorized for use against suspects who are actively resisting or exhibiting
active aggression, or to prevent individuals from harming themselves or others.”
GO # 400-26 also requires that officers keep CED cycling to a minimum, especially against
persons in an excited delirium, and use only the force necessary to apply traditional restraint
devices and affect an arrest. Furthermore the GO defines excited delirium as a state of extreme
mental and physiological excitement, characterized by extreme agitation, hyperthermia, hostility,
exceptional strength, or endurance without fatigue.
The officers are more accountable for the CED than any other intermediate weapon. The
officers are required to complete documentation related to any use of the CED for example.
GO # 600-17 requires that:
“In every situation in which a CED is discharged (cartridge firing or drive stun), even if
the suspect was not struck, officers will notify the Command Center and write a detailed
incident report, and the on-scene supervisor will make a supplement.”
Furthermore, officers are required to complete reports related to the replacement of CED
cartridges and DPMs. Roll call sergeants are required to document their witnessing of each
officer’s CED spark test at the beginning of each shift.

Mir&'ox
Rodrih'uez, P.C.
~""""",Ac<>Do>n_

22

Manufacturer’s Comments
To better understand (1) who has been purchasing CEDs, (2) certain legal liabilities and (3) risk
as well as the technology related to the CED, MFR requested certain CED background
information from the manufacturer of the TASER X26.
1. Who has Purchased CEDs?
TASER® is a registered trademark. The information provided by the Manufacturer is
summarized below. MFR did not perform an audit on the data provided by the
Manufacturer. The Manufacturer refers to the CED throughout their literature as an
electronic control device (ECD).
The Manufacturer claims to have sold 327,000 units to over 12,500 law enforcement and
U.S. military agencies, as well as law enforcement agencies in 44 countries overseas.
Over 4,300 agencies provide CEDs specifically to all their patrol officers. According to
the Manufacturer, their CEDs have been deployed approximately 470,000 times on
suspects and 610,000 times on volunteers.
Largest Texas TASER X26 CED sales were as follows:
1. Houston Police Department
2. Fort Worth Police Department
3. El Paso Police Department
4. Harris County Sheriff’s Department
5. Dallas Police Department
6. Lubbock Police Department
7. Austin Police Department
8. Pasadena Police Department
9. Montgomery County Sheriff’s Department
10. Fort Bend County Sheriff’s Department
11. San Antonio Police Department
12. Galveston Police Department

Mir&'ox
Rodrih'uez, P.C.
~""""",Ac<>Do>n_

23

2. Legal Liabilities 1 and Risk
According to the Vice President and General Counsel of the Manufacturer:
“Every law enforcement use-of-force creates some risk of civil and criminal liability and
litigation. For force equipment manufacturers and distributors, there is civil product
liability litigation risk. For law enforcement agencies there is civil risk and for individual
officers there are civil and criminal excessive use-of-force liability and litigation risks
under 42 U.S.C. § 1983 and state specific causes of action.
Law enforcement use-of-force risks and device manufacturer risks are separate and
distinct liability and litigation risks. Manufacturers and distributors are not liable for
excessive force claims and law enforcement agencies and individual officers are not
liable for product liability claims.
The liability and litigation risk profile varies among the different use-of-force tools and
techniques used by law enforcement. As a general rule, the greater the risk of injury to
suspects caused by the use-of-force tool, the greater the risk of liability and litigation.
Since the TASER® brand Electronic Control Device (ECD) has proven to significantly
reduce injuries to suspects by up to 79%, and an estimated 9,000 lives have been
saved by use of the TASER ECD, it has one of the lowest liability and litigation risk
profiles of any use-of-force tool or technique. Courts have generally held that proper
use of the TASER ECD is not excessive use-of-force and the reduction in use-of-force
claims against law enforcement resulting from use of the TASER ECD is well
documented and is one of the economic benefits of deploying TASER ECDs.
While TASER International, Inc. has been named in a number of product liability
lawsuits arising from law enforcement use of the TASER ECD, TASER International
has been successful in getting dismissals, summary judgments, or favorable jury
verdicts in 51 lawsuits to date with more expected. The suspect injury or death
lawsuits are frivolous and the plaintiffs have been unable to prove that the TASER
ECD is defective or was an unjustified cause of any injury or death, both of which are
essential elements to establish product liability. Autopsy reports and medical experts
have determined that the cause of in-custody deaths have been due to complications
from drug intoxication or pre-existing medical conditions and not from the TASER
ECD.”

1

According to HPD management, the City has been served with only one lawsuit related to a CED incident during the
Scope period.

Mir&'ox
Rodrih'uez, P.C.
~""""",Ac<>Do>n_

24

3. CED Technology
To better understand the underlying basics about electricity here are certain relevant
facts:

•
•
•

High voltage cannot cause injury if the current is very low.
Low voltage CAN injure you if the current is high enough.
CED devices operate at low average currents (0.0021 - 0.0036 Amperes).

A CED device produces 17-19 electrical pulses per second. Each electrical pulse
mimics the electrical signal which is sent from the brain to the skeleton muscles causing
the muscles to contract and release. One pulse causes a twitch; 17-19 pulses per
second causes the appearance of a smooth contraction and the incapacitating effect.
While the CED device produces an initial 50,000 volts to create a spark that will transmit
electricity through 2 inches of clothing, each short pulse of 400 volts actually enters the
body. When compared to a static shock from a doorknob (35,000 to 100,000 volts) or a
Van de Graff Generator (1,000,000 to 20,000,000 volts), a common display in science
museums which makes your hair stand on end, each pulse from a CED with 400 volts
and extremely low current is equally harmless.
When discussing how electricity will affect the human body, voltage becomes irrelevant
without a discussion of the corresponding amount of electric current (measured in
amperes). To say 400 volts is dangerous is inaccurate without also talking about the
current associated with that charge. Voltage, even high voltage, alone does not harm or
kill.
The average current delivered by a TASER X26 device is 0.0021 amperes or 2.1 mill
amperes. Compare this with the average Christmas tree light bulb which has
approximately 1 ampere of current, or the 16 amps from a typical 110-volt wall socket; it
should be readily apparent that the extremely low current of a CED is safe.
To further put this into perspective, the “power plant” of a CED is two lithium batteries,
similar to those placed in most digital cameras, which can produce approximately
100,000 CED electrical pulses.
The TASER X26 outputs 0.07 joules of energy per pulse compared to a cardiac
defibrillator that operates at 360 joules per pulse.
The following is a schematic diagram of the technical components of a TASER X26:

Mir&'ox
Rodrih'uez, P.C.
~""""",Ac<>Do>n_

25

The dimensions of the X26 Taser (without cartridge) are as follows:
Length
Height
Width
Weight

6.00”
3.20”
1.30”
7.20 oz.

The length of the X26 Taser with cartridge is 7.25”.

Mir8t0x
Rodriguez, P.e.
Certified Public Accountants

26

DETAILED BACKGROUND AND AUDIT METHODOLOGY
AUDIT METHODOLOGY
To accomplish the objectives and scope of this HPD CED Program Performance Audit, MFR
formed a multidisciplinary team consisting of experts from the UH CPP, Rice University, SHSU
CCJ, University of San Francisco, and PFM.
The Audit Team composition and audit methodology for each team was as follows:
MFR Team
The MFR Audit Team consisted of MFR professional internal auditors and information
technology professionals. MFR focused on obtaining certain data that would be beneficial to
all of the groups such as policies and procedures, background data, HPD GOs, and then
facilitating data requests for the Audit Team.
MFR reviewed the internal controls related to the procurement and maintenance of HPD’s
CED inventory and related CED cartridges located at both the Academy and certain police
stations. As part of the review, MFR conducted test counts and recorded the serial numbers
of the actual CED and cartridges carried by a sample of 100 HPD officers. The test counts
and related serial numbers were compared to the records maintained at the Academy.
MFR conducted six Focus Groups with approximately 25 officers in each group who had
been issued a CED. The purpose of the Focus Groups was to obtain input directly from the
officers related to their experience(s) with CED operations and related administrative
matters.
UH CPP Team
The UH CPP Audit Team was comprised of experts from the University of Houston Center
for Public Policy, Rice University, University of San Francisco, and PFM. UH CPP
coordinated and led this group of experts in their quantitative analysis of the use of CEDs by
HPD. The UH CPP and Rice experts statistically analyzed the HPD data in the categories of
call for service/incident reports, injuries, complaints, and substitution of alternative
intermediate weapons for more lethal weapons. Data sources for the Scope period came
from the City Health and Safety Unit and HPD including the Crime Analysis Division,
Technology Services, Training Division, Payroll Office, and the Internal Affairs Division.
PFM created a user friendly graphical database that interfaces with the statistical data
analysis. The general public will be able to use the graphical database and perform their
own analysis.

Mir&'ox
Rodrih'uez, P.C.
~""""",Ac<>Do>n_

27

DETAILED BACKGROUND AND AUDIT METHODOLOGY
SHSU CCJ Team
The SHSU CCJ Team consisted of the criminology and certain mathematical research
experts from SHSU and the College of Criminal Justice. They used qualitative analysis and
observation to study several issues relating to the use of CEDs by HPD. SHSU CCJ
reviewed all of the call for service/incident and on-site supervisor reports for each
deployment of a CED for the period December 4, 2004 through June 30, 2007. The
following issues were reviewed:
•
•
•
•
•
•
•
•

When officers deployed a CED, were they responding to a call for service for which
they were dispatched or did the officer self-initiate the call by making a traffic stop or
otherwise encountering the subject?
What was the nature of the original call for service?
When an officer deployed the CED, how many cycles were used?
In what percentage of the cases was the subject under the influence of alcohol
and/or drugs when the officer deployed the CED?
What appeared to be the effectiveness of the deployment of the CED?
Were officers in compliance with HPD policy in the deployment of the CED and the
reporting of the call for service/incident? If not, were factors present which made the
deployment appropriate?
Were there cases in which the CED was deployed where the officer would have
been justified in using deadly force?
During the training of cadets on the use of the CED, what was observed that might
be affecting the manner in which officers will deploy the CED?

The SHSU CCJ Team also observed certain CED training classes at the Academy as well as
reviewed certain HPD lesson plans that pertained to CEDs.
The results and observations from the three audit teams are consolidated into this report.

Mir&'ox
Rodrih'uez, P.C.
~""""",Ac<>Do>n_

28

Mi

r8t0x

Rodriguez,

p.e.

Certified Public Accountants

PART II

PART II
OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS

MIR•FOX & RODRIGUEZ, P.C.
OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS
TABLE OF CONTENTS
Page
HPD CED Focus Groups

29

CED Inventory Count

50

Digital Power Magazine

52

Training Cartridges

52

Unrecorded Inventory

53

Scanner

53

Spark Test

54

Mirst°x
Rodriguez, P.C.
Certified Public Accountants

OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS
HPD CED FOCUS GROUPS
Background
MFR analyzed the CED deployments by officer, and noted that 803 officers have deployed their
CEDs during the Scope period as summarized below. Our analysis further indicated that 803
officers deployed their CED for a total of 1,417 deployments during the Scope period that
resulted in 1,284 service/incident reports summarized as follows:
Number of Times an
Officer Deployed a
CED
1
2
3
4
5
6
7
8
9
10
11
12
13
Total

Total Number of
Officers
492
156
90
26
23
7
1
4
2
0
0
1
1
803

Total Number of
Deployments
492
312
270
104
115
42
7
32
18
0
0
12
13
1,417

In the table above it should be noted that more than one officer may have deployed their
CED related to an individual service/incident report.
MFR formulated six groups of officers who had been issued Conducted Energy Devices
(CEDs) to discuss their CED related issues. Officer in this context is defined as a classified
employee of the following ranks: Officer, Senior Police Officer, Sergeant, and/or Lieutenant.
Each group had approximately 25 participants. Each Focus Group session lasted
approximately two hours.

29

Focus Group Composition
The preliminary results of the statistical analysis performed by the University of Houston
Center for Public Policy (UH CCP) Team indicated that the deployment of CEDs varied
between certain racial groups of officers. To gain a better understanding of the preliminary
statistical results, six groups were formulated as follows:
Focus Group One
Officers were judgmentally selected from a list of all officers who were issued a CED and
had never deployed their CED in the line of duty.
Focus Group Two
Officers were judgmentally selected from a list of African American officers who had
either deployed or not deployed their CED in the line of duty.
Focus Group Three
Officers were judgmentally selected from a list of Latino officers who had either deployed
or not deployed their CED in the line of duty.
Focus Group Four
Officers were judgmentally selected from a list of Anglo and Asian officers who had
either deployed or not deployed their CED in the line of duty.
Focus Group Five
Officers were judgmentally selected from a list of officers (all races) who had deployed
their CED less than five times in the line of duty.
Focus Group Six
Officers were judgmentally selected from a list of officers (all races) who had deployed
their CED more than five times in the line of duty.
The officers that MFR selected for each group were required by HPD to attend their
assigned focus group. MFR checked off their names on the selection list to ensure that the
officers who were present were the ones asked to attend the focus group. Several of the
officers expressed their appreciation for the opportunity to give their comments related to the
CED and did so freely as there were no representatives from HPD executive level
management attending the Focus Group sessions.

Mir~x
Rodriguez, P.C.
Certified Public Accountants

30

OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS
Discussion Outline
The discussion format for each of the six Focus Groups was the same and the MFR
discussion leader used the following outline:
•

•
•

•

Introduction - The discussion leader assured the officers that their identity would not
be associated with their comments. The MFR auditors who served as note takers
during each session were instructed not to record either the officer’s name or badge
number.
Objectives and Scope of the Conducted Energy Device Program Performance Audit
were presented to the officers.
Officers were told that the purpose of the Focus Group meeting was to have an open
discussion about CEDs to identify potential areas of improvement by HPD and/or the
City.
General discussion questions were as follows:
1.
2.
3.
4.
5.
6.

7.
8.

9.

10.
11.

Do you see the CED as an effective tool in securing compliance?
Do you use the CED as a substitute for other intermediate devices (e.g.
baton)?
Do you feel that the CED is a safe tool to use?
Would you like to see CEDs discontinued?
Have you used your CED and then followed the use of the CED by another
intermediate device(s) or your revolver?
CED Training:
a. Do you feel that you have had adequate training related to the CED?
b. Do you feel other officers have had adequate training related to the
devices?
c. What additional training do you feel is needed?
Would you recommend that HDP/the City purchase the camera attachment
for your CED? If so, why? If not, why not?
Our preliminary statistical results indicated that African American officers use
CEDs less than Latino and Anglo officers, both in general and for suspects of
different races/ethnicities including African American, Anglo, and Latino. In
your opinion and given your experiences, “What are the possible reasons for
this difference?”
Based on our preliminary results we found that African American suspects
are more likely to be shocked by a CED than Latino, Anglo, and Asian
suspects. In your opinion and given your experiences, “What are the
possible reasons for this difference?”
In light of the comments made here today, are there any HPD policies related
to CEDs that HPD should consider for issuance and/or revision?
Any other comment and/or recommendation related to CEDs that you would
like us to consider?

31

OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS
The discussion leader asked the eleven questions in order to generate discussion within the
group. MFR is pleased to report that each officer in each group made a contribution. Some
officers contributed more than others.
Observation
Summary – Focus Group Comments
There were numerous comments made by the officers to each of the discussion questions that
are detailed later in this section of the report. The general overall responses made by each of
the Focus Groups to each of the questions are summarized in the following table:

1.

2.

3.
4.
5.

6.

7.

*
**

Discussion*
Question
Do you see the CED as an
effective tool in securing
compliance?
Do you use the CED as a
substitute for other
intermediate devices (e.g.
baton)?
Do you feel that the CED is a
safe tool to use?
Would you like to see CEDs
discontinued?
Have you used your CED
and then followed the use of
the CED by other
intermediate devices or your
revolver?
Do you feel that you have
had adequate training related
to the CED?
Would you recommend that
HDP/the City purchase the
camera attachment for your
CED?

Group**
One
Yes
(83%)

Group**
Two
Yes
(60%)

Group**
Three
Yes
(75%)

Group**
Four
Yes
(100%)

Group**
Five
Yes
(71%)

Group**
Six
Yes
(100%)

No
(100%)

Yes
(71%)

Yes
(50%)

Yes
(75%)

Yes
(100%)

Yes
(100%)

Yes
(100%)
No
(100%)
No
(100%)

Yes
(70%)
No
(100%)
No
(100%)

Yes
(75%)
No
(100%)
No
(100%)

Yes
(100%)
No
(100%)
No
(100%)

Yes
(100%)
No
(80%)
No
(90%)

Yes
(100%)
No
(100%)
No
(90%)

No
(90%)

No
(90%)

No
(100%)

No
(70%)

Yes
(75%)

Yes
(100%)

No
(100%)

No
(100%)

No
(100%)

No
(100%)

No
(100%)

No
(100%)

Discussion questions are described in detail in the previous section Discussion Outline
As described in the Background Section of HPD CED Focus Groups.
•

Group One - Officers who have not deployed a CED

•

Group Two - African American Officers who may or may not have deployed their CEDs in the
line of duty

•

Group Three - Latino Officers who may or may not have deployed their CEDs in the line of
duty

•

Group Four - Anglo and/or Asian Officers who may or may not have deployed their CEDs in
the Line of duty

•

Group Five - Officers (all races) who deployed a CED less than five times

•

Group Six - Officers (all races) who have deployed a CED more than five times

Mir~x
Rodriguez, P.C.
Certified Public Accountants

32

Focus Groups General Responses, Continued
Discussion*
Question
8. Our preliminary
statistical results
indicate that
African
American
officers use
CEDs less than
Latino and Anglo
officers, both in
general and for
suspects of
different
races/ethnicities
including African
American,
Anglo, and
Latino. In your
opinion and
given your
experiences
what are the
possible reasons
for this
difference?
9. Based on our
preliminary
results we find
that African
American
suspects are
more likely to be
shocked by a
CED than
Latino, Anglo,
and Asian
suspects. In
your opinion and
given your
experiences
what are the
possible reasons
for this
difference?
*
**

Group**
One
African
American
officers
may diffuse
the
situation
verbally.

Group**
Two
African
American
suspects
have more
respect for
African
American
Officers,
hence the
CED is not
required as
often.

Group**
Three
African
American
officers may
diffuse the
situation
verbally.

Group**
Four
African
Americans
do not think
situations
are out of
control.

Group**
Five
African
American
officers
may
diffuse the
situation
verbally.

Group**
Six
African
American
officers
may not
have
many
violent
suspects.

African
Americans
tend to
make more
calls for
services.

African
American
officers are
familiar
with
dealing
with
diverse
ethnic
groups.

Calls for
services tend
to be higher in
the African
American
communities.

African
Americans
are
shocked
more
because
they
commit
most of the
crimes.

More
African
American
suspects
in high
crime
areas.

African
American
suspects
will fight
police
before
Anglo or
Latino
suspects
will.

Discussion questions are described in detail in the previous section Discussion Outline
As described in the Background Section of HPD CED Focus Groups.

33

Focus Groups General Responses, Continued
Discussion*
Question
10. In light of the
comments made
here today, are
there any HPD
policies related
to CEDs that
HPD should
consider for
issuance and/or
revision.
11. Any other
comment and/or
recommendation related to
CEDs that you
would like us to
consider?

*
**

Group**
One
Reduce the
amount of
internal
paperwork
related to
CEDs.

Group**
Two
Reduce the
amount of
internal
paperwork
related to
CEDs.

Group**
Three
Reduce the
amount of
internal
paperwork
related to
CEDs.

Group**
Four
Reduce the
amount of
internal
paperwork
related to
CEDs.

Group**
Five
Reduce the
amount of
internal
paperwork
related to
CEDs.

Group**
Six
Reduce the
amount of
internal
paperwork
related to
CEDs.

Officers
would like
to have the
option of
what they
can wear
on their tool
belt as
many of
them are
running out
of room for
the CED
and
cartridges.

Officers
would like
to have the
option of
what they
can wear
on their tool
belt as
many of
them are
running out
of room for
the CED
and
cartridges.

Officers
would like
to have the
option of
what they
can wear
on their tool
belt as
many of
them are
running out
of room for
the CED
and
cartridges.

Officers
would like
to have the
option of
what they
can wear
on their tool
belt as
many of
them are
running out
of room for
the CED
and
cartridges.

Officers
would like
to have the
option of
what they
can wear
on their tool
belt as
many of
them are
running out
of room for
the CED
and
cartridges.

Officers felt
that they
use the
CED less in
relation to
other
officers as
they arrest
more
suspects.

Discussion questions are described in detail in the previous section Discussion Outline
As described in the Background Section of HPD CED Focus Groups.

Mir~x
Rodriguez, P.C.
Certified Public Accountants

34

Officers
would like
to have the
option of
what they
can wear
on their tool
belt as
many of
them are
running out
of room for
the CED
and
cartridges.

Detailed Focus Group Comments
Specific noteworthy comments made by the officers during the Focus Group sessions in relation
to the discussion question are as follows:
Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
1. “Do you see the CED as an effective tool in securing compliance?”

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

“Absolutely, it saves lives.
It depends on the situation.
It’s just another tool to have.
It’s simply an option.
Good deterrent. It makes people think twice.
It depends on the individual officer.
The taser puts the suspects on the ground.
Yes, the suspect is not going to fight with you when you ask them to get on
the ground.
It looks worse when someone is beaten with a baton or flashlight than when a
suspect is tased.
Presentation of the taser often secures compliance.
Psychologically, it’s a threat when presented to the suspect.
There is a lot more compliance when the suspect sees the device.
Yes, if the person is on drugs and has beaten his wife.
It’s a great tool when you are using it to protect yourself.
It depends on the circumstances.
Sometimes it works on a person and sometimes it doesn’t. It depends on the
clothing and level of intoxication.
Generally, an effective tool.
It does nothing to help suspects on PCP (Phencyclidine is a hallucinogen
similar to LSD but can be much more dangerous).
It doesn’t work on drug abuse suspects.
It doesn’t work on subjects with mental disabilities.
It’s not the first step in securing control.
It’s ineffective when suspects are so intoxicated that it seems like they are on
drugs.
Sometimes, the device has been known to fail to deploy.
It depends on what you are dealing with at the scene.
All the policies make the tool ineffective.
The paperwork makes you think twice about using it.”

35

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
2. “Do you use the CED as a substitute for other intermediate devices (e.g.
baton)?”

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Mir~x
Rodriguez, P.C.
Certified Public Accountants

“Yes.
The consequences are different for using it.
Training and policies are different.
Tenure plays a big part.
It depends on situation.
It depends on different factors (station you train with, shift, etc.)
There is confidence that the taser will work.
Demographical tenure determines the use of the taser.
Once I give you my command and you don’t comply, I’m gonna help you. I
have never discharged my weapon.
It’s a good tool.
It is a more advanced weapon and it is effective.
It’s better than anything else we have.
We use it when we’re in fear of getting hurt. When we’re being attacked by a
heavy-set person.
Absolutely. Before the taser, we used the baton or mace. These tools were
ineffective because you had to get close. If you weren’t close, you would end
up spraying other people with mace.
It’s better than shooting someone and is less harmful.
Yes, we are using it more because we are being trained more.
It’s more effective than the pistol.
The CED is more effective because pepper-spray gets into officers’ eyes
even after getting in the patrol car.
It’s a great tool. It should not be discontinued.
If he (the suspect) has something in his hand, I am going to pull my gun.
Forty years ago, we fought because we weren’t trained with tasers. There
are so many elements that you have to decide before you use it.
That’s how we react because of what could happen. We don’t think about
policies.
If a person is aggressive, we’re going to stop him. No, no, we gave you a
command. We don’t worry about the consequences.
In the old days, we would apprehend a person and write the letter tomorrow.
We already know what we need to do before we get out of the car. It’s going
to be my way.”
An additional comment by most officers that had not deployed a CED, was
that, “They would use it.”

36

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
2.

Officer’s responses continued:

•

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

“The department gave us this BS training. We go for what we know. If that’s
what you know, then you do what you know. I fight with these (puts up fists).
If I happen to look at my belt, I realize, Oh, I got this thing right here. Let me
use this taser. For years, we’ve done it this way (puts up fists)
Hesitant to use CED because of the way the policies are written.
Every time I come within pistol range (i.e. of the suspect) everything is fine.
We played with it in roll call, but we found out it didn’t click like it was
supposed to sometimes.
There’s not enough time to figure out which one to use.
The younger guys are more apt to use it.
The taser is not reliable.
It’s not always the safest tool. Sometimes when suspects are tased, they pull
darts out and officer has to threaten to shoot him/her.
No, because face plates (i.e. blast doors) fall off.
Reports that have to be written after deploying tasers are time-consuming so,
I would rather not use it.
I have used it and it didn’t work. So, I ended up shooting the suspect. I do
not trust the taser because it is not dependable.
Does not work when suspect is under the influence of drugs. For example, a
guy that was on drugs was tased 14 times until he went down. The next day
he was sent to jail and he was rejected back to the hospital.
I have not used it. The pepper spray works fine.
I would prefer hands on first rather than use CED.
I’ve seen it work and not work.
Sometimes you don’t have time to pull out the taser. The other weapons are
faster to use.
I have to reconsider using the CED because policies continue to change.
There are too many to keep up with and there is too much red tape involved
with it.
It’s not a replacement for the baton.
I can’t carry other intermediate devices (i.e. baton).
The situation must be almost perfect to use it. Now, we can’t use it on the
people that run.
In the drop program, we don’t care about the policies and procedures. It’s a
scare tactic with the younger people, We’re going to fire you, if you screw up.
We have not gotten anymore training than, unsnap, here it is, and that’s it.
It is the last weapon I would use because of the way they make us carry it.
It’s not accessible.”

37

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
3.

“Do you feel that the CED is a safe tool to use?”

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Mir~x
Rodriguez, P.C.
Certified Public Accountants

“Yes, it’s the safest tool we could have gotten if it works.
It’s not a fix all, but a great tool.
There are no lasting health repercussions.
It’s safe if it works on the individual that it’s used on.
Yes, if you already have it in your hand and it works. It is what it is. It’s a
safe tool sometimes if it works.
It’s safe based on the info you get from the department and the media.
Death is usually a result of a pre-existing condition.
It’s safer than the baton and the pepper spray.
The paperwork can take you up to 3 hours. It’s a safe tool ultimately, but the
public is not as aware as they should be.
There should be continued training and open dialogue on this device.
The protocol for having the taser on the belt is a bad decision and makes it
unsafe for officers.
There are some problems with the CED.
A gun is a safe tool; it depends on who uses it. There are a lot of problems
with the taser, especially, the cartridge.
The perception the public gets is not clear when the press gets, ‘No
comment’ responses from the department.
Has it been proven to cause death?
The public’s ignorance makes it seem like an unsafe tool.
The department doesn’t back up the officers in the media.”

38

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
4.

“Would you like to see CEDs discontinued?”

•
•
•
•
•
•
•
•
•
•
•
•
•
•

•
•
•
•
•
•
•
•
•
•

“No.
No, they should not get rid of it. We just weren’t adequately trained for this
tool.
I am 185 lbs and I don’t wanna have to fight him (as he points to a very large
officer). He can call me old, but I’m using the taser.
I think about the taser after the arrest. I should have tased that suspect. I
should have put that heat on his behind.
I would not like to see them discontinue the CED. However, I would like to
have a choice to carry it or not.
I would rather carry the taser. If other officers prefer other tools, they should
be able to carry what ever they feel comfortable with.
Keep them if officer is able to use it at their discretion/judgment. They don’t
have to watch us like children and tell us when to use the CED.
No, it’s a great tool. However it has administrative problems. It has saved
suspects’ lives.
Yes, take them away because there are too many restrictions for using it.
Are they going to take our guns away?
It’s not going to happen (i.e. take away the taser) no matter what we say.
I would like an option to carry it or not.
The department is putting too much into figuring out which tool to use.
The decision to use the CED has to be made in less than one second. There
are so many restrictions for using them that you need too much time to think
about whether to use it or not. There are too many rules. The policies make
it difficult to use.
Too much red tape makes it hard to use the CED.
It has too many connotations. A political game is played when the taser is
used.
It breaks more than it works.
Instructions in the beginning were to shoot suspects on passive resistance.
The department does not trust us enough to decide when to use the tool.
I would rather use the baton.
The department doesn’t trust the officers to use it wisely.
The tasers draw media.
The department needs to stand by what they told the officers in the
beginning.
It requires more paperwork for the supervisors.”

39

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
5.

“Have you used your CED and then followed the use of the CED by other
intermediate devices or your revolver?”
Officers’ responses were as follows:

•
•
•
•
•
•
•
•
6.

“Yes, because the CED did not work as intended. A homeless individual had
a heavy jacket and darts did not penetrate skin. Therefore, it did not work.
During a struggle an officer stunned himself and had to revert to hands-on
approach.
I had to use my hands after using the taser.
I never witnessed a CED that did not work.
Tenure makes a difference.
We use other tools because of habit or initial training.
People don’t want to deal with the effects of the taser. The suspects see it as
a threatening tool. Especially, when they see the laser dot on them.
If someone came up to another officer and pulled a trigger on him, you don’t
have that kind of time to think and respond.”

a. “Do you feel that you have had adequate training related to the CED?”
Officers’ responses were as follows:

•
•
•
•
•
•
•
•
•
•
•
•
•

Mir~x
Rodriguez, P.C.
Certified Public Accountants

“Are they training cadets to use the tool in mock crime scenes?
You have to be tased to carry the taser in smaller departments.
The new officers have had more training.
Annual training is not enough. However, I understand it is costly.
More training will make me more comfortable with working with the CED.
You can’t adequately train someone with the CED until you are in the actual
situation.
The officers are reluctant to use the CED when it doesn’t work. Hence, more
training is needed.
Not enough training may cause uncertainty on when and how to use the
CED.
Not enough training because we have too many accidental uses.
You need to have a heart and no fear.
You need to understand what causes a suspect to act a certain way.
Judgment is important.
How about consistent training?”

40

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
b. “Do you feel other officers have had adequate training related to the
device?”
Many of the officers responded to this question by saying that “The training
had not been consistent.”
c. “What additional training do you feel is needed?”
Officers’ responses were as follows:

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

“Consistent training is needed (instructions are not clear - they change on a
continuous basis).
There are inconsistent policies about removing darts. Judgment calls have to
be made continuously.
May need training to recognize signs of PCP.
CIT-Crisis Intervention Training. Forty hours are required.
Encourage media to go through training and get tased.
The younger officers are relying too much on the tasers. If it doesn’t work
they’re not sure what to do next.
Most female officers can talk a suspect down more often than male officers.
What more can be taught?
The taser can kill you too if you leave it on them for 45 seconds.
There is nothing in place and no adequate training.
What scares me about the training is that we may have to be shocked as part
of the training. Some departments have to do that and I don’t want to have to
do that.
The Fondren location structure was suggested for all stations. It has an
officer on each shift trained to use a CED.
During roll call it takes 10 minutes to update and review the CED.
We need more training on how to write more articulate reports.
The people at the top need to sit through some training or ride on a patrol.
In policy, it is stated that medical condition of suspect should be known.
The Sgt. or EMT can remove darts if it’s above the neck or in the groin.”

41

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
7.

“Would you recommend that HDP/the City purchase the camera attachment for
your CED? If so, why? If not, why not?”
Officers’ responses were as follows:

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Mir~x
Rodriguez, P.C.
Certified Public Accountants

“I don’t like it. The camera on the CED will not capture everything.
The camera only catches a portion of situation.
It will be another item that can be broken. It’s too costly.
It will be extra weight on our belts.
It would increase second guessing for using the CED.
No, in that case put a camera on our shoulder to see everything that happens
and not just what happens after the darts are deployed.
I would see this as a problem because it would only record when taser is
used and not what happened before it was used.
The people viewing tape will not judge situation correctly.
The cameras may be used against the officer.
No, because the camera would only be good to record last seconds of
situation. It will not capture entire situation.
The vulgar language would be captured. We are forced to use it because
that is the only way the suspects comply.
It’s not a good idea and it’s not necessary.
The camera captures piecemeal situations.
Adding the camera may decrease the use of the CED by officers.
You’re going to be guilty with the media no matter what.
The camera results will be on MySpace and YouTube.
When you use the taser camera, Channel 13 is going to show a clip of me
saying, ‘Listen *#*#*#*#’. Then they’re going to look at me and ask, ‘Now,
why did you tase him?’
They are too bulky.
They are designed to get the officer in trouble.
Policing in the City of Houston is not what you see on Law and Order. It is
not clean.
What is the camera going to do for the officers?
Police work is not clean. It’s violent, real, and not what’s on television.”

42

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
8.

“Our preliminary statistical results indicate that African American officers use
CEDs less than Latino and Anglo officers, both in general and for suspects of
different races/ethnicities including African American, Anglo, and Latino. In
your opinion and given your experiences, ‘What are the possible reasons for
this difference?’”
Officers’ responses were as follows:

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

“Most African Americans officers grew up fighting and wrestling. If you can’t
talk them down, you fight them.
It’s in the genes of the African American officers.
African American officers use the gift of gab.
Some officers are looking for a fight.
Look at the seniority of the officers that are using the CED’s.
It depends on the area of the City that you work in.
You have to be able to read people. I can talk to this guy or I need to fight
this guy. It comes with experience, you can’t teach this in the Academy.
You have to take control when you step on the scene.
The veterans are not training the new officers any longer. The veterans
taught the new officers when to use force.
Field training officers have tenure of two years which is not enough
experience to train new officers.
Tenure and shift are major factors. More CEDs are used on the night shift.
Cultural awareness in Academy doesn’t help with what you experience on the
street.
Tenure plays a part.
African American suspects are less likely to be compliant to a Anglo officer.
The style of African Americans officers. (communication, respect).
The rearing techniques of African Americans.
African American officers’ approach is different.
The CED may be used more in high crime rate areas.
Perhaps, they don’t feel like situation is out of control and don’t think it is
necessary to use the CED.
Perhaps, African Americans are more compassionate towards some people.
African American officers know the people where they patrol and don’t want
to use force against suspects.
Everyone just wants to get a paycheck and go home safely.
Perhaps, people calling police are calmer with the African American police
officers attending to the call.
Sometimes ethnicity affects the way the suspect responds. Sometimes
suspects will only respond to someone of their own ethnicity.
Officers working in higher crime rate areas will have more incidents.
More senior officers may be African American. Seniors are probably less
likely to use the CED.”

43

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
8.

Officer’s responses continued:

•
•
•
•
•
9.

“Sometimes people don’t want to speak to an African American officer.
A lot of African American people don’t like police, so they instill that mentality
into the younger generation. As soon as a Anglo officer gets out of the car, it
becomes a problem.
African American suspects can relate to African American officers better.
Different life experiences can play a role. Some African American officers
can talk down suspects.
It depends on the area you are patrolling in the City.”

“Based on our preliminary results we find that African American suspects are
more likely to be shocked by a CED than Latino, Anglo, and Asian suspects. In
your opinion and given your experiences, ‘What are the possible reasons for
this difference?”’
Officers’ responses were as follows:

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Mir~x
Rodriguez, P.C.
Certified Public Accountants

“The calls for service tend to be a lot higher in the African American
community.
The suspects have no future plans (i.e. they live for the moment).
Racist officers are doing the tasing.
New officers are trained a certain way by officers.
African American officers are used to dealing with diverse groups.
Stereotypes can play a role.
It can be based on fear and lack of knowledge.
When a person gets in trouble they bring up race issue and the race card is
thrown.
Whoever is committing the most crimes is being tased the most.
The suspects of African American ethnicity do not want to look bad in front of
their peers.
Cultural issues can play a role.
The crime rate is higher in the African American race.
The African American culture is more aggressive. The suspects have more
attitude and are more combative.
It is directly in line with who is committing most crimes.
The calls for service may contribute. African Americans are more likely to call
police.
Latinos may not call because of legal status.
Look at the crime rate amongst ethnicities.
The size of African American suspects can contribute to the amount.
If a Anglo officer works in an African American neighborhood, he or she is
going to taser more African American people.”

44

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
10.

“In light of the comments made here today, are there any HPD policies related
to CEDs that HPD should consider for issuance and/or revision?”
Officers’ responses were as follows:

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

•
•
•
•

“It was a good tool when it came out. However, policies make it hard to use.
Improve it or get rid of it.
Why does there have to be some aggressive force before using the CED?
You do not have to articulate some other form of force.
It’s not about what happens, but how it’s written up.
Assistance is needed for officers to write better reports.
No cameras for tasers.
Tasers are tested every morning to make sure they are operating correctly.
How do you classify taser on the call slip? (When to use it or not to use it)
If senior officers had partners, there probably wouldn’t be a great need for
tasers.
The department may not be thinking safety with the use of tasers.
How many more officers could we have if they use the money from the
purchase of tasers, and taser training, etc?
It’s funny that we have to test it every morning during roll call.
Testing it says how reliable they are. There is battery failure and holster snap
is weak. How many more options are we going to use? Why y’all tell us we
can’t have partners anymore?
Some things could be condensed. There is too much paperwork.
Different opinions form confusion about the situation. Too many reports have
to be written. Sometimes, supervisors write reports when they have not even
witnessed what occurred.
You have to articulate why you do what you do. It’s too complicated to use
the CED.
We should go back to if they are not compliant you should be able to tase
them.
We need to mirror the state law.
Liability is an issue that you have to worry about when using the CED.
IAD (Internal Affairs Division) complaints.
The City should implement the same amount of GOs for the CED as any
other compliance tool the officers have available.
The CED is the only situation where the supervisor has to write a
supplemental report. However, no supervisor supplement report is written for
any other compliance weapon. We recommend the elimination of the
supplement report.
Eliminate the incident letter for the taser.
Keep the CED in uniform with other weapons similar to it.
The CED was media written.
Paperwork should not be required for the accidental use of the CED.”

45

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
10.

Officer’s responses continued:

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Mir~x
Rodriguez, P.C.
Certified Public Accountants

“The policy needs to be revised. There is too much paperwork involved for
the CED.
The supplement report is beneficial because it will help the officer. The
supervisor gives an opinion.
IAD (Internal Affairs) complaint - two reports will support situation better than
one.
There are too many restrictions. There should be restrictions, but tasers
should be allowed to be used for people not complying.
Revise the restrictions and broaden the use of the CED.
The City would rather pay for broken windows on an officer’s car rather than
pay for a lawsuit. Do not suggest allowing officers to use taser to make
suspects comply even when they are in the car.
Dry stunning is not as effective as using the darts. There is a pain issue.
If tasers are kept, change the rules for using them.
They want us to go out there and see a neat situation.
There is too much controversy when tasers are used.
We do not do a good job of writing reports.
The policy has to be consistent and the training needs improvement.
The tasers break too easily.
There is not much faith in the device. Officers would rather jump on the
suspect because when they tried to use the taser the first time it didn’t work.
The frequency of failure is getting higher and causing more problems.
When one officer messes up, the policy should not change for everyone
across the board.
Management should come out on the street to see how it happens.
You got over worked officers in Houston. There are not enough officers on
the street.
The amount of money you put in the tasers could be the number of officers
back on the streets. Partners working on the street.”

46

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
11.

“Any other comment and/or recommendation related to CEDs that you would
like us to consider?”
Officers’ responses were as follows:

•
•
•
•
•
•
•
•

•
•
•
•
•
•
•
•
•
•
•

“We would like to see the CED issued to investigators.
Officers don’t wear the CED because it falls off easily and you have to write a
report.
There is an administrative letter you have to write every time the CED goes
off.
The tasers are under a lot of scrutiny.
Can you increase the voltage in the tasers?
A lot of people base things on what makes sense. A lot of times things don’t
make sense. I am a fifty year old man. This is how I do it. We deal with a
situation with a little more patience and so do the Latino officers.
We go through all these cultural diversity classes. If I get an African
American woman with a hand on her hip and shaking a finger in my face, I
won’t see it as a threat. I deal with that at home.
What comes out of this is not going to matter. They already made their minds
up. They could take a vote of no by 80% that we don’t want to carry it any
more. We are going to carry it. When it goes back up to Big Daddy we are
still going to carry it like it or not. We were ordered to attend this. For some
of us, it’s our day off or just getting off work. It’s her day off today. We were
ordered. If you came and asked who would come to this you will be talking to
yourself.
The holster needs to be improved. Is the City of Houston considering other
providers for them?
Manpower is an issue when using tasers.
Although it should be used as an intermediate device some of us have run
out of room on our belts for it.
I will go to my grave before I use it. I will use the baton or flashlight.
It gives an officer a false sense of security. With the CED as opposed to with
other weapons (baton, pepper-spray) you have to ask yourself, ‘What are you
doing?’
When you taser suspects you have the possibility of getting chastised by the
public. This makes you less likely to use the CED.
Scrutiny from the department is the biggest problem when using it.
They should have maintenance on the CED so it will function properly.
They need to hire more qualified policemen. The shortage of staff puts an
officer in a more likely position to use the CED.
The squad system is spread thin. When back up is called it takes too long to
respond.
When I call for back up on the night shift all I hear is crickets.”

47

Discussion leader asked the officers the following question(s) and the detailed
response(s) that were noted are listed below.
11.

Officer’s responses continued:

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Mir~x
Rodriguez, P.C.
Certified Public Accountants

“Language barrier is a factor (Asians & Latinos).
Gender comes into effect. Women may be more likely to use a taser on a
male. The size of the suspect matters.
Male officers are less likely to use it on female. However, if necessary it will
be used.
Dogs are being tased to get to suspects.
The department does a horrible job of explaining things.
The communication between levels of tiers (i.e. within HPD) does not exist.
The top tier cannot communicate with lower tier. The lowest tier cannot
communicate with highest tier.
Trigger happy officers mess it up for everyone.
More policies are written because of the bad officers.
When officers speak out, they are categorized as disgruntled and prone to
get fired.
Results without risks are expected.
HPD management feels like the CED results are irrelevant.
Lower rank’s opinions are taken, but not used.
It’s the Chief’s idea to have tasers in the first place.
They tell us to use it, but when we use it they question us about it.
If an officer speaks out he/she is fired.
This study means nothing. Nothing will change.
This discussion is not going anywhere. No one is going to do anything about
it.
The department just wants to be able to say they listened to their officers and
they performed a study.
The media has helped get the word out about the taser.
You have to have a perfect situation to use it on a suspect that is 21 feet
away.
Most female officers can talk a suspect down more often than male officers.
They gave us the taser, so we wouldn’t accidentally use our gun.
They said it was a toy and would help keep us from fighting. Then they came
up with policies and procedures. I want to give it back to them at this point.
The upper management needs to take calls.”

48

MFR noted that certain observations and comments kept reoccurring during the meetings and
they are summarized as follows:

•
•
•

•
•
•
•
•
•
•

Most officers in the sessions complained about the paperwork requirements
for CEDs.
Officers wanted more training in report writing.
Several officers want carrying the CED to be optional like the other
intermediate weapons. Many of the officers commented on the size of their
tool belt and how it affected their ability to enter and exit their patrol car as
well as the tool belt weighed them down during a foot pursuit.
Officers really appreciated the opportunity to be heard.
Several officers thought that the CED was effective and is becoming well
known because they have had suspects plead with them once the suspect
sees the laser dot from the CED on their body.
HPD management was perceived as a top down style and officers believe
that executive management did not adequately support them. Certain
officers appeared to have a lack of trust for executive management.
Most officers seemed to have little to no racial bias and certain officers
questioned why MFR had segregated Focus Groups.
Most officers wanted to retain the CED as an intermediate weapon; however,
HPD needs to reduce the paperwork requirements to that of other
intermediate weapons.
Several officers appeared to need clarification on the applicable GO’s.
Officers wanted the CED rules changed so they could deploy a CED on a
suspect who was fleeing. Currently, several of the officers believed they
could not use a CED on a fleeing suspect.

49

AUDIT DETAILS, OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS
CED INVENTORY COUNT
Background
The inventory records for the CEDs and the cartridges were maintained at the Academy.
Each police station is required to maintain a record of the replacement cartridges issued to
their officers from the station’s cartridge inventory.
GO # 400-26 - “Conducted Energy Devices” Section 6 requires that:
“Officers trained in the use of and issued a CED will wear it all times while wearing the
official uniform, even while working extra employment”. Furthermore “Officers may also
carry an approved pouch for carrying a spare CED cartridge (in lieu of attaching the
cartridge to the handle of the CED).”
In summary the CEDs were issued to each patrol officer with two cartridges and they should
be carrying one CED and two cartridges.
Observation
MFR selected a judgmental sample of 100 officers who were issued a CED to verify that the
serial number of their CED and the serial numbers of the CED cartridges that were assigned
to them by HPD were accurately recorded in the respective HPD inventory records. The
officers were selected from six roll calls at four police stations, Westside, Southeast, North,
and Northwest and certain officers were selected at the Academy.
MFR counted a total of 100 CEDs and 173 cartridges. Thus 26 of the 100 officers did not
retrieve their second cartridge for the Audit Team to record the serial number and one officer
had both cartridges missing. The officers gave the following reasons:

•
•
•
•
•
•
•
•
•

“Cartridge is in the trunk of my vehicle
Cartridge is at home
Blast doors fell off the cartridge
I lost my second cartridge while chasing a suspect
Cartridge will not stay on the handle of the CED
I have no room on my tool belt
Have not had time to do the paperwork
Did not want to be reprimanded for loss of City property
The officer in charge of the cartridges at our station only works the day shift”

In addition to the 26 officers noted above four officers went to their vehicle to get their
second cartridge so that the Audit Team could record the serial number.

Mir~x
Rodriguez, P.C.
Certified Public Accountants

50

AUDIT DETAILS, OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS
For the sample of 100 officers, MFR traced the serial numbers from all 100 CEDs to the
inventory records at the Academy, no exceptions were noted. Based on the results of the
CED testing, the risk for unaccounted CEDs is very low.
For the sample of 100 officers, MFR traced the serial numbers for only 110 of the 173
cartridges to the cartridge inventory records. The officer at the Academy had been on sick
leave and had a backlog of CED replacement cartridge serial numbers that had to be input
into the computerized cartridge inventory. MFR has provided the officer at the Academy the
serial numbers of the remaining 63 cartridges to trace into the inventory records, once he
has completed inputting the backlog of replacement cartridge serial numbers. MFR also
noted that all serial numbers of cartridges purchased by HPD were recorded on the
manufacturer’s shipping documentation that was received and maintained by the Academy.
Even though HPD has adequate physical safeguards over CED inventories and HPD is
aware of all of the CED cartridges that are in its possession, there is a risk that HPD may not
be able to determine to which officer a certain CED cartridge had been assigned.
Recommendation
MFR recommends that HPD review their policy related to replacement CED cartridges as
the officers have such a reluctance in carrying their extra cartridge primarily because of their
fear of losing and/or damaging it as well as the related paperwork. Additionally, HPD needs
to consider a more efficient record keeping system to reduce the risk of losing accountability
for cartridges.

51

AUDIT DETAILS, OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS
DIGITAL POWER MAGAZINE
Observation
A Digital Power Magazine (DPM) or commonly known as a battery, is an integrated power
magazine containing lithium battery cells and a solid state memory module for tracking the
CEDs power consumption. The memory module also has the date, time, power level
remaining, and contains the latest Manufacturer’s software update for the CED.
According to GO # 400-26, Section 12 CED or DPM Replacement, the officer is required to
write a letter to be reviewed and approved by the supervisor who in turn drafts a letter to the
commander. Upon approval, the officer will take copies of the approvals to the Taser Office
at the Academy for a replacement DPM. We were told that the DPMs were failing more
frequently now that certain CEDs were over 3 years old. Since the frequency of failure has
increased, DPMs are being replaced without the authorization letters in order to reduce the
officer’s time away from work.
Recommendation
We recommend that HPD review and assess the DPM failures and the amount of time
officers are spending to get them replaced. Consideration should be given to a more cost
effective and efficient replacement process.

TRAINING CARTRIDGES
Observation
The boxes of cartridges issued to the Academy Defensive Tactics Office for training
purposes had the same capacity as the ones issued to the officers and police stations.
During the audit we noted that the records did not indicate to whom the cartridges were
given during training.
Recommendation
To reduce the risk of inaccurate reporting of issuance of cartridges, MFR recommends that
HPD implement a process to improve the controls over training cartridge inventory.
Corrective Action
HPD has revised the training cartridge process. Process improvements include, the CED
training cartridges are now a blue color and can be easily identified by HPD.

Mir~x
Rodriguez, P.C.
Certified Public Accountants

52

UNRECORDED INVENTORY
Observation
The original cartridges that were received during the initial shipment of CEDs were not
recorded in HPD’s inventory records at the Academy’s Taser Office. The officer in charge of
the CED records told the Audit Team that the initial batches totaling approximately 1,300
cartridges were issued very quickly to the officers which did not give the Taser Office
adequate time to develop an inventory process. Subsequent to the initial issuance of the
cartridges, a process has been developed to record all cartridges.
MFR noted that there were six CEDs in the Defensive Tactics Office of the Training Division
that were not recorded in the CED inventory records. The Training Division told the Audit
Team that these CEDs were received from the manufacturer without charge and as a result
were not recorded in the inventory.
Recommendation
MFR recommends that upon receipt from the manufacturer, all the CEDs and cartridges that
are assigned to an officer be recorded in HPD’s CED inventory tracking system at the Taser
Office.
Corrective Action
HPD has recorded the six CEDs in the Defensive Tactics Office of the Training Division in
the HPD CED inventory tracking system. The Taser Office is continuing to work on
recording the 1,300 cartridges.

SCANNER
Observation
All CEDs, cartridges, and DPMs have a serial number and a corresponding bar code.
During the audit, MFR noted that the bar code scanner was more accurate and faster to
record the serial numbers into the HPD CED tracking system; however, during the audit, the
bar code scanner was temporarily out of order and has since been repaired. MFR did not
note any bar code scanner(s) or access to the Taser Office CED tracking system in the
police stations visited during the audit. The police stations were recording the serial
numbers manually. The manual numbers were then manually entered into the CED tracking
system at the Taser Office at the Academy.
Recommendation
To improve the system of internal control, MFR recommends that HPD consider installing
bar code scanners in all of the police stations to facilitate the recording and issuance of the
CEDs, cartridges, and DPMs. Furthermore the scanners could be part of a centralized HPD
CED tracking system that would contain the serial numbers of the CED and cartridges
assigned to each officer.

53

AUDIT DETAILS, OBSERVATIONS, RECOMMENDATIONS,
AND CORRECTIVE ACTIONS
SPARK TEST
Background
GO # 400-26 - “Conducted Energy Devices” Section 6 “Spark Test” requires that,
“Roll call supervisors will document witnessing spark tests on the roll call sheet and
ensure all officers on the sheet conducted a spark test.”
Observation
MFR noted that while attending roll calls during the CED and cartridge inventory test counts
that the roll call supervisors were not consistently recording the witnessing of the spark test.
Certain roll call supervisors had the officers record the Spark Test on their job card.
If the spark test is not performed regularly, the CED may fail to properly deploy the probes.
Recommendation
To ensure that the GO # 400-26 is being followed, MFR recommends that HPD review and
revise its policy to have a periodic internal review process to ensure that roll call supervisors
are adhering to GO # 400-26. The periodic review should be documented to evidence that
the Spark Test had been performed.

Mir~x
Rodriguez, P.C.
Certified Public Accountants

54

UNIVERSITY OF HOUSTON

(en:~~PUBlICPOlICY

PART III
UNIVERSITY OF HOUSTON
CENTER FOR PUBLIC POLICY
PART III

A STATISTICAL ANALYSIS OF THE
USE OF CONDUCTED ENERGY DEVICES BY
THE HOUSTON POLICE DEPARTMENT

UNIVERSITY OF HOUSTON CENTER FOR PUBLIC POLICY
A STATISTICAL ANALYSIS OF THE
USE OF CONDUCTED ENERGY DEVICES BY
THE HOUSTON POLICE DEPARTMENT
TABLE OF CONTENTS
Page
Summary

55

Background

59

Study Components

63

Data

67

Methods of Analysis

77

Results

79

Data Management

107

Conclusion

109

References

110

UMIYlt5J1Y Of I{)USJOII

''".::PUBlICPOLICY

SUMMARY
Using statistical analysis and data visualization/geo-spatial tools1, the research group
assembled by the University of Houston Center for Public Policy2 (UH CPP) studied the
following questions pertaining to the use of conducted energy devices (CEDs) within the
Houston Police Department (HPD):
•

Incidence: Who is subject to being shocked by a CED? What are the demographic
characteristics of suspects and officers in these events? Where have these incidents3
occurred?

•

Injuries: Have the number of injuries to officers and suspects been affected by the CED
policy?

•

Substitution: Are CEDs used as substitutes for alternative intermediate weapons or
lethal weapons?

•

Complaints: How many complaints have been filed for CED use? What are the
demographic characteristics of the complainant and the officer(s)? How many
complaints have been validated?

The results in this analysis are subject to data limitations; however, the available data are
sufficient for this exploratory analysis. The short duration of time (the period reviewed)
combined with an overall small number of incident reports (less than 1% of the 1.4 million cases
recorded during this period) disallowed strong causal interpretations. In future statistical
analyses, new control variables and the natural extension of the time period for investigation
can assist in providing greater certainty in answering the questions above.
Throughout this analysis we will note where data and design limitations limit the overall certainty
of our conclusions.

1

The use of visualization tools and “mapping” to find patterns and relations in quantitative data has a long history.
th
Among the more famous examples is Dr. John Snow’s investigation of a cholera epidemic in 19 century England
(see Tufte 2001). We thank Governor Bill Hobby for bringing this information to our attention.
2

The members of the research team that contributed to this analysis include Renée Cross (University of Houston
Center for Public Policy), Tom Duncavage (Prototype, Fusion & Modeling, LLC), Jim Granato (University of Houston
Center for Public Policy), Mark Jones (Rice University), Terry Mayes (Prototype, Fusion & Modeling, LLC), Bill Reed
(Rice University), Matt Soltis (Prototype, Fusion & Modeling, LLC), and M.C. Sunny Wong (University of San
Francisco). Stephanie Eguia (University of Houston Center for Public Policy) provided research assistance.
3

The term “incidents” refers to calls for service.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

55

Incidence Results
We noted that the incidence results must be viewed with caution. One challenge was the
lack of adequate data on suspect and officer characteristics. This affected the confidence
we put in the results of our suspect data analysis. On the other hand, the data utilized for
the officer data analysis allows us to draw inferences with much greater confidence. Yet
another complication was in the Council District analysis. While the Council District
analysis allowed us to control for important contextual factors, it was crucial to
remember that the number of CED events in most of the nine Council Districts was
sufficiently small so as to warrant caution in our interpretation.
For the period December 2004 to June 2007, the principal statistical and geo-spatial results
were as follows:
•
•

•
•
•
•
•
•

•
•
•
•
•

Of the 1.4 million incident reports, there were 1,284 (.08%) events where the CED
was deployed. This equates to 8 CED deployments for every 10,000 incidents.
There were approximately 700,000 incidents in the data where primary suspects
could be matched to an HPD officer (who could have deployed a CED). Of those
700,000 incidents, 1,030 involved the use of CEDs. This translated into
approximately a .14% likelihood of having a CED involved in an incident.
Alternatively for every 10,000 incidents, 14 involved the use of a CED.
Among suspects, African Americans had the greatest probability of having a CED
used on them. Latinos and Anglos followed in overall likelihood.
Among officers, there were no gender differences in the overall likelihood of
employing a CED.
Among suspects, males were more likely to be involved in a CED incident than
females.
African American officers were least likely to deploy a CED. Latino and Anglo
officers followed in overall likelihood, with both equally likely to use a CED.
African American officers had an equal likelihood of using a CED on African
American, Latino, and Anglo suspects.
Latino and Anglo officers had a much greater likelihood of using a CED on an African
American suspect than on Anglo or Latino suspects. Latino and Anglo officers were
equally likely to use a CED on an Anglo suspect. Latino officers had a greater
likelihood of using a CED on a Latino suspect than Anglo officers.
When looking at CED use within City Council Districts in Houston: Districts D and H
have the highest likelihood of CED deployment.
African American officers were just as likely to use a CED as were their Anglo
counterparts in Council District D.
Council Districts A, C, E, F, G, and I are similar in CED deployment.
Council District B has a greater likelihood of CED use than Council District F and G.
The relationships between officers and suspects disappear or change when
Council Districts are used as a statistical control.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

56

Injury Results
Due to the relatively short time period when the CED policy was in place (at the time
of this analysis) the findings presented here need to be interpreted with caution. With
the passage of time, it will be possible to find effects related to the CED policy.
For the period January 2000 to June 2007, the principal statistical results were as follows:
•
•
•
•
•

Injury indicators, in general, indicated incidence shifts (also known as structural
breaks) prior to the CED policy being instituted.
The estimated total number of workers’ compensation claims by the officers has
fallen by an accumulated 20% that began in June 2004.
The estimated level of monthly expenditures on claims shows an accumulated
reduction of approximately $50,000 per month (50%). This began in May 2003.
Both decreases began prior to the incorporation of the CED program at HPD.
The decline in the injury indicators has continued during the Scope period.

Substitution Results
As with the injury analysis, the substitution analysis and results presented here covered a
relatively short time period when the CED policy was in place. Again, the findings
presented here need to be interpreted with caution. Over time, it will be possible to find
affects related to the CED policy.
In addition, due to lack of available data, the results in this section did not cover intermediate
weapons so the test for substitution effects will need to be extended if the data become
available.
For the period January 2000 to June 2007, the principal statistical results were as follows:
•
•

•
•
•

There was no evidence that the introduction of CEDs served as a substitute for the
use of firearms by an officer.
There was evidence in the data of an incidence shift (structural break) in the
accidental discharge of firearms, but this occurred prior to the introduction of the
CED policy.
There was evidence of an incidence shift in citizen death due to the discharge of
firearms, but this occurred prior to the introduction of the CED policy.
There was evidence of an incidence shift in officer deaths due to the discharge of
firearms, but this occurred prior to the introduction of the CED policy.
There was evidence of an incidence shift in the total discharge of firearms, but this
occurred prior to the introduction of the CED policy.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

57

Complaint Results
For the period December 2004 to June 2007, the principal statistical results were as follows:
•
•
•

Since December 2004, there were 55 complaints filed where CEDs have been
mentioned in some manner.
Complaints were leveled at 57 male officers while 2 were directed at female officers
(note that the 55 complaints included a total 59 officers).
Of the 59 officers noted in the 55 complaints, 27 were Anglo, 20 were African
American, 9 were Latino, and 3 were Asian.

The disposition of complaints is summarized in Table 1a as follows:
TABLE 1a CED COMPLAINTS
CED Complaints4
Total
No Disposition
12
Exonerated
13
Information
1
Never Formalized
2
Not Sustained
9
Open Case
4
Sustained
3
Unfounded
11
Total CED Complaints
55

4

No Disposition - CED activity was not the focus of the complaint and the investigation found CED usage to be
proper and appropriate.

According to GO # 200-03:
Exonerated: Incident occurred, but was lawful and proper.
Information: No evidence to prove that an incident even occurred.
Never Formalized: Complainant refused to make a formal written statement or if a written statement was made,
refused to swear or affirm that the statement was true (notarized).
Not Sustained: insufficient evidence to either prove or disprove justification for the incident.
Open Case: Investigation is on-going.
Sustained: Evidence is sufficient to prove the allegation.
Unfounded: Allegation is false or not factual.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

58

BACKGROUND
The introduction and use of CEDs or what are frequently referred to as Tasers has produced
considerable controversy.5 HPD introduced CEDs in December 2004. By March 2005, all HPD
patrol officers were issued a CED upon completion of a training course.
City Controller Annise D. Parker included a Taser Performance Audit in her 2007 Audit Plan that
was issued to the Mayor and City Council on August 10, 2006. Controller Parker subsequently
contracted with Mir•Fox & Rodriguez, P.C. (MFR) to audit CED use by HPD. The UH CPP was
subcontracted by MFR to conduct the statistical analysis of the CED Performance Audit.6
The public concern about CED incidence was also echoed in the 2007 Houston Area Survey
conducted by Stephen Klineberg (http://houstonareasurvey.org/). Dr. Klineberg’s survey
included responses from 656 people in the Houston area. For a sample of 650, there is a 95percent probability that the data found in the survey will be true for the entire Harris County adult
population within a margin of error of plus or minus 3.5 percent.
Regarding CEDs, the survey asked whether the respondent agrees or disagrees with the
following statements:
1. The use of Taser devices by the police makes deadly force less likely.
Agree: 60.7%
Disagree: 29.1%
Do Not Know or Refuse To Answer: 10.2%.
2. The police are more likely to use Taser devices than less aggressive methods when
the suspect is African American or Latino.
Agree: 49.4%
Disagree: 35.7%
Do Not Know or Refuse To Answer: 14.9%.
However these results mask substantial variations. In particular, the cross-tabulations of the
responses showed distinct cleavages along racial and ethnic lines. In general, Anglos were
more likely than African Americans or Latinos to have a positive view regarding the use of
CEDs.
This polarization of opinion is consistent with the most available data on CEDs provided by
HPD.

5

We use the term Conductive Energy Device (CED) in this document since it is not a commercial term. Taser is a
brand name.
6

Since the implementation of the CED policy, allegations were made that the HPD frequently applies racial profiling
when using CEDs. On November 30, 2006, it was reported in the Houston Chronicle that Mayor Bill White supported
a statistical analysis of CED incidence. The Chronicle reports, “With Houston police facing complaints about Tasers
being deployed disproportionately on African Americans, Mayor Bill White said Wednesday that he wants an
independent, statistical analysis of how the department has used the devices.”

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

59

Table 1b shows that CED incidence was not equally distributed under a variety of
categories. Furthermore, Table 1 summarizes HPD police divisions; shifts where CED
events took place; the number of HPD officers that deployed CEDs for the particular
event; the suspect’s race; the suspect’s gender; and the suspect’s age.7
See Table 1b on next page.

7

In addition the Audit Team noted total CED deployments during the Scope period represent 0.47% of the
approximately 273,000 individuals who were incarcerated in the City’s Jail system.
UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

60

TABLE 1b CED INCIDENCE SUMMARY: DECEMBER 2004 TO JUNE 2007
Source: Crime Analysis and Training Divisions
DIVISION

# OF OFFICERS THAT DEPLOYED TASERS

AIRPORT

2

CENTRAL

112

2 OFFICERS

107

36

3 OFFICERS

16
8

CLEAR LAKE
CRIME ANALYSIS & COMMAND CENTER

I OFFICER

1

4OFFICERS

EASTSIDE

31

5 OFFICERS

FONDREN

69

MULTIPLE SUSPECTS

KINGWOOD

7

NARCOTICS

4

TOTAL

1,133

3
17
1,284

# OF OFFICERS SERIOUSLY INJURED AT TASER
EVENTS

NORTH DIVISION

190

NORTHEAST

216

NO

0

NORTHWEST

40

YES

3

SOUTH CENTRAL

78

TOTAL

SOUTHEAST

192

SOUTHWEST

67

SPECIAL OPS

5

SWAT

4

ASIAN

8

AFRICAN AMERICAN

TRAFFIC
WESTSIDE
X-JOB
TOTAL

SUSPECT RACE

285

82

ANGLO

162

ANIMAL

17

1,267

305

EVENINGS

518

NIGHTS

444

OFFICER SELF INITIATED'ON-VIEW

520

ACCIDENTAL DISCHARGE
TOTAL

17

YES
TOTAL

79

OFFICER

1

24

17-22

197

23-28 (most violent prone years)

331

1,095

29-34

245

53

35-42

242

131

43-49

144

5

50-69

71

UNKNOWN

12

ANIMAL

17

1,284

267

OFFICER

1

TOTAL

1,284

1,017
1,284

# OF SUSPECTS SERIOUSLY INJURED AT TASER
EVENTS

TOTAL

('"~:PUBlICPOLICY

17
1,284

15-16

NONE

UNIVUSITY OF HOU5TON

1,187

FEMALE

SUSPECT AVERAGE AGE

1,284

SUCCESSFUL TASER DEPLOYMENT
NO

1,284

TOTAL

REASON FOR TASER DEPLOYMENT

VERBAL AGGRESSION PHYSICAL GESTURE

TOTAL

ANIMAL
747

THREATENED OFFICER W-WEAPON

1

MALE

1,267

OFFICER DISPATCHED

COMBATIVE RESISTING

OFFICER

SUSPECT GENDER

REASON FOR POLICE RESPONSE TO TASER
EVENTS

TOTAL

810

LATINO

DAYS

MULTIPLE SUSPECTS

9

123

SHIFT

TOTAL

3

61

0
0

During the period December 2004 to June 2007, the data for the 1,284 CED incidents are
summarized below:
•
•
•
•
•
•

•

47% of all incidents occurred in the Northeast (216), Southeast (192), and North
(190) divisions.
75% of all events occurred in the evening (518) or night shifts (444).
88% of all events involving one police officer at the scene were deploying the
CED (1,133).
63% of all suspects were African American (810), 22% were Latino (285), 13%
were Anglo (162), and .7% were Asian (9).8.
92% of all suspects were male (1187) and 6% were female (79)9.
60% of all suspects were between the ages of 17 and 34 (773). 5% of all
suspects were over the age of 50 (71). The modal category was between the
ages 23 and 28.
The total CED deployments during the Scope period represent 0.47% of the
approximately 273,000 individuals who were incarcerated in the City’s Jail
system.

Therefore it was not surprising to find differences in public opinion across geographic and
demographic lines when it comes to this issue.
In general, we noted that the summary data in Table 1 is typical of our experience with other
social science data. Specifically, the data was not equally distributed across a variety of
categories. Rather, there was a concentration or clustering. This clustering raises important
questions for the subsequent analysis.

8
9

The remaining 17 were animals (e.g., dogs) and one officer accidentally deployed the device on himself.
The remaining 17 were animals (e.g., dogs) and one officer accidentally deployed the device on himself.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

62

STUDY COMPONENTS
HPD’s CED program is intended to accomplish the following:
•
•
•
•

Assist officers in securing and controlling combative individuals,
Reduce injuries to officers and suspects,
Reduce financial impact of civil liability in use-of-force incidents, and
In limited situations, provide an alternative to deadly force.

The analysis of these aspects involved the following: statistical, research design, and
measurement challenges that could corrupt valid inference. Addressing these challenges was
fundamental if the policy evaluation conclusions were to have any validity.

Statistical Challenges
To obtain the valid aggregation level and inference, while also accounting for potentially
confounding factors, we examined the probability of CED incidents as a function of individual
and contextual factors, both individually and combined.
Our preliminary examination of CED incidence data suggested that the data possess unique
measurement, sampling, and timing challenges. These challenges required a fairly
comprehensive approach involving several tools that, when taken as a whole, minimized the
threat of drawing false inferences from the data.
We addressed issues of measurement accuracy, sampling validity, and timing through the
following:
•

•

•

Measurement. The first step in the statistical analysis subjected the CED incidence to
tests for measurement, validity, and reliability. This type of assessment was extended
when feasible to other data collected for the analysis.
Sampling. Along with assessing the measurement accuracy of the samples, we broke
the data down by various aggregation levels including HPD division, City Council District,
zip code, and police “beat” levels by combining the CED City incident data with data
contained in the HPD’s Offense Incident Report database. An array of statistical
methodologies including basic descriptive analysis of the main variables of interest and
complex rare-event analysis of matched samples were then employed.
Timing. An intervention analysis was employed to determine if statistically significant
changes occurred in the metrics of interest after the CED technology was announced in
December 2004. The determinants of CED use and the consequences of CEDs that
have evolved from December 2004 to June 2007 were evaluated as well as before and
after the policy was announced or implemented. We chose January 2000 as the
beginning date.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

63

Design Challenges
A common mistake in interpreting data is to take the facts and then directly interpret causal
mechanisms from these facts and correlations. For example, Table 1 shows a set of facts, but
in no way can facts and correlations substitute for causal reasoning.
The Audit Team emphasized that these observations are preliminary and are only the start of a
process to enhance our understanding. This statistical study employs the protocols of social
science, and in particular, we wanted to separate systematic effects of the CED policy from
random chance. A scientific bias requires us to set the barrier high before making any causal
pronouncements. False claims of causation only harm the process of public policy decision
making.
In what sense do we refer to the word cause? Two variables are related if certain values of one
variable tend to coincide with the values of the other variable, but the relation could be purely
episodic. On the other hand, when values of one variable produce the values of the other
variable, then the relation is causal. In other words, correlation is about variables moving
together (they coincide), but causality involves saying not only that two variables coincide but
one variable’s values produce distinct values of the other.
Isolating a causal relation requires the use of controls and holding variables constant. If two
variables, say A and B, “move” together the practice of holding a variable constant means we
introduce a third variable, call it C, and then determine if the introduction of variable C influences
variables A and B such that they no longer move together. Take a hypothetical example where
we find that people with blond hair are more likely to vote for a particular political party (Shively
2008: 76). The fact that a variable representing an adult’s particular hair color is associated with
voting for a particular party’s candidates may or may not be a causal relation. Now, if we add a
third variable, socio-economic status, and take people of the same socio-economic status (i.e.,
hold socio-economic status constant) we may find that there is no difference between people
with blond hair and everyone else when it comes to voting for a particular party’s candidates.
The goal of the research designs the Audit Team employed was to isolate the effects of the
CED policy, and in doing so, separate facts and correlations from causes. In an ideal world we
would want to use a true or natural experiment. A “true” experiment involves a process that
follows the sequence (Shively 2008: 82-84).
Step 1:
Step 2:
Step 3:
Step 4:
Step 5:

The random assignment of subjects to a test group and a control group.
The measurements of the dependent variable for both groups.
A treatment administered to the test group.
A subsequent measure of the dependent variable for the test and
control group.
If the test group “measurements” differ between the first and second
measurements (and subsequent measurements if feasible) then there is
support that the treatment has an effect.

Natural experiments follow a similar structure but the analyst does not have the ability to
assign subjects into test and control groups.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

64

Unfortunately neither of these designs was an option for this study since obtaining control
groups or control locations with the current data was unavailable. As a result, we did not
have a way to randomize the “treatment” over individuals.
The second design challenge was that there was a difference between when the CED policy
was announced and when it was actually implemented. It is difficult to determine with
confidence the date on which implementation of the intervention began and even more
difficult to determine the date of any effect without looking at the data.
To address these matters, we utilized both time series and cross section analysis as
“second best” alternatives to uncover causal patterns. The time series analysis was applied
to policy “intervention” questions. The combination of time series and cross section analysis
was applied to CED incidents where we controlled for characteristics of behavior (with the
available data).

Securing Valid Metrics for the Analysis
While issues of measurement, sampling, and timing are essential to avoiding invalid
inferences, a truly comprehensive research design should make use of contextual
information. This contextual information has enormous potential in making for an accurate
assessment of the true causal factors in any analysis. For this study, the contextual
components of the analysis included several variables that were currently available and can
be linked to data or dates of the CED policy.
There were four categories in this analysis:
1. Incidence: Who is subject to being shocked by a CED?
To determine who is shocked by a CED, the following information was sought:
•

•
•
•

Demographic information on the individuals who come into contact with
the HPD officers to identify the correct population (as well as various subpopulations based on the nature of the contact)
Demographic information on HPD officers
The number of CED incidents in this population
The inclusion of contextual factors such as:
• Location
• Time
• Number of HPD officers present
• Reason for contact (potential violent/potential non-violent offense).

2. Injuries: Have the number of injuries to HPD officers and suspects been affected
by the CED policy?
The following information allowed an assessment of CED related injuries or reduction
in injuries:
•
•

The number of injuries to HPD officers before and after the CED policy
went into effect
The use of the CED compared to alternative methods that involve greater
physical harm (e.g., firearm, baton, flashlight, physical restraint, or
negotiation)

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

65

3. Substitution: Are CEDs used as substitutes for alternative intermediate weapons
or lethal weapons?
The following information led to a determination of whether the CED was used as a
substitute for other weapons:
•

The use of the CED compared to alternative methods that involved
greater physical harm (e.g., firearm, baton, flashlight, physical restraint, or
negotiation)

4. Complaints: How many complaints have been filed against HPD officers for CED
use? What are the demographic characteristics of the complainant and the HPD
officer(s)? How many complaints have been validated?
The following information was used to analyze the CED complaints against HPD
Officers:
•
•

The number of complaints filed and the corresponding demographic data
about CED usage
An analysis of CED usage with HPD officers that used it on more than
one occasion

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

66

DATA
The data categories we analyzed deal with the issues of incidence, injury, substitution, and
complaints. Data sources came from the HPD Crime Analysis Division, the HPD Payroll Office,
and the Internal Affairs Division.10
Incident Data: The best measurable representations that exist in current HPD databases
provide the following information:
•
•
•
•
•

Location (zip code, City Council District)
Type of Incident (UCR subject code)11
Shift (three point: days, evenings, and nights)
Officer Characteristics (race/ethnicity, gender)
Suspect Characteristics (race/ethnicity, gender)

The data are in daily intervals. The period covered was December 2004 to June 2007.
Injury Data: Data on injuries was collected from the City Health and Safety Unit’s workers’
compensation claims. Data was screened to ensure that only claims related to physical
altercations were used in the analysis.12 Data collected pertained to the number of cases that
involved:
•
•
•
•
•
•
•
•
•
•
•
•

Physical altercation (variable name: Altercation)
Foot pursuit that ends in physical altercation (variable name: Pursuit)
Total amount of physical altercations (variable name: Total Comp)
Cost due to physical altercation (variable name: Altercation$)
Cost due to foot pursuit that ends in physical altercation (variable name: Pursuit$)
Total cost of physical altercations (variable name: Total$)
Lost days due to physical altercation (variable name: Altercation Days Lost)
Lost days due to foot pursuit that ends in physical altercation (variable name: Foot
Days Lost)
Lost time due to physical altercation (variable name: Altercation Lost Time) which is
equivalent to the total number of filed claims.
Lost time due to foot pursuit that ends in physical altercation (variable name: Foot
Lost Time) which is equivalent to the number of filed claims.
Total amount of lost days due to physical altercations (variable name: Total Days
Lost)
Total amount of lost time due to physical altercations (variable name: Total Lost
Time) which is equivalent to the total number of filed claims.

10

Note that we also consider the effects of population changes and these changes can influence magnitudes.
Population dynamics are likely to be highly correlated across geographic units (i.e., zip code, council district, etc.),
particularly since the period of analysis is no more than 7 years (within the same Census period). To account for this
potential threat, we create an alternative data scale that standardizes the data as deviation from means. We find no
statistically meaningful difference between using scaled data (deviation from means) and the raw data.
11

Background information on the Uniform Crime Report Program (UCR) can be found at
http://www.fbi.gov/ucr/ucr.htm.
12

The total number of claims for the period January 2000 to June 2007 was 6,260. Of this total, 1,971 (31.5%) were
due to aggressive acts (involving physical altercation).

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

67

The data covered the period January 2000 to June 2007. The descriptive statistics are
presented in Tables 2a and 2b.
TABLE 2a. MONTHLY WORKERS’ COMPENSATION SUMMARY STATISTICS: JANUARY
2000 TO JUNE 2007

Altercation

Altercation$

Pursuit

Pursuit$

Total
Comp

Total$

Mean

15.84

$82,148.20

6.05

$29,524.76

21.90

$111,673.00

Median

16.00

$57,324.50

6.00

$10,290.33

22.00

$85,304.92

Maximum

25.00

$705,392.60

15.00

$184,145.00

32.00

$708,218.40

Minimum

8.00

$130.28

0.00

$0.00

13.00

$2,441.15

Std. Dev.

3.87

$99,049.24

2.94

$40,423.04

4.75

$106,320.90

1,426.00

$7,393,338.00

545.00 $2,657,228.00

1,971.00

$10,050,566.00

Total
number of
altercations

Source: HPD Payroll Office

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

68

TABLE 2b. MONTHLY LOST DAYS AND LOST TIME SUMMARY STATISTICS: JANUARY
2000 TO JUNE 2007
Foot
Lost
Time

Foot Days
Lost

Mean

1.69

Median

Altercation
Days Lost

Altercation
Lost Time

65.20

188.00

4.02

5.71

253.00

1.50

18.50

158.00

4.00

5.50

226.00

Maximum

8.00

563.00

778.00

10.00

14.00

844.00

Minimum

0.00

0.00

0.00

0.00

1.00

4.00

Std. Dev

1.54

96.88

160.00

2.03

2.71

181.64

152.00

5,867.00

16,893.00

362.00

514.00

22,760.00

Total

Total Lost
Time

Total Days
Lost

Source: HPD Payroll Office
According to the HPD Payroll office “Days Lost” in Table 2b refers to the total number of days
lost. “Lost Time” refers to the number of claims that pertain to the 8 hour work shifts that were
lost.
Among the results (see Table 2a), the mean level of monthly expenditures on workers’
compensation claims was $111,673, with approximately 22 claims made per month. The
maximum dollar expenditure for non foot pursuit claims (Altercation$) occurred in February 2003
with a total of $705,393 and the minimum occurred in June 2007 with a total of $130.28.13 The
total dollars spent for the entire period was $10,050,566 (see Table 2a) while the total days lost
for the period was 22,760 days (see Table 2b). In Table 2b the average monthly total for days
lost (Total Days Lost) was 253 days with a maximum of 844 days and a minimum of 4 days.
In Figures 1 through 4, are the time series behavior of total monthly claims (Total Comp), total
monthly expenditures (Total$), total days lost due to aggressive acts (Total Days Lost), and total
lost time due to aggressive acts (Total Lost Time).

13

The expenditures were calculated so that any expenses after the claim were always rolled into the date of the
original claim. This may be a reason for the low total in June 2007.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

69

FIGURE 1

70

FIGURE 2

Total Expenditures on Injury Claims Due to Aggressive Acts:
January 2000 to June 2007
Monthly
Dollars

800000
700000
600000
500000
400000
300000
200000
100000
O--+n,.'T"T"T"T"'I'T"T"T"T"'I'I"TT'T"l'TTTT'T'TTTT'T'T'TTTT"T'TTTT"rTTTT'1"TTTT'l"TTTT'l'T'TTT'T'T'TTT'T'T"T"T"T"'ITT'T'TTTT'T'TTT'T'i-TTT"T"iTI-~

2000

2001

2002

2003

2004

Year

UNIVERSITY OF HOUSTON

cenl:;PUBlICPOlICY

71

2005

2006

2007

FIGURE 3

Total Days Lost Due to Aggressive Acts:
January 2000 to June 2007
Monthly
Totals

900
800700600500400-

(\

I

300-

I
V

11
200-

f'J

v

~

100-

If"

o--+rrr'T"T"T"T"T'T'T"T"T"T"~~"I"T"T"T"T'T"T"T"T"T'TT'Tv"I"T"T"T"T'T"T"T"T"T'T"T"T"T"T"~"TT'T"T"T"I"T"T"T"T\JT'T"T"T"T"~r"T"T"'i-r1"'T"T"T"T"TTTTT'fI
I

2000

I

2001

I

2002

I

2003

'I

2004

Year

UNIVERSITY OF HOUSTON

cenl:;PUBlICPOlICY

72

I

2005

I

2006

2007

FIGURE 4

Total Lost Time Due to Aggressive Acts:
January 2000 to June 2007

0~""""""'''''''''''''''r'''r''l'''l''~~.,.,..,..,.,..,.,..,..,.,.,.,.,..,..,..~~.,..,.,..,..,.,..,..,.,..,..,..,..,.,..,..~~.,.,..,..,.,.~1'''l'''l'''!I
I
I
I
I
I
I
I

2000

2001

2002

2003

2004

Year

UNIVERSITY OF HOUSTON

cenl:;PUBlICPOlICY

73

2005

2006

2007

Substitution Data: Data on weapon substitution was collected from the HPD Internal Affairs
Division (see Table 3). Existing data collected pertained to the number of cases that involved:
•
•
•
•
•
•
•
•
•

Discharge of firearms that was an accident (variable name: Accident)
Discharge of firearms that resulted in a citizen’s death (variable name: Citizen Death)
Discharge of firearms that resulted in a citizen’s injury (variable name: Citizen Injury)
Discharge of firearms that resulted in a citizen’s death and injury (variable name:
Citizen Death/Injury)
Discharge of firearms that resulted in an officer’s death (variable name: Officer
Death)
Discharge of firearms that resulted in an officer’s injury (variable name: Officer Injury)
Discharge of firearms that resulted in an officer’s death and injury (variable name:
Officer Death/Injury)
Discharge of firearms that resulted in property damage (variable name: Property
Damage)
Discharge of firearms - total from categories above (variable name: Total/No Animal).

The data covered the period January 2000 to June 2007. The descriptive statistics are
presented in Table 3.
TABLE 3. MONTHLY DISCHARGE OF FIREARMS SUMMARY STATISTICS: JANUARY
2000 TO JUNE 2007

Accident

Citizen
Citizen Citizen
Death/
Death Injury
Injury

Officer
Death

Officer
Injury

Officer
Death/
Injury

Property Total/No
Damage Animal

Mean

0.58

0.59

1.08

1.67

0.08

0.04

0.12

0.06

4.62

Median

0.00

0.00

1.00

2.00

0.00

0.00

0.00

0.00

4.00

Maximum

3.00

3.00

4.00

5.00

2.00

1.00

2.00

1.00

14.00

Minimum

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

1.00

Std. Dev.

0.72

0.70

1.07

1.34

0.34

0.21

0.39

0.23

2.26

52.00

53.00

97.00

150.00

7.0

4.0

11.00

5.00

416.00

Total

Source: HPD Internal Affairs Division
We noted that the mean level for the total discharge of firearms (not involving animals) was 4.62
and the total for the period was 416. The monthly maximum for the total discharge of firearms
total was 14 and the minimum was 1. For citizen and officer deaths due to the discharge of
firearms, the totals for the period were 53 and 7 respectively.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

74

Figure 5 provides a summary of the time series behavior for a selected variable: the total
discharge of firearms (Total/No Animals).
FIGURE 5

Total Discharge of Firearms:
January 2000 to June 2007
Monthly
Totals

16
14 -

12108-

A

6-

~

/~

4-

~

V ..,

f\
~

rt
I~
IN
{~

V

~

r

20---1TTTT'T"T"T"T"T'TT'T"T"~"'T"T"T'TT'T"T"T"T"'ITTTT'T"T'TT'T"T"I'T'T"T"T"l"TTT'T"T'T"T"T"T"'I'T"T"T'TTT'T'T"T"T"I'T'T"T"T"l"TTT'T"T'T"T"T"T"'I'T"T"T'TTT"T"'i-n-rT"T"'I
I
'I
I
I
I
I
I

2000

2001

2002

2003

2004

2005

2006

2007

Year
Complaint Data: The data found in Table 4 contains all complaints filed against HPD officers in
regard to the use of CEDs between December 2004 and June 2007. The data contained the
following information:
•
•
•
•

Demographic characteristics of complainants and officers
Number of officers and complainants present
Reason for contact (potential violent/potential non-violent offense)
Disposition of the complaint

UNIVERSITY OF HOUSTON

cenl:;PUBlICPOlICY

75

TABLE 4. CED COMPLAINT STATISTICS: DECEMBER 2004 TO JUNE 2007
CED Allegation Totals

CED Disposition Totals

CHIEF INFORMATION OFFICER (CIO) ISSUE

3

5%

No Disposition***

12

22%

CONDUCT AND BEHAVIOR

5

9%

EXONERATED

13

24%

CRIMINAL ACTIVITY

1

2%

INFORMATION

1

2%

DEATH IN CUSTODY*

1

2%

NEVER FORMALIZED

2

4%

IMPROPER POLICE PROCEDURE

6

11%

NOT SUSTAINED

9

16%

MISCONDUCT

4

7%

OPEN CASE

4

7%

OPEN CASE

4

7%

SUSTAINED

3

5%

TASER NOT ADDRESSED IN SYNOPSIS**

9

16%

11

20%

22

40%

55

100%

55

100%

USE OF FORCE
Total:
* In custody death. Harris County
Medical examiner ruled death due
to cocaine toxicity. Death not related to
CEDs.

UNFOUNDED

Total:
*** No Disposition - CIO Issues, or CED
was not addressed in the synopsis.
In these instances. CED activity was not
the focus of the complaint and the investigation
found CED usage to be proper and appropriate.

** CED was used. but was not the
focus of the complaint or investigation.
CED use was deemed to be appropriate.
Complainant Demographics

Officer Demographics
Total Officers

59

Total Complainants:

Race

51

Race

Anglo:

27

46%

Anglo:

7

14%

African America:

20

34%

African American:

36

71%

Latino:

9

15%

Latino:

7

14%

Asian:

3

5%

Asian:

1

2%

Other or Unknown:

0

0%

Other or Unknown:

0

0%

Male:

39

76%

Female:

12

24%

Gender

Gender

Male:
Female:

57

97%

2

3%

Some officers were involved in multiple complaints. Some complaints had more than one officer involved.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

76

METHODS OF ANALYSIS
Tests for Incidence
There were four pieces of data that we used in our analysis of CED utilization. We started with
the universe of incidents over our time period.14 We then created subsets by breaking these
incident data down to incidents that could be matched with a suspect. In the case that an
incident had more than one suspect, we used the first listed suspect. Next, we matched officers
to this data. Finally, we matched this data with the CED data.
Econometric Analysis
We constructed a statistical model of CED utilization that could provide an estimate of the
probability of CED use as a function of incident characteristics, suspect characteristics, and
officer characteristics. Specifically, we estimated a linear probability model to evaluate the
relative effects of incident, suspect, and officer characteristics. From these models we
simulated the predicted probability and their 90% confidence intervals of CED utilization as a
function of our observable variables. We conducted several robustness checks to be sure our
results were not a function of any particular assumption of the linear probability model. To
address the fact that CED use is an indicator variable, we used maximum likelihood estimation
to obtain the estimated probability of CED use. This robustness check suggested that our initial
results from the linear probability model were consistent. In addition, we used a model to
correct for the empirical fact that CED use was a rare event. Although these models do not tend
to fit the data particularly well, on the whole, the results were comparable to the results obtained
from the linear probability model.
Visualization Analysis
We augmented this particular statistical analysis with a visualization platform. The visualization
platform maps all incidence data by geographic placement within Harris County and over time.
The platform is available on the UH CPP website at http://www.uh.edu/cpp.

14

We chose the universe of cases, where the universe of cases for this period of analysis involved merging primary
suspect data with an incident and an HPD officer (who could have deployed a CED). We relied on using a data set
that was much larger than the number of CED incidents. In particular, we wanted to provide a control group to make
probabilistic comparisons of how the distribution of observable variables in the CED data may differ from the
distribution of the observable variables in the incidents where a CED was not deployed. There were many ways to
segment the data, from the very broadest categorization to much narrower ones. In this initial analysis, we chose the
broadest categorization because it required us to make no assumptions about how the probability of CED utilization
might shift as a function of observable variables. We allowed the data to speak rather than make such assumptions.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

77

Tests for Policy Effects
The tests for policy effects involved the use of time series data. The analysis involved
the following:
•
•

Determination of incidence shifts (or structural breaks) in the data
The combination of linking data persistence with incidence shifts when possible
(intervention analysis with dummy variables)

To identify incidence shifts in the data and to determine if they corresponded with the
introduction of the CED policy, we employed two types of tests.15 The first test showed
the timing of the largest break in the data (Andrews 1993). The second test analyzed
how many breaks occurred in the data (rolling paired t-test). The tests are described as
follows:
The Andrews Test
Instead of setting break points by some subjective assessment of timing, the
Andrews test uses the entire time series to determine if any break points exist. The
focus on the largest and most significant break point secures information on
whether the changes in the series structure occurred before or after policy changes.
The Andrews statistic is calculated as follows. First, compute and find the
maximum Wald statistic for the entire series.16 Then determine if the maximum
Wald statistic exceeds the critical value. Maximum values for a given series that
are larger than the critical value are interpreted as rejections of the null hypothesis.17
There are two limitations to the Andrews test. First, it allows for only one break in
the time series. Second, the Andrews test is tied to a specific regression
specification.
The Rolling Paired t-test
An alternative estimation is to compare the mean of variable for two sample periods
- before and after the treatment - using a rolling paired t-test (Cureton 1957). While
the Andrews test determines a break point in a regression form where controlled
variables are necessarily included, the rolling paired t-test only examines the
equality of means in two groups (or periods). We relaxed the restriction of
controlling variables in regressions and searched for break points by running rolling
paired t-tests over the break points to search break points with the highest
t-statistics.18

15

In the case of the CED policy, HPD announced the policy in December 2004, but the training and distribution of the
CED was not completed until March 2005.
16

In applying the test we followed Andrew’s “trimming” rule. Trimming involves how deep into the sample (a
proportion) to look for structural breaks (shifts) as well as how close to the end of the sample to end the search. The
proportions should be large enough to include sufficient data points and small enough to encompass the largest
number of potential breaks. In this paper, we used 25 percent trimming as a baseline, but compared these results
with other trimming proportions. The conclusions were robust to alternative proportions.
17

The null hypothesis is traditionally set to indicate no break (i.e., no program effect).

18

This type of test is similar to Quandt (1958), Goldfeld and Quandt (1973), Hinkley (1971), and McGee and Carlton
(1970). They also run regressions over the break point periods to search for break points with the minimum residual
sum of squares (RSS).

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

78

Intervention Analysis
We combined the results of the Andrews and rolling paired t-tests with an intervention
analysis. The dates for incidence shifts were identified by these two tests, but we also
placed these dates within a regression framework to determine the actual change in the
level of the dependent variable. In addition, one of the attributes of time series analysis
was that point estimates for the immediate effect were adjusted to determine the longrun or steady state effect. It was sometimes the case that analysts ignored the
cumulative effect and focus on the point estimates. This would be a mistake as the point
estimate effect could be dwarfed by the long-run cumulative effect. A useful way to test
a hypothesis is to examine the effect of a specific policy change. These possibilities are
great since many subjects in the social sciences are influenced by changes in regime or
policy. The interventions can be characterized in many ways, but they generally can be
categorized as either temporary or permanent.

RESULTS
Incidence Analysis
The main purpose of the incidence analysis was to examine the role of suspect and
officer race/ethnicity in the use of CEDs by HPD officers between December 2004 and
June 2007. The results of the analysis focused on suspects, then officers, and then
relevant governmental/geographical units (the nine Houston City Council Districts).
To conduct this analysis we merged data from three separate datasets: Offense
Incident, Suspect, and Officer Employee Number for the period December 2004 through
June 2007. HPD officers were equipped with CEDs (analysis was also done excluding
the first four months when not all officers were equipped with CEDs). Four sets of
variables were employed in the analysis:
•
•
•
•

Suspect
Officer
Crime Context
Geographic/Temporal Context19

Two types of Suspect data were utilized: suspect race/ethnicity (African American,
Anglo, Latino, or other) and gender (male or female).20 All offenses without a suspect or
when the suspect was not human were excluded from the analysis. Information on an
officer's physical characteristics (i.e., height, weight, strength) was not available.

19

Recall that the variables we used involved:
•
Incident Location (zip code, Council District)
•
Type of Incident (UCR subject code)
•
Shift (three point: days, evenings, and nights)
•
Officer Characteristics (race/ethnicity, gender)
•
Suspect Characteristics (race/ethnicity, gender)

20

An insufficient number of cases with Asian suspects and Asian officers exist to conduct reliable analysis when
these two groups are examined separately.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

79

Additional desirable suspect data such as physical characteristics (weight/height/size)
were unavailable for a large majority of suspects, and even in those instances when
available, the data was considered unreliable (or at best, extremely imprecise). A
possible critical variable, information on a suspect’s past criminal record, was not
available. The lack of information on suspect characteristics required that we be very
cautious in interpreting all of the suspect related data analysis (i.e., that analysis which
examines the impact of suspect race/ethnicity on the likelihood that they are the subject
of a CED event). In contrast, we are more comfortable making interpretations based on
the officer related data (i.e., the analysis which examines the impact of officer
race/ethnicity on the likelihood that an officer employs a CED), since most important
contextual factors are controlled for in this analysis by the essentially random
assignment of officers to incidents.
As of 2006, the City’s population had the following racial/ethnic distribution: Latino
(41.9%), Anglo (27.6%), African American (24.7%), and Other Groups (5.8%).
The Audit Team compared the race/ethnicity of suspects in the Total Service/Incident
Reports Analysis Population to the race/ethnicity of the suspects noted by HPD in the
CED Service/Incident Reports. The results of our comparison are as follows:
Table 5a summarizes the Race/Ethnicity of the suspects that were in the Analysis
Population of the Service/Incident Reports and the Race/Ethnicity of the CED
Service/Incident Reports.
TABLE 5a. RACE/ETHNICITY OF SUSPECTS

Suspect’s
Race/Ethnicity
African American
Latino
Anglo
Other Groups

Total
Service/Incident
Reports Analysis
Population
46.0 %
28.2 %
24.4 %
1.4 %

CED
Service/Incident
Reports
66.9 %
23.5 %
9.0 %
0.6%

Difference
20.9 %
-4.7 %
-15.4 %
-0.8 %

Based on the above analysis, African American suspects were involved in a
proportionally greater number of total Service/Incident Reports analyzed as well as CED
service/incident reports. In addition, the proportion of CED Service/Incident Reports was
20.9% more than the total service/incident reports. The Latino, Anglo, and Other Group
suspects were involved in proportionally less service/incident reports.
According to HPD, the Department-wide officer demographics (see table 5b) during the
Scope period were as follows:

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

80

TABLE 5b. NUMBER OF CLASSIFIED OFFICERS BY RACE/ETHNICITY AND GENDER
2004 Classified

Male
W

P

A

B

H

**********

2005 Classified

Male
W

P

A

B

Total
M

Female
W

P

A

B

H

Total F

Total M/F

A

B

H

Total F

Total M/F

No data available **********

Total
M

Female

H

W

P

January

1,361

79

3

445

508

2,396

132

4

-

93

51

280

2,676

February

1,333

78

3

442

501

2,357

131

3

-

91

50

275

2,632

March

1,327

78

3

440

498

2,346

129

3

-

91

49

272

2,618

April

1,304

79

3

439

495

2,320

126

3

-

90

47

266

2,586

May

1,289

78

3

438

492

2,300

125

3

-

89

47

264

2,564

June

1,281

77

3

439

487

2,287

124

3

-

88

47

262

2,549

July

1,277

77

3

436

487

2,280

124

3

-

88

47

262

2,542

August

1,272

77

3

434

489

2,275

125

3

-

87

47

262

2,537

September

1,237

85

3

440

507

2,272

125

3

-

90

46

264

2,536

October

1,286

84

3

440

501

2,314

126

3

-

90

46

265

2,579

November

1,275

84

3

438

500

2,300

126

3

-

89

46

264

2,564

December

1,268

82

3

431

499

2,283

124

3

-

87

45

259

2,542

2005 Average

1,293

80

3

439

497

2,311

126

3

-

89

47

266

2,577

A

H

Total
M

Female

P

W

P

B

H

Total F

Total M/F

2006 Classified

Male
W

B

A

January

1,260

84

3

436

496

2,279

120

3

-

88

47

258

2,537

February

1,274

89

3

454

526

2,346

125

3

-

95

55

278

2,624

March

1,263

90

3

453

524

2,333

125

3

-

95

56

279

2,612

April

1,267

91

4

459

531

2,352

126

3

-

110

55

294

2,646

May

1,265

91

4

456

525

2,341

124

4

-

110

53

291

2,632

June

1,258

91

4

457

533

2,343

126

4

-

110

54

294

2,637

July

1,258

91

6

460

532

2,347

129

4

-

112

55

300

2,647

August

1,256

98

6

463

542

2,365

130

4

-

112

52

298

2,663

September

1,255

98

6

467

541

2,367

132

4

-

111

53

300

2,667

October

1,248

98

5

461

544

2,356

129

3

1

110

53

296

2,652

November

1,247

96

5

459

537

2,344

130

4

1

110

54

299

2,643

December

1,249

97

5

460

537

2,348

130

4

1

113

51

299

2,647

2006 Average

1,258

93

5

457

531

2,343

127

4

0

106

53

291

2,634

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

81

TABLE 5b (continued). NUMBER OF CLASSIFIED OFFICERS BY RACE/ETHNICITY AND
GENDER

2007 Classified

Male
W

P

A

B

Total
M

Female

H

W

P

A

B

H

Total F

Total M/F

January

1,245

94

4

457

551

2,351

127

4

1

11

47

290

2,641

February

1,246

98

4

458

550

2,356

128

4

1

110

49

292

2,648

March

1,246

100

3

455

554

2,358

123

4

1

114

46

288

2,646

April

1,256

101

3

462

561

2,383

123

4

1

116

49

293

2,676

May

1,263

105

3

460

564

2,395

123

5

1

119

49

297

2,692

June

1,263

107

4

453

569

2,396

119

6

117

49

291

2,687

2007 Average

1,253

101

4

458

558

2,373

124

5

98

48

292

2,665

W=

White

P=
A=

Asian or Pacific Islander
American Indian or Alaskan
Native

B=

Black

H=

Hispanic

1

The Audit Team compared the race/ethnicity of officers in the Total Service/Incident
Reports Analysis Population to the race/ethnicity of the officers noted by HPD in the
CED Service/Incident Reports. Table 5c summarizes the results of our comparison are
as follows:

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

82

Table 5c summarizes the Race/Ethnicity of the Officers that were in the Analysis
Population of the Service/Incident Reports and the Race/Ethnicity of the CED
Service/Incident Reports.
TABLE 5c RACE/ETHNICITY OF OFFICER’S
Total Service/Incident Reports
CED
Officer’s
Analysis Population
Service/Incident Reports Difference
Race/Ethnicity
African American
25.1 %
17.3 %
-7.8 %
Latino
24.3 %
27.9 %
3.6 %
Anglo
46.2 %
52.3 %
6.1 %
Other Groups
4.4 %
2.5 %
-1.9 %
Based on the above analysis, the positive values in the Difference column indicate the
officer racial/ethnic group was involved in a proportion of CED events that was larger
than the proportion represented by it for all HPD incidents in the Analysis Population.
Negative values indicate the officer racial/ethnic group was involved in a proportion of
CED events that was smaller than the proportion represented by it for all HPD incidents
in the Analysis Population.
For our statistical analysis, the crime context was measured using splines (both
individual and grouped) based on the UCR subject codes. Geographic context was
measured using two types of data: zip code splines and splines for the nine City Council
Districts. We focused on the City Council Districts in the analysis presented here given
their more substantively meaningful status within this analysis.21 The nine City Council
Districts were employed as substantively meaningful dummy variables which allowed us
to control for one key contextual variable (the geographic location of the officer-suspect
interaction).
The City Council Districts had the advantage of representing distinct regions of the City.
They were independent of any HPD or investigator decisions (e.g., they could not be
altered to affect the outcome of the analysis). They were mutually exclusive (i.e., events
can occur in one, and only one Council District) and, in part as a consequence of the
Voting Rights Act (combined with moderate to high racial/ethnic housing segregation)
are representative of the ethnic/racial context within which officer-suspect interactions
take place in the City.

21

Recall that we used a host of statistical procedures in the incidence analysis. In particular, we used Ordinary Least
Squares regression analysis, Logistic regression analysis, and Rare Events regression analysis. In addition, under
each of these econometric regimes a variety of diagnostics were employed ranging from extreme bounds tests to
sub-group evaluation. The results presented here represented our summary judgment regarding the impact of the
variables taking into account the combined results of this meta-analysis. The main analysis population was
approximately 570,000 individual cases (due to missing data issues, the actual number varied somewhat depending
on the specific analysis population).

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

83

In contrast, alternative geographic variables, such as HPD Divisions, could not provide
context since they were not mutually exclusive in terms of their geographic coverage
(some Divisions were City-wide and thematic while others were geographically based).
At the same time the reliability of the decision-rules utilized to place incidents were under
different HPD Divisions was not entirely clear.
The City Council District level analysis opened an important window on the context in
which officer-suspect interaction took place in the City of Houston, providing important
caveats to broad-brush City-wide interpretations of the data. For instance (see Table 9),
while City-wide Anglo officers were more likely to utilize their CED when interacting with
African American suspects than were African American officers, in District D (the Council
District in which the largest number of African American suspects were involved in a
CED event), African American officers were just as likely to use a CED as were their
Anglo counterparts (both when speaking of all suspects as well as when limiting the
analysis to African American suspects).
For the present data analysis, the number of CED events was too fragmented across
HPD's 19 Divisions to allow for valid City-wide analysis of CED events while employing
HPD Divisions rather than Council Districts as the contextual control variables. In fact,
even if we focused on Division-level analysis comparable to that conducted for the nine
City Council Districts, it would only be possible to conduct this analysis for 9 of the 19
HPD Divisions (eight of which are geographically defined Divisions, and one of which is
the City-wide "Extra Jobs" Division).

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

84

Council Districts
Guided by social science protocol, we determined that the most objective unit of analysis
was City Council Districts.
Using U.S. Census 2000 data, the racial/ethnic breakdown of the various City Council
Districts are shown in Figure 6.
FIGURE 6. THE RACIAL/ETHNIC DISTRIBUTION OF THE POPULATION OF THE
HOUSTON CITY COUNCIL DISTRICTS

PERCENTAGE OF THE POPULATION (US CENSUS 2000)

100
90
80
70
70

66
60

59

60

66

58
African American
Anglo

50

44
40

43

Latino

40

40
31
30

26

25

26
22
18

19
16

20
10

9

8

10

15

18
15

16
11

9

0

A

B

C

D

E

F

CITY COUNCIL DISTRICT

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

85

G

H

I

CED deployments took place in all City Council Districts during the scope period. Table 6 below
summarizes the total number of CED deployments by Council District for the periods November
and December 2004, 2005, 2006, and January through June 2007.
TABLE 6. CED DEPLOYMENTS BY CITY COUNCIL DISTRICT
Scope Period
Council
District
A
B
C
D
E
F
G
H
I
Total

November and December
2004
3
5
2
7
4
3
3
4
4
35

2005 2006
31
25
101
85
48
49
123
93
31
20
29
37
20
24
85
82
62
64
530
479

January through June
2007
18
42
21
37
10
16
13
26
22
205

Total
77
233
120
260
65
85
60
197
152
1,249

Suspect Component
The results suggest that African American suspects were significantly more likely to be
subject to a CED shock than Anglo or Latino suspects (see Tables 7A, 7B, and 7C).
This is an observation that was significant, for both City-wide and within five of the nine
City Council Districts of Anglos and three of the nine City Council Districts for Latinos
(see Table 8).22 Latino suspects were significantly more likely to have a CED used on
them than Anglo suspects, although this observation was not especially strong as well
as only present in one of the nine Council Districts. Male suspects were significantly
more likely to be subject to a CED shock than female suspects (see Table 9). This
observation was present in all nine Council Districts.

22

A result that was strong and significant in some Districts does not imply that it was not present in others; rather it
means that we do not consider the finding as strong as what we found at the City-wide level. We did not find results
“flipping” signs in a significant manner at the District level.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

86

TABLE 7a. CED USE IN HOUSTON, AFRICAN AMERICAN SUSPECTS

COMPARISON SUSPECT GROUP

THE LIKELIHOOD THAT A CED IS USED WHEN AN AFRICAN AMERICAN IS THE SUSPECT
COMPARED TO:

Anglo Suspect

Much Higher

Much Higher

Equal

Equal

Much Higher

Much Higher

Latino Suspect

Much Higher

Much Higher

Equal

Equal

Much Higher

Equal

Full

Male
Officers

Female
Officers

African
American
Officers

Anglo
Officers

Latino
Officers

Analysis Population

TABLE 7b. CED USE IN HOUSTON, LATINO SUSPECTS
COMPARISON SUSPECT GROUP
African American Suspect
Anglo Suspect

THE LIKELIHOOD THAT A CED IS USED WHEN A LATINO IS THE SUSPECT COMPARED TO:
Much Lower

Much Lower

Equal

Equal

Much Lower

Equal

Higher

Higher

Equal

Equal

Equal

Higher

Female
Officers

African
American
Officers

Anglo
Officers

Latino
Officers

Male
Analysis Population
Full
Officers
TABLE 7c. CED USE IN HOUSTON, ANGLO SUSPECTS
COMPARISON SUSPECT GROUP
African American Suspect
Latino Suspect

Analysis Population

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

THE LIKELIHOOD THAT A CED IS USED WHEN AN ANGLO IS THE SUSPECT COMPARED TO:
Much Lower

Much Lower

Equal

Equal

Much Lower

Much Lower

Lower

Lower

Equal

Equal

Equal

Lower

Full

Male
Officers

Female
Officers

African
American
Officers

Anglo
Officers

Latino
Officers

87

TABLE 8. LIKELIHOOD A SUSPECT OF THE GROUP IN COLUMN A WILL BE INVOLVED
IN A CED DEPLOYMENT COMPARED TO A SUSPECT OF THE GROUP IN COLUMN B:
FULL POPULATION AND INVOLVING OFFICERS BY RACE/ETHNICITY
ANALYSIS
POPULATION

Full

African American
Officers

Anglo Officers

Latino officers

SUSPECT
COLUMN A

SUSPECT
COLUMN B

African American

Anglo

African American

Latino

Latino

Anglo

African American

Anglo

African American

Latino

Latino

Anglo

African American

Anglo

African American

Latino

Latino

Anglo

African American

Anglo

African American

Latino

Latino

Anglo

RED = Much Higher Likelihood
ORANGE = Higher Likelihood
Blank = Equal Likelihood

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

88

CITY COUNCIL DISTRICT

A

B

C

++ ++

D E F G

H

I

++ ++
++ ++

++

++ ++
++ ++

++

TABLE 9. CED USE IN HOUSTON, FEMALE SUSPECTS
COMPARISON
SUSPECT GROUP
A MALE SUSPECT

ANALYSIS
POPULATION

THE LIKELIHOOD THAT A CED IS USED WHEN A FEMALE
IS THE SUSPECT COMPARED TO:

Much
Lower

Much
Lower

Lower

Much
Lower

Much
Lower

Much
Lower

Full

Male
Officers

Female
Officers

African
American
Officers

Anglo
Officers

Latino
Officers

When the analysis population was restricted to incidents involving African American Officers
(see Tables 7A, 7B, and 7C), there were no racial/ethnic differences in the probability of
suspects having a CED used on them. In other words, all suspects - whether African American,
Anglo, or Latino - were all equally likely to be subject to a CED shock by African American
officers.
When the analysis population was restricted to incidents involving Latino officers (see Tables 7A
7B, and 7C), African American suspects were significantly more likely to be subject to a CED
shock than Anglo suspects. This was a strong and significant relationship that was present in
three council districts (see Table 8). Latino suspects were slightly more likely to have a CED
used on them than Anglo suspects. This latter relationship was relatively modest City-wide and
present in a significant manner in only one City Council District.
When the population was restricted to incidents involving Anglo officers (see Tables 7A, 7B, and
7C), African American suspects were significantly more likely to be subject to a CED shock than
Anglo suspects. Latino suspects were very marginally more likely to have a CED used on them
than Anglo suspects, and African American suspects were marginally more likely to be subject
to a CED shock than Latino suspects. Both of these latter results were not especially strong,
with the former noteworthy City-wide but not at the City Council District level (except in one
district) while the latter is not noteworthy at the City-wide level, but was strong and significant in
four City Council Districts (see Table 8).
The main conclusion from the analysis above was that African American suspects were
significantly more likely to be subject to a CED shock than Anglo and Latino suspects.
However, this greater probability of having a CED used on them was only the case when the
officer was Anglo or Latino. The results also demonstrate that Latinos were marginally more
likely to be subject to a CED shock than Anglos (though only when the officer is Anglo or
Latino). This observation is substantially less robust than that regarding African American
suspects.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

89

As mentioned earlier, the lack of adequate suspect data (height/weight/size; criminal history)
required that we treat the above results with considerable skepticism. It is very likely that our
models suffered from omitted variable bias and that if proper controls regarding the suspect
characteristics were included, many of the significant results we identified would vanish.23
Suspect and Officer Component
The results suggest that African American officers were significantly less likely to use a CED on
suspects than both Anglo and Latino Officers (see Tables 10A, 10B, and 10C). This result was
significant City-wide as well as in four of the nine City Council Districts for Anglos and four of the
nine City Council Districts for Latinos. Anglo, and Latino Officers were equally likely to use a
CED on suspects (see Table 11).
This result is present City-wide as well as in all City Council Districts. Male and female officers
were equally likely to use a CED on suspects (see Table 12).
TABLE 10a. CED USE IN HOUSTON, AFRICAN AMERICAN OFFICERS
COMPARISON
OFFICER GROUP
AN ANGLO
OFFICER
A LATINO OFFICER

ANALYSIS
POPULATION

THE LIKELIHOOD THAT A CED IS USED WHEN AN
AFRICAN AMERICAN IS THE OFFICER COMPARED TO:

Much
Lower
Much
Lower

Full

Much
Lower
Much
Lower

Much
Lower
Much
Lower

Much
Lower
Much
Lower

African
Female American
Male
Suspects Suspects Suspects

23

Equal

Equal

Equal

Lower

Anglo
Suspects

Latino
Suspects

In a separate analysis we also controlled for the number of years an officer had been on the force, utilizing a variety
of functional forms. By including this additional variable/set of variables to control for years on the force, we reduced
the overall analysis population by approximately one-fifth (due to the lack of data for officer years on the force for
many cases). Furthermore, analysis controlling for years on the force provided general conclusions similar to those
presented here (although in a few specific instances, some sub-conclusions were altered slightly, although this was
also in part due to the reduction in the number of overall cases analyzed). As a result of the above factors, we did not
include the analysis that incorporated the years in service control set here (the results of this analysis can be
obtained from the authors upon request).

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

90

TABLE 10b. CED USE IN HOUSTON, LATINO OFFICERS
COMPARISON
OFFICER GROUP
AN AFRICAN
AMERICAN
OFFICER
AN ANGLO
OFFICER

THE LIKELIHOOD THAT A CED IS USED WHEN A LATINO IS
THE OFFICER COMPARED TO:

ANALYSIS
POPULATION

Much
Higher

Much
Higher

Much
Higher

Much
Higher

Equal

Higher

Equal

Equal

Equal

Equal

Equal

Equal

Female
Suspects

African
American
Suspects

Anglo
Suspects

Latino
Suspects

Full

Male
Suspects

TABLE 10c. CED USE IN HOUSTON, ANGLO OFFICERS
COMPARISON
OFFICER
GROUP
AN AFRICAN
AMERICAN
OFFICER
A LATINO
OFFICER

THE LIKELIHOOD THAT A CED IS USED WHEN AN ANGLO IS
THE OFFICER COMPARED TO:

Much
Higher

Much
Higher

Much
Higher

Much
Higher

Equal

Equal

Equal

Equal

Equal

Equal

Equal

Equal

Full

Male
Suspects

Female
Suspects

African
American
Suspects

Anglo
Suspects

Latino
Suspects

ANALYSIS
POPULATION

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

91

TABLE 11. LIKELIHOOD AN OFFICER OF THE GROUP IN COLUMN A WILL BE
INVOLVED IN A CED DEPLOYMENT COMPARED TO AN OFFICER OF THE GROUP IN
COLUMN B: FULL POPULATION AND INVOLVING SUSPECTS BY RACE/ETHNICITY
ANALYSIS
POPULATION
FULL
(i.e. All Suspects)

AFRICAN
AMERICAN SUSPECTS

ANGLO
SUSPECTS

LATINO
SUSPECTS

OFFICER
COLUMN A

OFFICER
COLUMN B

A

COUNCIL DISTRICT
B C D E F G H

AFRICAN AMERICAN

ANGLO

**

AFRICAN AMERICAN

LATINO

** **

ANGLO

LATINO

AFRICAN AMERICAN

ANGLO

AFRICAN AMERICAN

LATINO

ANGLO

LATINO

AFRICAN AMERICAN

ANGLO

AFRICAN AMERICAN

LATINO

ANGLO

LATINO

AFRICAN AMERICAN

ANGLO

AFRICAN AMERICAN

LATINO

**

ANGLO

LATINO

**

GREEN = Much Lower Likelihood
BLUE = Lower Likelihood
Blank = Equal Likelihood
ORANGE = Higher Likelihood

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

92

I

** **

**
**

**

**

** **

TABLE 12. CED USE IN HOUSTON, FEMALE OFFICERS
COMPARISON
OFFICER GROUP
A MALE OFFICER

ANALYSIS
POPULATION

THE LIKELIHOOD THAT A CED IS USED WHEN A FEMALE IS THE
OFFICER COMPARED TO:

Equal

Equal

Equal

Equal

Equal

Equal

Full

Male
Suspects

Female
Suspects

African
American
Suspects

Anglo
Suspects

Latino
Suspects

When the analysis was restricted to African American suspects (see Tables 10A, 10B, and
10C), we noted that both Anglo and Latino Officers were significantly more likely to use a CED
on suspects than African American Officers. This finding was present City-wide as well as in six
Districts (Anglo versus African American Officers) and four City Council Districts (Latino versus
African American Officers) (see Table 11). There were no differences in the probability of CED
usage among Anglo and Latino Officers.
If the analysis was restricted to Latino suspects, we noted virtually no differences among the
officers. Anglo, Latino and African American Officers were equally likely to use a CED on Latino
suspects. The only observation, and it is relatively modest, is that Latino Officers were more
likely to use a CED on suspects than African American Officers (but this is a weak finding Citywide and is significant only in two City Council Districts) (see Table 11).
The analysis also considered Anglo suspects. Differences were not identified among the
African American, Anglo, and Latino Officers in terms of their probability of using a CED on a
suspect. This observation holds up in all of the City Council Districts, with one very minor
exception.
Unlike the case for the Suspect data analysis, where the specter of omitted variable bias
required considerable caution in interpreting the results, here we had no such concerns. Given
the quasi-experimental nature of our analysis (similar context, with only officer race/ethnicity
varying), we were quite confident that these results would withstand any addition of omitted
variables. These results made clear that among the officers, there were virtually no differences
in terms of the probability of using their CED when the suspect was an Anglo or Latino (with the
minor exception that Latino Officers were slightly more likely to use a CED on Latino suspects
than African American Officers. In two Districts, Anglo Officers were less likely than Latino
Officers to use their CED when the suspect is a Latino). When the suspect was an African
American, African American Officers were significantly less likely to employ their CED than
Anglo or Latino Officers (who were equally likely to utilize their CED).
A final note on sample sensitivity is merited when discussing the Council District level analysis
provided in Tables 8 and 11. While this analysis is important in that it allows us to control
for one key contextual variable (i.e., the geographic location of the officer-suspect
interaction), the relatively small number of CED cases per district limits the accuracy of
the results.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

93

Geographic Component: Council Districts
Two City Council Districts stood apart from the rest when CED deployment was statistically
analyzed. CED use in Districts D and H was significantly greater than all other Districts with the
exception of District B. District B had a CED use that was significantly greater than Districts F
and G. Other significant differences did not exist.
Given the finding that African American suspects were significantly more likely to be subject to a
CED deployment than Anglo and Latino suspects, these results were not particularly surprising.
However, of some interest was the greater use of CEDs in District D compared to District B (as
both have comparable racial and socioeconomic demographics). It was not immediately clear
why District H had a high CED usage or why District I, with similar demographics to District H,
did not have a similar number of CED deployments.
If we focus on the three City Council Districts with the highest CED probabilities (D, H, and B),
we noted the following three sets of relationships (see Tables 8 and 11).
In terms of the suspect data (see Table 8), in District H, African American suspects were
significantly more likely to be the subject of a CED use than either Anglo or Latino suspects.
This observation was driven primarily by the greater tendency of a CED to be used on an
African American suspect when the officer was an Anglo. There were no significant differences
present when the officer was an African American or Latino.
In City Council Districts B and D, African American suspects were more likely to have a CED
used on them than Anglo suspects, but not Latino suspects. This significant relationship was
driven in part by the greater tendency of CED use when the suspect was an African American
and when the officer was a Latino.
For the officer data (see Table 11), we noted that in City Council Districts H and B that African
American Officers were less likely to deploy their CEDs than Anglo and Latino Officers. This
result was driven primarily by the much lower tendency of these African American Officers to
use the CED than their Anglo and Latino counterparts when the suspect was an African
American. In District D however, we did not observe any racial/ethnic differences among the
officers in CED use. There was a minor exception when the suspect was an African American.
African American Officers were noticeably less likely to use a CED than Latino Officers.
Overall, there appears to be some behavioral differences in City Council District D (e.g., those
HPD divisions that were dominant in these Districts) that were worthy of future investigation.
Summary
The absence of adequate data on suspect and officer characteristics limits the inferences that
can be made from the results of our suspect data analysis. Nonetheless, the results do
highlight several relationships between suspect race/ethnicity and CED use that merit further
scrutiny. The stronger research design and data utilized for the officer data analysis allows us
to draw inferences with much greater confidence. Lastly, while the Council District crosssectional analysis allows us to control for important contextual factors, it is crucial to remember
that the number of CED events in most of the nine Council Districts is sufficiently small so as to
warrant caution in our interpretation of the Council District analysis results.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

94

As mentioned previously, given the data limitations, considerable caution must be
exercised in the interpretation of the suspect related results. The data analysis above
however allowed three observations:
•
•

•

African American suspects were significantly more likely to have the CED used on them
than Anglo and Latino suspects.
African American Officers were significantly less likely to use their CED than Anglo and
Latino Officers. The explanation for this observation was most likely hinges on a
complex set of factors related to the way in which the suspect interacted/responded to
the officer and in which the officer interacted/responded to the suspect.
Latino suspects were somewhat more likely to be subjected to a CED deployment than
Anglo suspects. This difference was modest, and driven primarily by the greater
tendency of Latino Officers to utilize their CED when a suspect was Latino, compared to
when the suspect was an Anglo.

Recommendation: Diverse Patrol Experiment
It is clear that a complex set of factors has yet to be investigated. Among these variables are
measures (to be developed) that capture the threat that officers face, the general context in
which the CED incident occurs, as well as the relation between an officer's productivity, arrest
history, and their use of CEDs. In order to obtain a more thorough and complete understanding
of the dynamics of these new variables, it would be advisable to conduct a series of natural
experiments. These natural experiments would be designed to evaluate, for example, the role
of officer and suspect race and ethnicity in the probability that a CED incident occurs.
Injury Analysis
The analysis of the injuries to HPD Officers was conducted from workers’ compensation claims
data. Time series intervention analysis was used where applicable. Recall that the variables
examined (see tables 2a and 2b) included the following:
•
•
•
•
•
•
•
•
•
•
•
•

Physical altercation (variable name: Altercation)
Foot pursuit that ends in physical altercation (variable name: Pursuit)
Total amount of physical altercations (variable name: Total Comp)
Cost due to physical altercation (variable name: Altercation$)
Cost due to foot pursuit that ends in physical altercation (variable name: Pursuit$)
Total cost of physical altercations (variable name: Total$)
Lost days due to physical altercation (variable name: Altercation Days Lost)
Lost days due to foot pursuit that ends in physical altercation (variable name: Foot
Days Lost)
Lost time due to physical altercation (variable name: Altercation Lost Time) which is
equivalent to the number of filed claims.
Lost time due to foot pursuit that ends in physical altercation (variable name: Foot
Lost Time) which is equivalent to the number of filed claims.
Total amount of lost days due to physical altercations (variable name: Total Days
Lost)
Total amount of lost time due to physical altercations (variable name: Total Lost
Time) which is equivalent to the number of filed claims.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

95

Andrews Test Results
The results are summarized in Table 13, and Figures 7 and 8. There were two variables that
shifted (structural breaks). The cost due to physical alternation (Altercation$) had a shift) in July
2002 (p-value < .01) and the total cost of physical altercations (Total$) had a break in April 2003
(p-value < .01).
TABLE 13. ANDREWS TESTS FOR STRUCTURAL BREAKS IN INJURY DATA: JANUARY
2000 TO JUNE 2007

Model
Altercation
Altercation$

Maximum Test
Statistic

Month of
Maximum

0.966

June 2006

25%

14.08*

July 2003

25%

Trimming

Pursuit

3.825

March 2002

25%

Pursuit$

5.58

October 2004

25%

Total Comp

2.97

May 2003

25%

20.95*

April 2003

25%

Foot Days Lost

5.60

October 2004

25%

Foot Lost Time

2.05

January 2003

25%

Altercation Days Lost

3.45

August 2002

25%

Altercation Lost Time

4.75

August 2004

25%

Total Days Lost

4.44

August 2002

25%

Total Lost Time

2.23

May 2005

25%

Total$

Notes: N = 90 months.
The Chi-squared critical value for testing for a break in a single parameter with 25 percent
trimming is 11.48 at the 1% level (*). (See Andrews 1993, Table 1).

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

96

FIGURE 7

Andrews Test for Maximum Structural Break:
Expenditures on Injury Claims Due to Aggressive Acts (non-foot pursuit)
Wald Test
Statistics

16~--------------------,

14
12
10
8
6

4

2

2000 2001

2002 2003 2004 2005 2006 2007
Year

UNIVERSITY OF HOUSTON

cenl:;PUBlICPOlICY

97

FIGURE 8

Andrews Test for Maximum Structural Break:
Total Expenditures on Injury Claims Due to Aggressive Acts
Wald Test
Statistics

24----------------------.
20
16
12
8
4

2000 2001

2002 2003 2004 2005 2006 2007
Year

UNIVERSITY OF HOUSTON

cenl:;PUBlICPOlICY

98

The Rolling Paired t-test Results
The rolling paired t-test (summarized in Table 14) results indicated that initial shifts occurred in
Altercation$ (April 2004); foot pursuits that ended in physical altercations (Pursuit-January
2003); the total amount of physical altercations (Total Comp-July 2004); Total$ (May 2003) (see
Figure 9); and the total amount of lost days due to physical altercations (Total Days Lost-July
2005) (see Figure 10).
In addition, shifts occurred for the following variables on January 2002: lost days and lost time
due to physical altercations (Pursuit$, Foot Lost Time, Altercation Lost Time, and Total Lost
Time).
TABLE 14. ROLLING PAIRED t-tests FOR STRUCTURAL BREAKS IN INJURY DATA:
JANUARY 2000 TO JUNE 2007
t-statistic
(maximum)

Month of
Maximum

Initial
Effect

Altercation

1.36

May 2003

NA

NA

Altercation$

2.65**

August 2004

NA

NA

Pursuit

3.56**

July 2005

NA

NA

Pursuit$

3.15*

July 2005

NA

NA

Total Comp

3.07**

July 2005

Total$

3.53**

September 2004

Foot Days Lost

1.98*

July 2005

NA

NA

Foot Lost Time

2.94**

December 2003

NA

NA

Altercation Days
Lost

1.48

August 2004

NA

NA

Altercation Lost
Time

2.84**

September 2002

NA

NA

Total Days Lost

2.21*

July 2005

NA

NA

Total Lost Time

3.51**

August 2003

NA

NA

Model

Notes: N = 90 months. * indicates p < .05, ** indicates p < .01.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

99

(2.63)
($38,081)

Cumulative

(3.24)
($50,774)

FIGURE 9

Rolling Break Point t-tests:
Total Expenditures on Injury Claims Due to Aggressive Acts
Wald Test
Statistics

3.6
3.2
2.8
2.4
2.0
1.6
1.2
0.8
0.4
0.0
2000

2001

2002
Year

UNIVERSITY OF HOUSTON

cenl:;PUBlICPOlICY

100

2003

2004

FIGURE 10

Rolling Break Point t-tests:
Total Days Lost Due to Aggressive Acts
t-statistics

1.6

1.2

0.8

0.4

2000

2001

2002

2003

2004

Year

Intervention Results

The intervention analysis for using the dates for the Andrews test indicated no effect. However,
the rolling paired dates resulted in the following cumulative policy effects for particular variables.
As reported in Table 14, the variables where a cumulative effect occurs were Total Comp and
Total$. In the case of Total Comp, there was a drop in the level of monthly claims (starting in
July 2004) to -2.63 and that over time dropped to -3.24. The reduction in the intercept was
approximately 18% (3.24/18.27). For Total$ the initial reduction in the level, starting in May
2003 was $38,081 which accumulates to a reduction of $50,774. This equals a total
expenditure of 49% ($50,774/$102,635).

UNIVERSITY OF HOUSTON

cenl:;PUBlICPOlICY

101

Summary
The results on injury analysis indicated there have been shifts in a variety of injury indicators.
There have been substantial reductions in the number of compensation claims as well as the
total expenditures. These reductions began prior to the institution of the CED policy and have
continued through the scope period. While the CED policy cannot be the initial cause for the
change and the scope period with CED capability is short, with the passage of time, it will be
possible to see the full statistical effect of the CED policy.
Substitution Analysis
Here we present the results of the HPD discharge of firearms for the period January 2000 to
June 2007. Recall that our analysis for weapon substitution effects includes the following
variables:
•
•
•
•
•
•
•
•
•

Discharge of firearms that was an accident (variable name: Accident)
Discharge of firearms that resulted in a citizen’s death (variable name: Citizen Death)
Discharge of firearms that resulted in a citizen’s injury (variable name: Citizen Injury)
Discharge of firearms that resulted in a citizen’s death and injury (variable name:
Citizen Death/Injury)
Discharge of firearms that resulted in an officer’s death (variable name: Officer
Death)
Discharge of firearms that resulted in an officer’s injury (variable name: Officer Injury)
Discharge of firearms that resulted in an officer’s death and injury (variable name:
Officer Death/Injury)
Discharge of firearms that resulted in property damage (variable name: Property
Damage)
Discharge of firearms - total from categories above (variable name: Total/No Animal).

Andrews Test Results
The results are summarized in Table 15. For the Andrews test, the only variable that showed a
shift was citizen’s death. As Table 14 indicates, the break in the series occurs in September
2004 (p-value < .01).

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

102

TABLE 15. ANDREWS TESTS FOR STRUCTURAL BREAKS IN SUBSTITUTION DATA: JANUARY
2000 TO JUNE 2007
Model

Maximum Test Statistic Month of Maximum

Accident

Trimming

3.60

January 2003

37.5%

Citizen Death

11.97*

September 2004

25%

Citizen Injury

0.76

February 2003

25%

Citizen Death Injury

6.10

September 2004

25%

Officer Death

NA

NA

NA

Officer Death Injury

NA

NA

NA

Property Damage

NA

NA

NA

Total/No Animal

3.10

December 2004

25%

Notes: N = 90 months.
The Chi-squared critical value for testing for a break in a single parameter with 25 percent
trimming is 11.48 at the 1% level (*). (See Andrews 1993, Table 1).
The Rolling Paired t-test Results
As reported in Table 16, the variables with shifts using this test were: the discharge of firearms
that was an accident (Accident-January 2002), the discharge of firearms that resulted in a
citizen’s death (Citizen Death-August 2002), the discharge of firearms that resulted in an
officer’s death (Officer Death-January 2002), and the total number of events (Total-November
2003).

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

103

TABLE 16. ROLLING t-tests FOR STRUCTURAL BREAKS IN SUBSTITUTION DATA: JANUARY
2000 TO JUNE 2007
Model

t-statistic (maximum) Month of Maximum Initial Effect Cumulative

Accident

4.77**

September 2003

-.50

---

Citizen Death

2.39*

January 2005

.31

---

Citizen Injury

1.46

November 2003

NA

NA

Citizen Death Injury

1.29

June 2005

NA

NA

Officer Death

2.38*

August 2002

NA

NA

Officer Death Injury

1.67

October 2002

NA

NA

Property Damage

3.24**

December 2004

NA

NA

Total/No Animal

2.10*

January 2004

NA

NA

Notes: N = 90 months. * indicates p < .05, ** indicates p < .01.
Intervention Results
There was no evidence of persistence in any of these variables. However, it was possible to
find evidence of initial effects. The Accident variable showed a drop in its mean of -.50. This
constituted a reduction of 59% (-.50/.85). On the other hand, there was evidence of an increase
in the level of Citizen Death, an increase from .42 to .73 or 74%.
Summary
As with the injury analysis, we noted that shifts or structural breaks in the data occurred prior to
the introduction of the CED policy. This was not entirely surprising since the only data available
that could be associated with weapons pertains to firearms. Note also that the CED policy has
been in existence for such a short duration. Over a period of time, it will be possible to see the
full statistical effect of the CED policy. A more direct test for substitution effects would involve
the use of batons or flashlights, which are more readily associated with the intermediate weapon
status a CED possesses. These tests, along with the addition of an extended time period would
improve the overall research design.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

104

Complaint Analysis
Between December 2004 and June 2007, there were 55 complaints filed against HPD Officers
where CEDs were mentioned in the complaint. A reading of these complaints (see Table 4 for a
summary) indicated the following:
Gender Breakdown
•
•

Of the 59 officers noted in the 55 complaints, 97% of the complaints were leveled at
male officers while 3% of the complaints were directed at female officers.
76% of the complaints were made by males and 24% were made by females.

Racial Breakdown of Officers (59)
•
•
•
•

27 or 46% of the complaints were directed at Anglo Officers
20 or 34% of the complaints were directed at African American Officers
9 or 15% of the complaints were directed at Latino Officers
3 or 5% complaints were directed at Asian Officers

Racial Breakdown of Complainants (51)
•
•
•
•

7 or 14% of the complaints were made by an Anglo
36 or 71% of the complaints were made by an African American
7 or 13% of the complaints were made by a Latino
1 or 2% of the complaints were made by an Asian

Racial/Gender Breakdown of Officers and Complainants
•

•

•

•
•

Of the 23 complaints leveled at Anglo male Officers, 3 were made by Anglo males,
14 by African American males, 2 by African American females, 2 by Latino males,
and 2 by Latino females.
Of the 14 complaints leveled at African American male Officers, 1 was made by an
Anglo male, 8 by African American males, 1 by a Latino male, 1 by an Anglo female,
and 3 by African American females.
Of the 9 recorded complaints leveled at Latino male Officers, 2 were made by Anglo
males, 1 by an Asian female, 4 by African American males, 1 by an African American
female, and 1 by a Latino female.
The 3 complaints leveled at Asian male Officers were made by 1 African American
male, 1 African American female, and 1 Asian female.
The 2 complaints leveled at African American female Officers were made by 1
African American male and 1 African American female.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

105

Disposition of the Complaints
Of the 55 complaints noted in HPD documentation, 3 were sustained and 9 were not. In the
remaining cases investigated, the following outcomes occurred: the officer was exonerated (13);
there was no evidence or insufficient evidence to prove the incident occurred (1); the complaint
was never formalized (2); the CED was not the focus of the complaint and the investigation
found the CED usage appropriate (12); and the allegation was false or not factual (11).
Four cases remained open.
Summary
In summary, this statistical analysis identified racial and gender differences in the breakdown of
both officers and complainants involved with the use of CEDs. Anglo and African American
male Officers had the most complaints filed against them. In particular, the mode (most
frequent category) was Anglo Officers that had a complaint filed by an African American
complainant.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

106

DATA MANAGEMENT
The process for collecting the data raised some important issues that affected the analysis. The
research team found the personnel of HPD to be fully cooperative in all requests for information.
However, there were delays in acquiring and assembling the data as the data collection process
was underway. Much of the delay was attributed to an outdated database system so it is
important to keep in mind that HPD’s planned transition to a new system should alleviate some
of the problems.
For example, HPD fielded approximately 1.4 million calls for service/incident reports that were
recorded in multiple databases during the Scope period. Approximately 48% of the 1.4 million
electronic police call for service/incident reports did not contain suspect information (e.g., the
incident was reported after the suspects had long left the scene of the incident, no suspect was
involved in the incident, and/or no information on the suspect was collected). The addition of
key explanatory variables (suspect race/ethnicity, Uniform Crime Report (UCR) code, zip code
of incident location, City Council District of incident location) resulted in the exclusion of
approximately 110,000 cases while the lack of officer data for an incident led to the exclusion of
approximately 50,000 additional cases. This left the Audit Team with a final analysis population
of approximately 570,000 merged records (the Analysis Population).
The original electronic data was of poor quality, incomplete, inconsistent, and retrieval was
difficult. The physical size (weight, height) of the suspect was often not recorded in either the
electronic or hardcopy reports and if it was we noted that the majority of the suspects were 175
pounds. Included in the approximately 700,000 calls for service/incident reports were 1,284
incidents where a CED was deployed. Only 951 (75%) of the 1,284 CED deployments could be
statistically analyzed primarily because of the data merging challenges. In summary, the Audit
Team reviewed all 1,284 of the hardcopy CED calls for service/incident reports; however, they
could not include CED incidents that were lost during the electronic data merges without biasing
the results of the analysis.
There appeared to be two organizational barriers to data management at HPD. The current
data management process was fragmented or de-centralized. Rather than storing data in one
central location with the use of common software and universally defined units of analysis, there
was more than one sub-organization that had its own method of data management. Data
management knowledge was concentrated in too few individuals.
The lack of coordination among the various data management units within HPD combined with
insufficient knowledge diversification within each unit results in the following process and
delivery:
•

•

•

Due to incompatible and sometimes outdated software, HPD analysts must
frequently engage in inefficient data acquisition practices (i.e., open and close
various programs and manually write data outcomes).
Those who ask for data (such as the Audit Team) are faced with delays in data
merges, including further requests for clarification due to coding error, coding
ambiguity, and missing data from the original source.
Several HPD call for service/incident databases were developed in COBOL, a
seldom used computer language in modern systems. As a result HPD is at risk of
inadequate staffing resources in the event of staff turnover. COBOL is a computer
language that is generally neither mandated nor offered in current college
curriculums.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

107

The impairment of process and delivery is compounded by the additional factor that HPD
appears to be understaffed in the data management area. In addition to requests for data,
particularly large data requests such as this, there are daily requests that are a function of Open
Records requirements.
Summary
The impairment of process and delivery was compounded by the additional factor that HPD
appeared to be understaffed in the data management area. In addition to requests for data,
particularly large data requests such as the ones for this CED audit, there were numerous Open
Records requests.
The current structure for data management is organizationally deficient and under-staffed. This
combination of factors may produce inefficiencies in data transmission, increases in
measurement and coding errors, and an overall inability to create a template for connecting
disparate pieces of information to support overall HPD Management processes. The
implications are even more severe however, if there are efforts to increase situational
awareness for HPD officers that require data in real time.
Recommendation: An Audit on Data Management Processes
To end delays in data dissemination and to provide a process to enhance the forthcoming
modernization in data management, we believe a process audit is imperative. The current
structure for data management at HPD seriously impairs efficient data processing and data
acquisition. The audit would seek to merge efficient processes, remove impediments to efficient
processing, and combine these methods with the new data processing capabilities that are now
being constructed.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

108

CONCLUSION
Before summarizing the statistical conclusions, it bears repeating that this study faced some
important data limitations. The limitations meant that a variety of alternative explanations
have yet to be evaluated and important statistical controls have not been included. The fact
that the CED policy has been in existence for such a brief period means the passage of time
could lead to new results and conclusions.
With these caveats in mind we provide some summary thoughts. In regard to incidence, the
results from the CED analysis suggest that certain combinations of officer and suspect
characteristics resulted in an increased probability of CED utilization. Depending on how
the race of the officer and the race of the suspect were paired, it was possible to see
significant increases and decreases in the rate of CED utilization. Although these City level
results were robust to numerous statistical controls, it was important to note that we
observed interesting deviations from these general patterns when we conducted our
analysis at the City Council District level. The results for injuries and substitution indicated
that nearly all statistically discernible changes occurred prior to the implementation of the
CED policy. We also noted that the data for injuries and substitution had little persistence in
their behavior. The effects of policy changes for the most part occurred quickly.
The complaint analysis indicated that complaints in which a CED was mentioned did have a
distinctive racial and gender propensity. More males and African Americans filed
complaints. The mode (most frequent category) of HPD officers receiving complaints were
Anglo males. We also noted that few complaints have been sustained.
A final observation centers on data management. Despite cooperation by HPD in providing
the data, we found that there were delays that can be traced to organizational rigidity and a
general lack of staffing.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

109

REFERENCES
Andrews, D. W. K. 1993. “Tests for Parameter Instability and Structural Change with
Unknown Change Point.” Econometrica 61, 4 (July): 821-856.
Cureton, E. E. 1957. “Further Notes on the Two-Group t Test.” The American
Statistician 11, 4: 21.
Goldfeld, S. M. and R. E. Quandt. 1973. “The Estimation of Structural Shifts by
Switching Regressions.” Annals of Economic and Social Measurement 2:
475-485.
Hinkley, D. V. 1971. “Inference in Two-Phase Regression.” Journal of the American
Statistical Association 66: 736-743.
Manski, C. F. 2006. “Profiling: Introduction to the Feature.” The Economic Journal
116, 515: F347-F450.
McGee, V. E., and W. T. Carlton. 1970. “Piecewise Regression.” Journal of the
American Statistical Association 65, 1109-1124.
Quandt, R. E. 1958. “The Estimation of the Parameters of a Linear Regression System
Obeying Two Separate Regimes.” Journal of the American Statistical Association
53: 873-880.
Shively, W. P. 2008. The Craft of Political Research, 7th Edition. Upper Saddle River,
New Jersey: Pearson Prentice Hall.
Tufte, E. R. 2001. The Visual Display of Quantitative Information. Cheshire,
Connecticut: Graphics Press.

UNIVUSITY OF HOU5TON

('"~:PUBlICPOLICY

110

Sam Houston State University,
A Member of the Texas State University System
College of Criminal Justice

PART IV
A QUALITATIVE ANALYSIS OF THE
USE OF CONDUCTED ENERGY DEVICES BY THE HOUSTON
POLICE DEPARTMENT

PART IV

SAM HOUSTON STATE UNIVERSITY,
A QUALITATIVE ANALYSIS OF THE USE OF CONDUCTED ENERGY DEVICES
BY THE HOUSTON POLICE DEPARTMENT
TABLE OF CONTENTS
Page
Summary

111

Background

114

Study Components and Method of Analysis

118

Results of Compliance Review of Houston Police
Department Incident Reports

120

Houston Police Department Incident Report Review

132

Houston Police Department Training Academy - CED Training

139

Comments on UH CPP Team Statistical Analysis of the Use of CEDs

143

Recommendations

147

SUMMARY
The study period was from the implementation of the Conducted Energy Device (CED)
program, December 4, 2004, through June 30, 2007. Using qualitative analysis and
observation, the research group from Sam Houston State University College of Criminal
Justice (SHSU CCJ)1 studied the following issues in relation to the use of CEDs by the
Houston Police Department (HPD).
•

•
•
•

•

•

•
•

When officers deployed a CED, were they responding to a call for service for which
they were dispatched, or did the officer self-initiate the call by making a traffic stop,
or otherwise encountering the suspect?
What was the nature of the original call for service? In other words, why was the
officer there?
When an officer deployed the CED, how many cycles were used?
In what percentage of the cases in which an officer deployed the CED, was it noted
that the suspect on whom the CED was deployed appeared to have used alcohol
and/or drugs?
What appeared to be the effectiveness of the deployment of the CED in controlling
the suspect’s behavior? When the device appeared to be ineffective, what was the
apparent reason?
Were officers in compliance with HPD policy in the deployment of the CED and the
reporting of the incident? If not, were factors present which made the deployment
appropriate despite the apparent failure to comply with policy?
Were there cases in which the CED was deployed in which the officer would have
been justified in using deadly force?
During the training of cadets on the use of CED, what was observed that might be
affecting the manner in which officers deploy the CED?

Initiation of Call for Service
Fifty-five percent (55%) of the time, in the incidents which resulted in the deployment of a
CED, the officer had responded to the scene as a result of being dispatched. In 33% of the
cases, the contact with the suspect against whom the CED was eventually deployed, was
initiated by the officer. In 12% of the cases, the initial reason for the contact with the
suspect was a traffic issue. A majority of the self-initiated contacts were drug related.

1

The principal investigator from Sam Houston State University was Margo L. Frasier, J.D. Assisting was Jennifer
Schulenberg, Ph.D., and Brian Lawton, Ph.D.

111

Nature of Call for Service
The two most frequent types of calls for service were drug related (15%)2 and disturbance
related (15%).3 The incidents involving drugs were notable in that, with a few exceptions,
they were calls which were self-initiated by an officer.4 Assault was the basis for 13% of the
calls. In addition to the 13% of the calls which were assault based, 7% of the calls were
based on a report of family violence. Criminal trespass was the basis of 6% of the calls.
Traffic offenses were the basis for 11% of the calls. Other misdemeanors such as Driving
While Intoxicated, Theft, and Criminal Mischief comprised the basis of 13% of the calls.
Other felonies comprised the basis of 9% of the calls. Approximately one-third of the
felonies were related to automobile theft. Eleven percent of the calls were based on the
need for crisis intervention by officers with a person suspected of having a mental health
issue.
Number of CED Cycles Used When Deployed
A standard cycle of the CED is five seconds. In 38% of the incidents, the officer only
deployed one cycle. In 31% of the incidents, the officer deployed two cycles. In 15% of the
incidents, three cycles. In 5% of the incidents, the CED was cycled four times. In 3% of the
cases, the CED was cycled five times. In 5% of the cases, the CED was cycled between six
and ten times. In 3% of the cases, the CED was cycled over ten times.
Impact of Alcohol and/or Drugs Consumed by Suspects
In 58% of the incidents, officers did not note that the suspect appeared to be under the
influence of drugs and/or alcohol. The researcher did not attempt to determine whether the
officer, if asked, would have said that drugs and/or alcohol might have been a contributing
factor to the behavior which resulted in the need for CED deployment. However, it is highly
likely that a significant percentage of these suspects were under the influence of drugs or
alcohol given the behavior that was noted by the officer.
In 15% of the incidents, the officers noted in their reports that the suspect appeared to be
under the influence of alcohol. In 27% of the incidents, the officers noted in their reports that
the suspect appeared to be under the influence of drugs and/or possessing drugs.

2

For the purposes of this study, drug related calls include the possession of and/or sale of drugs.
For the purposes of this study, disturbance related calls include the crimes normally associated with disorderly
conduct and public intoxication.
4
The fact that a call was self-initiated by an officer does not mean that citizen complaints were not the basis for
officers being in the area.
3

112

Effectiveness of CED in Controlling a Suspect
The device was reported as being effective in assisting in controlling the suspect in 77% of the
incidents. It should be noted that this does not mean that the device was effective in its initial
deployment or effective in controlling the suspect’s behavior with only one cycle of the CED. In
some cases, the officer deployed the device in probe mode and then resorted to using the CED
in the drive stun mode to gain control. In other cases, the officer used multiple cycles in the
probe mode to gain control of the suspect.
In 4% of the incidents, the report indicated that the CED was not effective due to the clothing the
person was wearing. In 4% of the incidents, the officer missed the suspect with one or more
darts. In 2% of the cases, it was reported that the darts hit the suspect, but the suspect pulled
out the darts. In 2% of the incidents, it was reported that the CED was defective and did not fire
the probes. In an additional 11% of the incidents, the officer indicated that the CED was not
effective without attributing the ineffectiveness to any of the reasons previously identified. In the
majority of the cases where the officer indicated that the CED was not effective, the officer
reported that the suspect appeared to be under the influence of alcohol and/or drugs or
experiencing some form of mental health crisis.
HPD CED Policies
HPD CED policies, in general, followed the best practices of the law enforcement
profession. Additionally, the majority of the reports indicate that the officer used the CED in
compliance with HPD CED policies. Any concerns about the use of CEDs did not appear to be
a result of any shortcoming with HPD CED policies. Instead, as detailed later, the concerns
appear to be as a result of training and/or implementation of HPD CED policies.
Use of CED Instead of Deadly Force
The results of the study by the University of Houston Center for Public Policy (UH CPP) did not
indicate a reduction in the number of officer involved shootings since the introduction of the
CED. This was not surprising as the opportunities to use a CED instead of a firearm are very
limited. For example, it would be highly unusual for an officer to rely on a CED if the suspect
was armed with a firearm.
However, there were incident reports which indicate that officers would have been legally
justified in using deadly force and the officers chose, instead, to deploy the CED. It should be
noted that just because the officer would have been legally justified in using deadly force, it
does not mean that the officer would definitely have chosen this option if not for being CED
equipped. However, a review of the reports indicated situations in which other intermediate
weapons were unlikely to have been used due to officer safety concerns; thus, if the CED had
not been available, the use of deadly force would have been more likely.
HPD Training Academy - CED Training
There were numerous issues involved in the training of cadets on CED use. Many of the
issues, which are detailed later in this report, center around the lack of emphasis on the
potential danger to a suspect if numerous cycles of a CED are deployed; particularly on a
suspect who might be at risk for excited delirium. Additional concerns center around the lack of
training in a manner which more accurately reflects the situations described in the actual
incident reports of HPD officers. Also, while the cadets went through the motion of deploying
the CED, demonstration of true proficiency was not required of the cadets.

113

BACKGROUND
Introduction
Law enforcement officers perform their duties, day and night, without much notice from the
public. The general public gives little thought to how officers go about maintaining order and
the challenges officers face in pursuing that goal.
Further, there may be a disconnect between pubic perception and legal reality. Statutes
and court decisions direct officers that they may use the amount of force that is reasonably
necessary to control a situation given the facts and circumstances facing the officers.5 The
public, however, tends to believe that officers should use the absolute minimum amount of
force necessary to control the situation. The criticism leveled at officers often centers
around a perception that officers should attempt the lowest level of force, and, if not
successful, progress to the next higher level of force. Only if, each level of force proves
unsuccessful should the officer, in the public’s opinion, resort to a greater level of force.
Meanwhile, officers are trained to use a level of force which is reasonable under the totality
of the circumstances. This does not necessarily translate to an officer using the least
amount of force justified. In fact, officers are trained to use the level of force that, while
reasonable, will control the situation.
These observations were not meant to be critical of the public or law enforcement. They
were meant to portray the reality in which use-of-force issues were dealt with by law
enforcement, viewed by the public, and portrayed by the media.
There are two circumstances in which the public tends to pay particularly close attention to
the use-of-force by law enforcement officers. The first is when a law enforcement officer
uses deadly force. When a law enforcement officer uses force that results in a death, the
public wants to know “why.” Why didn’t the officer wound the suspect instead of killing
him/her? Why didn’t the officer use less than deadly force to resolve the situation?
The second circumstance is when the public, or a segment of the public, perceives that law
enforcement officers have used force in a disproportionate or discriminatory manner towards
one segment of the community. Once again the public asks questions. Would the situation
have been handled differently if the suspect had been of a different race or ethnicity? Was
more force used because of who was involved as opposed to what the person did to warrant
law enforcement action?

5

Texas Penal Code 9.32

114

When CEDs are considered, both circumstances may come into play. CEDs are viewed by
a majority of the general public as an alternative to deadly force.6 When an officer has
resorted to the use of a firearm resulting in the death of a suspect, often one of the first
questions asked was, “Why didn’t the officer use a Taser?”7 On the other hand, when a
CED was used to control a situation, the public sometimes asked, “Why did the officer have
to use a Taser?” HPD’s utilization of CEDs has been called into question by some;
particularly concerning the issue of whether certain racial minority groups have had CEDs
deployed against them in a disproportionate and/or discriminatory manner. Also, since the
introduction of CEDs at HPD, when officers have used deadly force, one of the first
questions posed is, “Why a CED was not used instead of deadly force?”
Basics of Conducted Energy Devices
Most of situations confronted by law enforcement officers are controlled simply by the
presence of officers. Only 1% of the contacts between law enforcement and the public,
nationwide, result in a use-of-force at a level higher than a verbal command and
handcuffing.8 For circumstances in which physical force is required, law enforcement
agencies have long sought tools to assist the officer in controlling the situation.
Traditionally, officers were equipped with a handgun and some type of striking instrument.
With the development of chemical munitions for non-military purposes, law enforcement
agencies began using tear gas to disperse crowds during protests. During the 1960s, there
was an emphasis on the advancement of weapons which would give an officer an
alternative to the lethal force option of using a firearm. The modern day results are pepper
spray, which causes irritation to the suspect, and the expandable baton. Some officers have
found pepper spray to be problematic due to the fact that the situations are limited in which
its use is appropriate and the officer often becomes contaminated by the residue of the
pepper spray when handling a suspect who has been sprayed. Additionally, there are
individuals who appear to be unaffected by pepper spray and others who appear to have an
adverse reaction to the spray. Batons, expandable or fixed, are a useful device, but require
the officer to be in very close proximity to the suspect in order for the device to be used.
Additionally, a suspect struck with a baton will often suffer some level of injury. While rarer
than reports of ineffectiveness of pepper spray, there are reports of some individuals being
virtually unaffected by being struck with a baton. For many law enforcement agencies,
utilization of CEDs is just another step in the development and use of what are commonly
referred to as intermediate weapons.

6

The Annual Houston Area Survey found that 60% of the public thought equipping law enforcement with CEDs made
it less likely they would use deadly force. www.houstonareasurvey.org.
7
While “Taser” is the name of a particular brand of conducted energy devices, the public has come to refer to CEDs
generally as Tasers. In fact, during the training at HPD, officers are instructed to announce “Taser, Taser, Taser”
when deploying a CED as the suspect might be confused if the terms “CED” or “conducted energy device” were used
instead.
8
U.S. Bureau of Justice Statistics (1999). Contacts between police and the public: Findings from the 1999 National
Survey, Washington D.C.; U.S. Department of Justice.

115

Approximately thirty-five years ago, NASA scientist Jack Cover developed the TASER, an
acronym for a device used by the hero in the 1911 fictional adventure story, “Thomas A.
Swift’s Electric Rifle.” The development of this technology went basically unnoticed for the
first decade. While some members of the law enforcement community used CEDs during
the 1980s and 1990s, the practice was confined primarily to the West Coast. Most likely,
the public would have been unaware of these devices prior to 2000 had it not been for the
infamous Rodney King incident which brought widespread public attention to the Los
Angeles Police Department’s use of a CED. On television screens across America, the
public watched as law enforcement officers delivered what appeared, to many, as never
ending strikes with batons and jolts with a CED to Rodney King while he was on the ground
surrounded by a group of officers. Whether one believed the use-of-force to be appropriate
or not, for most of the public, it was the first time they had seen a CED used on a suspect.
In 1999, a CED, known as the TASER was produced and introduced by TASER,
International. The TASER administers an electric charge that causes muscular dysfunction
and temporarily incapacitates a person; particularly if used in “probe” mode. If a CED is
used in the “drive stun” mode,9 it is regarded as a compliance tool as it often does not cause
incapacitation, but does cause pain and discomfort. If officers use the device in the drive
stun mode, they place it against the person’s body and pull the trigger to deliver the shock.
The term “drive” refers to the method of delivery as the officer is instructed to place the CED
against the suspect’s body and use pressure to maintain contact when the suspect moves
away from the CED, whether intentionally or involuntarily. If officers deploy the CED in the
probe mode, the trigger is pulled and a cartridge fires two darts. The CED then delivers
50,000 volts of electric current through the filament line which is attached to the darts. In
both of these modes, the pulling and releasing of the trigger results in a standard five
second cycle. However, if the officer does not release the trigger, the cycle will continue
until the trigger is released. If the officer wishes to shorten the five second cycle, the device
must be turned off.
The incapacitation of the suspect is due to the electrical charge overriding the central
nervous system. The idea is that the suspect simply cannot control his/her neuromuscular
system and that the incapacitation is not affected by the suspect’s mental state or use of
drugs and/or alcohol. However, the incapacitation ceases as soon as the electrical charge
ceases. Thus, at the end of the cycle, the suspect must either voluntarily submit to control
or the officer must use the incapacitation period as an opportunity to gain control.
Additionally, as will be seen later in the analysis, the CED does not appear to affect some
individuals and there are times when one or both probes miss,10 the probes become
dislodged either due to the suspect pulling them out or due to a struggle, or the probes do
not penetrate the suspect’s skin.11

9

Some agencies refer to it as the “dry stun” mode. HPD uses the term “drive stun” or “drive” so that is the term that
will be used herein.
10
Both probes must make contact in order to complete the cycle so that the voltage is delivered. If only one probe
makes contact, the circuit may be completed by placing the CED electrodes against the suspect (Drive Stun Mode).
However, this method requires close contact with the suspect.
11
While it is not necessary for the probes to penetrate the suspect skin for the voltage to be delivered, experience
shows that the CED is less effective if at least one probe does not penetrate the skin.

116

HPD’s Use of CEDs
A study conducted by the Texas Law Enforcement Management and Administrative
Statistics Program of the SHSU CCJ in 2005, found that 54% of the Texas law enforcement
agencies who were surveyed were currently utilizing CEDs and an additional 10% were
considering the use of CEDs.12 Notably, the study found that the Texas Department of Public
Safety did not equip its troopers with CEDs and did not have plans to do so. Currently,
nationwide, over 8,000 law enforcement agencies utilize CEDs in some manner.
A review of newspaper accounts reflect that many Texas law enforcement agencies made
the decision to add CEDs to their officers’ equipment belts as a result of a particular incident
in which an officer used deadly force and the public questioned whether the use of an
intermediate weapon, including a CED, would have resulted in a life being spared. Houston
was no different. The timing of the acquisition of CEDs in relation to an incident where
deadly force was used tends to reinforce the public’s perception that CEDs are an
alternative to deadly force as opposed to simply being one of the intermediate, or less lethal,
weapons available to an officer.
In December, 2004, HPD began issuing CEDs to its officers. By March, 2005, all HPD
street officers were being issued a CED upon completion of a training course.13 Now, all
new cadets are issued a CED during their basic training at the HPD Training Academy
(Academy) upon the completion of a CED training course. Presently, with a few exceptions,
all Officers have been trained in the use of CEDs and authorized to deploy CEDs. HPD
policy requires that officers, who are issued a CED, carry the CED when in uniform.
Questions regarding the use of CEDs arose shortly after their use by HPD began. One
issue was whether CEDs were being used disproportionately and/or in a discriminatory
manner against certain groups; particularly, African American males. As part of the Fiscal
Year 2007 Audit Plan submitted by City Controller, Annise D. Parker, in August, 2006, the
City Controller informed the Mayor and City Council that her Audit Division intended to
perform an independent assessment of CED use by HPD. In November, 2006, Mayor Bill
White also called for an independent analysis into the use of CEDs. In March 2007, the City
Controller and Mir•Fox & Rodriguez, P.C. (MFR) assembled a team to conduct a
Performance Audit of the use of CEDs. A group from the SHSU CCJ responded to a
request for a proposal to assist in the conduct of the audit.

12
13

Frasier, M.L. (2005). The use of conducted energy devices (Tasers). Telemasp Bulletin, 12 (6), 2.
Originally, the training program was four hours in length; now it is six hours.

117

STUDY COMPONENTS AND METHOD OF ANALYSIS
Policy Compliance Review
The review for policy compliance took several steps. First, past, and current policies were
reviewed in light of what is considered “best practices” by the law enforcement profession.
One of the difficulties encountered was that HPD’s policy on CED use had undergone
several changes since its original development and implementation. Some of the initial
changes in the policy were the result of HPD hosting a symposium on the use of CEDs in
conjunction with the Police Executive Research Forum (PERF) in 2005. Other changes
were a reflection of experience. In other words, as CEDs were introduced and used, a
decision was made that the policy needed to be changed in some manner. HPD continued
to review CED use with the most recent changes taking place in March 2007.
Second, each CED incident report was analyzed in light of the policy in effect at the time to
determine whether the officers followed HPD policy. Additionally, particular attention was
paid to any incident of non-compliance to determine whether the non-compliance appeared
to reflect a problem with the policy, training, and/or implementation of the policy, and/or
training.
Report Review
One of the most challenging aspects of any analysis of CED use by HPD was the gathering
of the data. As outlined in the report by the UH CPP, there also have been statistical
challenges due to the lack of data.
One of the major obstacles was the CED Incident reporting form used by HPD. The
incident reports reflected the final charge, if any, against the suspect on whom the CED
was deployed. It also reflected any additional charges filed. However, the initial reason for
the call for service was not readily available for a majority of the reports. For instance, the
report may have reflected that the suspect was arrested for Resisting Arrest, Search, or
Transportation. However, our review of the report may have identified that the officer was
initially dispatched to a disturbance call due to a reported verbal altercation, and, when the
officer attempted to investigate the incident, the suspect became aggressive and physically
resisted the officer. Certain incident reports listed Resisting Arrest, Search, or
Transportation as the reason for the call, but the actual reason for the officer being at the
location in the first place was a disturbance call.
Additionally, there was no easy method to determine the number of cycles for which the
CED was deployed; whether the suspect appeared to be under the influence of alcohol or
drugs at the time of the deployment; whether the officer initiated the call or was dispatched
to the call; and the apparent effectiveness of the CED deployment. While there was a
specific report required by policy to be generated by the reviewing supervisor as to the
number of cycles deployed, it was not captured electronically in a form that made it
electronically retrievable without reviewing the actual report.

118

The only way that the above information could be reliably obtained for the purposes of this
study was to read the actual police incident reports. This task was undertaken by the SHSU
CCJ team. Each of the reports related to the 1,284 incidents was read and analyzed so as
to obtain the above information in addition to making a determination as to whether the use
of the CED was in compliance with the applicable HPD policy.
Training Observation
By the time the audit began, almost all of the current patrol officers of HPD had undergone
the initial CED training. The members of the audit team were afforded an opportunity to
attend a demonstration of the in service training received by current officers. Members of
the team were given the same information officers received in the classroom training and
the hands on training. This included deploying a CED in the same manner as officers were
trained.
Additionally, a member of the SHSU CCJ team attended the actual CED training received by
cadets at HPD. The students and instructors were observed during the classroom and
hands on portions of the training. In order to make the observation more beneficial, the
same team member who had read and analyzed the CED incident reports was used to
perform the observation. Additionally, this observation was purposely delayed until the team
member had the opportunity to complete the reading and initial analysis of the reports.
During the observation, the team member paid particular attention to whether the CED
policy was being followed and whether any concerns noted during the analysis of the reports
might be attributable or influenced by the training provided. Due to time restraints, the team
member was not able to attend all of the portions of the Basic Academy which dealt with the
use of force and/or dealing with individuals with potential mental health problems. However,
the team members did have conversations with various members of the Academy staff and
reviewed, in written form, the scenarios used during other portions of the training.

119

RESULTS OF COMPLIANCE REVIEW OF HOUSTON POLICE DEPARTMENT
INCIDENT REPORTS
Review of HPD’s Policy
HPD’s implementation of a CED program was extremely aggressive. While some, if not
most, agencies have utilized a test period and phased in the use of CEDs over a period of
months or years, HPD trained, equipped, and implemented the use of CEDs to its patrol
force in less than six months. This was a particularly difficult task given that 3,600 officers
were trained and equipped with CEDs during this period.
Such an approach, undoubtedly, resulted in a CED Policy being written based on the
experience of other agencies’ use of CEDs as opposed to being tested within the confines of
HPD. Not surprisingly, such an approach resulted in there needing to be changes made in
the policy shortly after it was implemented.
Additionally, some of the changes in the policy within the first year of CED use were the
result of HPD hosting a symposium on the use of CEDs in conjunction with the PERF in
2005. In October, 2005, PERF developed and published National Guidelines for the Use of
CEDs. HPD revised its policy in accordance with those guidelines. However, it did not
adopt all of them verbatim.
HPD policy, in general, followed the best practices of the law enforcement profession. Any
concerns about CED use did not appear to be as a result of any shortcomings in the policy.
Instead, as detailed later, any problems that existed, appeared to be the result of issues
involving training and/or implementation of the policy.
PERF Recommendations
PERF describes itself as being a “national membership organization of progressive police
executives from the largest city, county, and state law enforcement agencies” and as being
“dedicated to improving policing and advancing professionalism through research and
involvement in public policy debate.”14 When CED usage became prevalent in 2005, PERF
assembled a group of law enforcement professionals and researchers to discuss what
recommendations, if any, should be made regarding CEDs. The national meeting was held
in Houston, Texas, and co-hosted by HPD.
As a result of that meeting, surveys, and additional research, PERF published its guidelines
in October, 2005. It should be noted that PERF did not recommend a model policy, but,
instead, provided policy and training guidelines for consideration by agencies.
PERF has 52 guidelines regarding CEDs. While the entire list is available through PERF,
the following guidelines and corresponding guideline numbers were considered to address
the more important issues and be relevant to CED use by HPD.

14

Police Executive Research Forum website, www.policeforum.org.

120

Review of HPD Policy and/or Practice:
The guideline number is the actual number assigned by PERF in their guidelines. The comment
in italics after the guideline reflects the SHSU CCJ review and analysis of the HPD policy and/or
practice.
PERF Guideline No: 1
CEDs should only be used against persons who are actively resisting or exhibiting active
aggression, or to prevent individuals from harming themselves or others. CEDs should not
be used against a passive suspect.
Observation
HPD policy followed the guideline. As noted in the compliance section, there are reports
which call into question whether the measuring stick for deployment is active aggression or
passive resistance.
PERF Guideline No: 2
No more than one officer should activate a CED against a person at a time.
Observation
HPD policy followed the guideline.
PERF Guideline No: 3
When activating a CED, law enforcement officers should use it for one standard cycle and
stop to evaluate the situation (a standard cycle is five seconds). If subsequent cycles are
necessary, agency policy should restrict the number and duration of these cycles to the
minimum activations necessary to place the suspect in custody.
Observation
HPD policy called for an officer to reevaluate the need for additional cycles after the initial
cycle. As noted in the compliance section, given the large percentage of uses which
involved more than one cycle, it may be that the language of the policy and resulting training
needs to continue to be emphasized.
PERF Guideline No: 4
Training protocols should emphasize that multiple activations and continuous cycling of a
CED appear to increase the risk of death or serious injury and should be avoided where
practical.
Observation
As will be reflected in the Training Observation section, this did not appear to be sufficiently
occurring. In fact, as indicated, the exact opposite message may have been perceived by
the cadets.
121

PERF Guideline No: 5
Training should include recognizing the limitations of CED activation and being prepared to
transition to other force options as needed.
Observation
While the language of this guideline was relayed in the training, the “how to” was lacking in
the CED specific training. While transition is covered in other sections, it should be
mentioned during the CED specific training.
PERF Guideline No: 6
That a suspect is fleeing should not be the sole justification for police CED use. Severity of
offense and other circumstances should be considered before officers’ CED use on the
fleeing suspect.
Observation
This guideline was reflected in HPD policy. However, as reflected in the Compliance
section, there were a significant number of reports in which fleeing, either alone or fleeing
accompanied by resistance, such as pulling away or pushing the officer was the justification
given for using the CED.
PERF Guideline No: 7
CEDs should not generally be used against pregnant women, elderly persons, young
children, and visibly frail persons unless exigent circumstances exist.
Observation
HPD policy followed the guideline.
PERF Guideline No: 8
CEDs should not be used on handcuffed persons unless they are actively resisting or
exhibiting aggression, and/or to prevent individuals from harming themselves or others.
Observation
HPD policy followed the guideline.
PERF Guideline No: 16
Following a CED activation, officers should use a restraint technique that does not impair
respiration.
Observation
Officers are trained, in general, to use a restraint technique which does not impair
respiration. However, during the CED specific training, the issue was not addressed.
122

PERF Guideline No: 25
CEDs should be maintained in a holster on an officer’s weak (support) side to avoid the
accidental drawing and/or firing of an officer’s sidearm.
Observation
HPD policy followed the guideline.
PERF Guideline No: 28
A warning should be given to a person prior to activating the CED unless to do so would
place any other person at risk.
Observation
HPD policy followed the guideline. As reflected in the Compliance section, a large number
of the incident reports did not indicate any such warning. If the warning was given, it was
not being documented in the incident report. Additionally, the incident reports often did not
reflect that a warning is being given prior to subsequent cycles in a large percentage of the
cases. However, some reports reflect a warning being prior to subsequent use and the
suspect choosing to comply with the officer’s command in order to avoid another cycle.
PERF Guideline No: 29
When applicable, an announcement should be made to other officers on the scene that a
CED is going to be activated.
Observation
During training, officers were told to make such an announcement. Generally, reports did
not reflect any such announcement being made.15
PERF Guideline No: 30
A supervisor should respond to all incident scenes where a CED was activated.
Observation
HPD policy followed the guideline.

15

When it is noted that the report does not reflect the warning or announcement being made, it is not intended as a
finding that the warning or announcement is not being made, but that it is often not reflected in the incident report.

123

PERF Guideline No: 31
A supervisor should conduct an initial review of a CED activation.
Observation
HPD policy followed the guideline.
PERF Guideline No: 32
Every instance of CED use, including an accidental discharge, should be accounted for in a
use-of-force report.
Observation
HPD policy followed the guideline.
PERF Guideline No: 37
CED activations should be tracked in the Early Intervention System (EIS).
Observation
HPD policy followed the guideline to the extent that officers with multiple uses of CEDs were
reviewed. The tracking of those who used multiple cycles of the CED during one incident
did not appear to occur.
PERF Guideline No: 40
Departments should not solely rely on training curriculum provided by a CED manufacturer.
Agencies should ensure that manufacturers’ training does not contradict their use-of-force
policies and values. Agencies should ensure that their CED curriculum is integrated into
their overall use-of-force system.
Observation
As reflected in the Training Observation section, the training did not appear to follow this
guideline.
PERF Guideline No: 50
Officers should be aware that there is a higher risk of sudden death in people under the
influence of drugs and/or symptoms associated with excited delirium.
Observation
As reflected in the Training Observation section, the training did not appear to sufficiently
stress this issue.

124

Compliance with Policy
While the incident reports were read and analyzed in light of HPD policy in effect at the time,
HPD policy has not changed significantly for the purposes of this portion of the audit. One
notable change to the policy was that a CED should not be used on someone who is
attempting to ingest drugs orally. Other than this noted change in the policy, compliance
testing with HPD policy has been based on the current edition of the policy.
Overall, incident reports indicate that officers were complying with HPD policy on use of
CEDs.16 For uniformity, the PERF guidelines are used once again and the italicized remarks
reflect whether SHSU CCJ’s review found the use of the CED to be in compliance with HPD
policy.
PERF Guideline No: 1
CEDs should only be used against persons who are actively resisting or exhibiting active
aggression, or to prevent individuals from harming themselves or others. CEDs should not
be used against a passive suspect.
Observation
There were sufficient incidents in which CEDs appear to be used against those who were
not aggressively resisting so as to raise a question as to what was being used as the
measuring stick for deployment by some officers; active aggression or passive resistance.
In a number of cases, the suspect was reported as refusing to follow an officer’s orders, but
it was not clear whether the suspect was physically resisting or simply not doing what the
officer was requesting. The determining of what level of resistance was involved was made
more difficult by officers’ frequent reliance on “boilerplate language” as opposed to actually
describing what the suspect did and/or said.17 Training recommendations will be addressed
in the Recommendations section
PERF Guideline No: 2
No more than one officer should activate a CED against a person at a time.
Observation
There were incident reports, although few, which reflect a violation of this policy provision.
The failure to follow this guideline may be linked to a failure to follow guideline #29, in that
incident reports did not reflect that officers were announcing that they intended to use the
CED. There were a few incidents in which the officers consciously decided to deploy their
CEDs at the same time. These decisions may have been the result of officers not
appreciating the concern as to whether multiple cycles, and, in particular, multiple cycles by
two or more officers at the same time, may have a negative health effect on the suspect.

16

One of the difficulties in conducting any research and/or analysis based on incident reports is that the researcher
does not have the capability of independently assessing the accuracy of what is reflected in the reports.
17
For instance, officers would report that the suspect “aggressively resisted” and/or that the officers “applied
reasonable force,” but would not describe the behavior and/or words of the suspect.

125

PERF Guideline No: 3
When activating a CED, law enforcement officers should use it for one standard cycle and
stop to evaluate the situation. If subsequent cycles are necessary, agency policy should
restrict the number and duration of these cycles to the minimum activations necessary to
place the suspect in custody.
Observation
Approximately 38% of deployments of a CED consisted of one cycle. Another 31% of the
deployments were two cycles in duration. Three cycles were used in 15% of the
deployments. Four cycles were used in 5% of the deployments, while five cycles account
for 3% of the deployments. In 5% of the deployments, between six and ten cycles were
used. In 3% of the deployments more than ten cycles were used. It appears that what
officers were evaluating was not whether they should use another cycle of the CED, but
whether the CED had produced the desired results: the individual complying with their
orders. In the cases in which the CED was cycled more than five times, it often
appeared that not only had the device not produced the desired results, but that
additional cycles did not change the ultimate outcome. In these incidents, it was
suggested that the officer should be evaluating other options. Part of the problem, as
detailed in the Training Observation section, was that some of the films used in the training
appeared to send a message that multiple cycles were an acceptable practice, as opposed
to emphasizing that it was unknown as to whether multiple cycles posed a health risk for the
suspect. Also, what appeared to be an issue was a lack of integrated training as to how an
officer should transition from one type of use-of-force to another.
PERF Guideline No: 4
Training protocols should emphasize that multiple activations and continuous cycling of a
CED appear to increase the risk of death or serious injury and should be avoided where
practical.
Observation
As noted above, and as reflected in the Training Observation section, this did not appear to
be sufficiently occurring. In fact, as indicated, the exact opposite message might be
perceived by at least some of the cadets in that some of the films used in the training
appeared to send a message that multiple cycles were acceptable as opposed to
emphasizing that it was unknown as to whether multiple cycles pose a health risk for the
suspect. Also, as indicated above, there appeared to be a lack of integrated training as to
how an officer should transition from one type use-of-force to another.

126

PERF Guideline No: 5
Training should include recognizing the limitations of CED activation and being prepared to
transition to other force options as needed.
Observation
As far as transition to other force options, the “how to” appeared to be lacking in the CED
specific training as reflected in the Training Observation and Recommendations sections.
Given the large number of deployments which result in multiple cycles being used, it may be
that officers lack an understanding as to how to reevaluate the situation and determine if a
different approach is appropriate, including the use of a different form of force. As noted
above, in over 30% of the incidents, three or more cycles were used. In approximately 3%
of the deployments, over ten cycles were used.
PERF Guideline No: 6
That a suspect is fleeing should not be the sole justification for police CED use. Severity of
offense and other circumstances should be considered before officers’ CED use on the
fleeing suspect.
Observation
There were incident reports in which fleeing, either alone or fleeing accompanied by
resistance, such as pulling away from the officer or pushing the officer was the justification
for using the CED. In a number of cases, particularly self-initiated drug investigations, it
appeared that a CED was deployed even though the suspect was never touched by the
officer. In those incidents, the incident report indicated that the suspect began to flee either
upon seeing the officer or as the officer was approaching the suspect. Some reports
indicated that the suspect was reaching into his waistband area or pockets while fleeing.
PERF Guideline No: 7
CEDs should not generally be used against pregnant women, elderly persons, young
children, and visibly frail persons unless exigent circumstances exist.
Observation
Very few reports indicated that this policy was violated. When a CED was used on one of
these populations, the reports indicated exigent circumstances.

127

PERF Guideline No: 8
CEDs should not be used on handcuffed persons unless they are actively resisting or
exhibiting aggression, and/or to prevent individuals from harming themselves or others.
Observation
There were incidents when CEDs have been deployed on handcuffed individuals. The
reports, with very few exceptions, articulated the active resistance or aggression of the
suspect.
PERF Guideline No: 16
Following a CED activation, officers should use a restraint technique that does not impair
respiration.
Observation
As reflected in the Training Observation section, this did not appear to be occurring in the
CED specific training. In other parts of training, this was covered generally. In general,
there did not appear to be sufficient emphasis on the potential health risks to the suspect
upon which the CED was deployed.
PERF Guideline No: 25
CEDs should be maintained in a holster on an officer’s weak (support) side to avoid the
accidental drawing and/or firing of an officer’s sidearm.
Observation
This policy was followed. In a few reports, it appeared that the officer drew both the CED
and sidearm (although the officer did not necessarily have both of them out of their holsters
at the same time). HPD does not use the bright yellow CEDs. One of the advantages of the
bright yellow CED is that it a visual reminder to the officer as to whether the officer is holding
the CED or a sidearm.
PERF Guideline No: 28
A warning should be given to a person prior to activating the CED unless to do so would
place any other person at risk.
Observation
The majority of the reports did not indicate any such warning. If the warning was given, it
was not articulated in the incident reports. Additionally, the majority of the incident reports
did not indicate that a warning was given prior to additional cycles. However, some reports
reflected a warning was given prior to subsequent use and the suspect choosing to comply
with the officer’s orders in order to avoid another cycle. As more of the public becomes
aware of the CED, greater use of a warning may result in compliance with an officer’s order
without having to resort to deployment. HPD does not track the number of times that an
officer displayed the CED or warned that the CED would be used and did not deploy the
CED. We had inadequate information to determine if such actions by officers were effective
in gaining cooperation without deployment of the CED.

128

PERF Guideline No: 29
When applicable, an announcement should be made to other officers on the scene that a
CED is going to be activated.
Observation
During training, officers were told to make such an announcement. Generally, reports did
not reflect any such announcement being made. It was unclear whether the announcement
was being made, but not being reflected in the report or not being made. This may be a
result of lack of hands on training and practical situational training in CED use. Studies
show that telling someone what to do is not as effective in instilling the behavior by having
them practice the behavior.
PERF Guideline No: 30
A supervisor should respond to all incident scenes where a CED was activated.
Observation
It was very rare that a supervisor did not respond. Under early versions of the policy, it was
not clear who was to respond to scenes where officers deployed their CED while working an
extra job or off duty job. The policy has been clarified and on duty sergeants now respond.
PERF Guideline No: 31
A supervisor should conduct an initial review of a CED activation.
Observation
Sergeants conduct an initial review. One of the problems noted, particularly in early reports,
was that the sergeant would simply write “download reflects agreement with report” without
either the officer or the sergeant detailing how many times and for how long the CED had
been deployed. A download report exists, but it was frequently not incorporated in either the
incident report or the sergeant’s report. Inclusion of the information in the sergeant’s report
would be beneficial for determining compliance with policy and tracking of individual officers’
CED use.

129

PERF Guideline No: 32
Every instance of CED use, including an accidental discharge, should be accounted for in a
use-of-force report.
Observation
Generally, it appears that this policy was being followed and that HPD’s control over the
CED cartridges helped to insure that all uses were documented. During the Scope period,
HPD improved their system of internal control by documenting their review of each CED
Incident Report by an Executive Assistant Chief. A new policy is still required to insure that
each officer has their required two cartridges and is not deploying or losing a cartridge and
not reporting it.
PERF Guideline No: 37
CED activations should be tracked in the EIS.
Observation
There were instances in which a particular officer’s multiple CED uses were reviewed. It
would be beneficial to have those cases independently reviewed. According to the results of
the Focus Groups conducted by MFR, when an officer’s frequent CED use is questioned,
the explanation that the officer makes a lot of arrests or works a particular assignment such
as street narcotics is given. As with other uses of force, the more complete analysis is to
compare the officer’s use of the CED with other officers who have similar levels of arrests or
similar assignments. Additionally, the tracking of those who use multiple CED cycles during
one incident did not appear to be occurring.
PERF Guideline No: 40
Departments should not solely rely on training curriculum provided by a CED manufacturer.
Agencies should ensure that manufacturers’ training does not contradict their use-of-force
policies and values. Agencies should ensure that their CED curriculum is integrated into
their overall use-of-force system.
Observation
As reflected in the Training Observation section, this did not appear to be sufficiently
occurring. The officers conducting the training have received training through TASER,
International, and appeared to often be simply repeating the training they received.

130

PERF Guideline No: 50
Officers should be aware that there is a higher risk of sudden death in people under the
influence of drugs and/or symptoms associated with excited delirium.
Observation
As reflected in the Training Observation section, this did not appear to be sufficiently
occurring. In fact, as noted in that section, officers were told that CEDs have not been found
to be responsible for any deaths. Whether that is an accurate statement is not the issue, the
issue is whether officers have an appreciation for the potential danger to the suspect who is
under the influence of drugs and/or exhibiting symptoms associated with excited delirium.

131

HOUSTON POLICE DEPARTMENT INCIDENT REPORT REVIEW
Introduction
During the design of the audit, it was determined that several issues should be reviewed by
SHSU CPJ. For instance, when a CED was used, had the officer become involved in the
incident as a result of being dispatched, having initiated the contact, or as a result of a traffic
stop? Additionally, what types of calls were resulting in CED use? When a CED was
deployed, how many cycles were being used on suspects? Were the suspects noted to be
under the influence of alcohol or drugs at the time of the deployment? When a CED was
used, was it effective? If the CED was not effective in controlling the suspect, why was it not
effective?
As noted previously, one of the major obstacles to determining compliance with policy and
ascertaining any trends is the reporting form used by HPD. In order to obtain the answers to
the above questions, each report had to be read and analyzed.
Initiation of Call for Service
Fifty-five percent (55%) of the time, in the incidents which resulted in the CED deployment,
the officer responded to the scene as a result of being dispatched. In other words, a
member of the public telephoned HPD and reported either a crime or a situation which
warranted law enforcement attention.
In 33% of the cases, the contact with the suspect, against whom the CED was eventually
deployed, was initiated by the officer. A large percentage of these cases were drug
investigations in which HPD was proactively patrolling areas known for drug sales and/or
use. Another group which falls into this category was the situation in which an officer
noticed someone behaving in a bizarre manner so as to indicate the subject was either
under the influence of drugs, alcohol, and/or experiencing a mental health problem.

132

In 12% of the cases, the initial reason for the contact with the suspect was a traffic issue. It
should be noted that in many of these cases, it appeared that the officer also suspected the
person to be involved with other criminal activity, particularly drugs.
Table 1 - Summary of Calls for Service and
Whether They Were Dispatched, Self-initiated, or as a Result of a Traffic Stop

Traffic
12%

Dispatched
55%

Self Initated
33%

Nature of Call for Service
As indicated previously, the incident reports were designed so that the nature of the call was
captured based on the charges ultimately filed against a suspect, if any. For instance, the
charge might reflect that the suspect was charged with Escape. However, it would be
extremely unusual for an officer to respond to a call for service for an Escape. After review
of the incident report, SHSU CPJ determined that the officer responded to the call for
another underlying reason, arrested the person, and the person fled or escaped from the
officer.
While some reports reflect the initial basis for the call, many do not. What made it even
more challenging was that the only way to definitely determine the nature of the original call
was to read the actual, detailed incident report.

133

The two most frequent types of calls for service were drug related (15%)18 and disturbance
related (15%).19 The incidents involving drugs were notable in that, with a few exceptions,
they were calls which were self-initiated by an officer. Some were, as noted above, the
result of a traffic stop where the officer had probable cause to make the traffic stop based on
a traffic violation, but the traffic violation appeared to be a pretext to determine whether the
occupants in the automobile were involved in illegal activity.20
Assault was the basis for 13% of the calls. This was in addition to the 7% of the calls which
were based on a report of family violence. Criminal trespass was the basis of 6% of the
calls. Traffic offenses were the basis for 11% of the calls. Other misdemeanors such as
Driving While Intoxicated, Theft, and Criminal Mischief comprised the basis for 13% of the
calls. Other felonies comprised the basis for 9% of the calls. Approximately one-third of the
felonies were related to automobile theft. The incident reports indicated that these suspects
often abandoned the stolen vehicle and fled on foot.
The last category of calls for service was related to mental health issues. While the person
involved may have committed a crime, the criminal charge was often not filed so that the
person could be committed to receive the necessary mental health care. Eleven percent of
the calls were based on the need for crisis intervention by officers with individuals
experiencing a mental health issue.

18

For purposes of this study, drug related calls include possession of drugs and the sale of drugs.
For purposes of this study, disturbance related calls include the crimes normally associated with disorderly conduct
and public intoxication.
20
It should be noted that the United States Supreme Court has ruled that pretext traffic stops are not a violation of the
Fourth Amendment to the United States Constitution.
19

134

Table 2 -Summary of the Original Type of Call for Service Which Resulted in CED
Deployment

20
18
16

Percent

14
12
10
8
6
4
2

is
de
m
ea
no
rs
Ot
he
rF
el
on
ie
s

Tr
af
fic

Ot
he
rM

As
sa
ul
t

Cr
Dr
im
ug
in
al
Tr
es
sp
Di
as
so
s
rd
er
ly
Co
nd
uc
t
M
en
ta
lH
ea
lth

Fa
m

ily

Vi
ol
en
ce

0

Type of Call
Number of CED Cycles Used When Deployed
A standard cycle of the CED is five seconds. This occurs when the officer pulls the trigger
and releases it. If the officer continues to hold the trigger down, the CED has a continuous
cycle lasting as long as the trigger is pulled.
In 38% of the incidents, the officer only deployed one cycle of the CED. In 31% of the
incidents, the officer deployed two cycles of the CED. In reviewing the reports, the most
common justification for a second cycle was that, while the initial deployment resulted in the
person being incapacitated, once the suspect was no longer incapacitated, the suspect did
not comply with the cuffing instructions.21

21

It appears that this usually consisted of the person refusing to put his hands behind his back.

135

In 15% of the incidents, three cycles of the CED were used. Again, the most common
reason was that the individual would not allow himself to be cuffed. In 5% of the incidents,
the CED was cycled four times. In 3% of the cases, the CED was cycled five times. In 5%
of the cases, the CED was cycled between six and ten times. In 3% of the cases, the CED
was cycled over ten times.
In many of the cases in which the CED was cycled in excess of five times (8% of the cases),
the report indicated that the CED was not effective in controlling the suspect. This calls into
question whether the officer should have continued to use additional cycles or should have
reevaluated the situation to determine whether a different approach was required to resolve
the situation.
Table 3 -Summary of Number of CED Cycles Deployed

40
35
30
25
Percentage 20
15
10
5
0

Number of
Deployments in
Percent

1

2

3

4

5

6-10 10+

Number Of Deployment
Impact of Alcohol and/or Drugs Consumed by Suspects
In 58% of the incidents, officers did not note that the suspect appeared to be under the
influence of drugs and/or alcohol. The researcher did not attempt to determine whether the
officer, if asked, would have said that drugs and/or alcohol might have been a contributing
factor to the behavior which resulted in the need for the CED deployment. However, it is
highly likely that a significant percentage of these suspects were under the influence of
drugs and/or alcohol given the behavior that was noted by the officer.
In 15% of the incidents, the reports noted that the suspect appeared to be under the
influence of alcohol. In 27% of the incidents, the reports noted that the suspect appeared to
be under the influence of drugs and/or possession of drugs.

136

PERF Guideline #50 states that officers should be aware that there is a higher risk of
sudden death in people under the influence of drugs and/or symptoms associated with
excited delirium. The reason for the caution is the suspected correlation between drug use,
particularly methamphetamine and cocaine, and excited delirium. What is unknown is the
additional effect of repeated cycles of the CED on the health of someone who is under the
influence of drugs and/or experiencing excited delirium. The majority of the cases in which
the CED was cycled more than ten times involved a subject who was described as being
“high” on drugs and/or having a mental health issue.
Table 4—Summary of Impact of Alcohol and/or Drugs Consumed by Suspects

None Noted

27%

Alcohol
58%

Drugs

15%

Effectiveness of CED in Controlling a Suspect
The CED was reported as being effective in assisting in controlling the suspect in 77% of the
incidents. It should be noted that this does not mean that the CED was effective in its initial
deployment or effective with only one cycle. In some cases the officer deployed the CED in the
probe mode and then resorted to using the CED in the drive stun mode or used multiple cycles
in the probe mode to gain control.
In 4% of the incidents, the incident report indicated that the CED was not effective due to the
clothing the person was wearing. By this it was meant that the clothing would not allow the dart
to penetrate the skin. In northern states, longer probes are utilized by law enforcement to
overcome this problem. Given the low number of incidents in which clothing appears to be an
issue, it does not appear the longer probes are warranted.

137

In 4% of the incidents, the officer missed the suspect with one or more darts. In 2% of the
cases, it was reported that the darts hit the suspect, but the suspect pulled out the darts. In
2% of the incidents, it is reported that the CED was defective and would not fire the probes.
In an additional 11% of the incidents, the officer indicated that the CED was not effective
without attributing the ineffectiveness to any of the reasons previously identified. In the
majority of the cases where the officer indicated that the CED was not effective, the officer
reported that the subject appeared to be under the influence of alcohol and/or drugs or
experiencing some form of mental health crisis.
While the CED was effective in three out of four situations, the quarter of the time it was not
effective, highlights the need for officers to have another option in mind to gain control of the
situation. It is suspected that the reports of ineffectiveness of the CED in the drive stun
mode may be related to training. As is detailed in the Training Observation, emphasis is
lacking in the training on how to properly administer a drive stun.
During the Focus Groups, there was a difference of opinion as to the effectiveness of the
CED. The officers appeared to be of the opinion that the CED was less likely to work on
someone who was high on drugs, intoxicated, or had a mental health issue. The reports
appear to support their anecdotal based belief.
Table 5 - Summary of the Reasons Given When the CED Deployment was Not Effective

2
11

Instances

Instances

2
Instances

4
Instances

4
Instances

Clothing

Missed

Pulled Out

138

Defective

Not Noted

HOUSTON POLICE DEPARTMENT TRAINING ACADEMY - CED TRAINING
Classroom Instruction
A SHSU CCJ team member attended the CED cadet training on August 29, 2007. The
instruction was provided by members of the training staff from the Academy.
At the beginning of the class, the instructor asked for volunteers for the afternoon. The
students were obviously aware that what they were volunteering for was to have a CED
deployed on them. Students were instructed to place their names in a hat for a drawing for
thirteen volunteers. From the beginning, this created a stir of excitement among the
students. The discussion continued amongst the students during the breaks and lunch hour
(they already knew whose names had been drawn) with students who were not chosen
often professing their desire to experience the CED and students whose names had been
drawn professing their lack of fear.
During the actual classroom training, instructors highlighted the various provisions of the
policy. Some of the notable ones were:
•

•
•
•
•

•
•
•

CED is an intermediate weapon and not a substitute for lethal force. Students were
told “do not use a Taser to stop a knife.” This was emphasized several times by the
instructor.
CED is to be used on those who are actively resisting. Actively resisting was
explained as “kicking, fighting, etc.”
CED is not to be used on those who are passively resisting, handcuffed, or known to
be mentally ill. Protesters were used as an example of passive resistance.
CED is not to be used on suspects who are pregnant, elderly, visibly frail, or young
children.
CED is not to be used on someone simply because they are fleeing. The instructor
went on to say that if the person had been fighting with the officer right before fleeing
then it was okay to deploy the CED.
CED is not to be used on someone who is only verbally non-compliant.
Number of cycles is to be kept to a minimum. The instructor expressed that excited
delirium is a concern and that the officer should reevaluate after one cycle.
Warnings are to be given to the suspect and backup officers that a CED is going to
be deployed.

Also notable were some of the other comments by instructors. When describing that the
device delivers 50,000 volts, the instructor said, “The media always plays this up. It isn’t
going to kill anybody.” Students were also told that multiple cycles were appropriate to keep
the suspect incapacitated. In contrast, there was only one mention of the danger of excited
delirium and little recognition of there having been cases in which the issue of the role the
CED had played in the death of a suspect upon whom a CED had been deployed had been
raised.

139

Part of the classroom instruction was spent watching a slide demonstration which appeared
to have been distributed by the manufacturer. Most of the information was technical in
nature and appeared to contribute little to the cadets’ understanding of when CED
deployment was appropriate. It appeared to reinforce a message that the CED was
extremely reliable. Students were told that the CED had a 94.3% success rate and the
instructor indicated that most failures were “failures in use” as opposed to a failure of the
CED.
A series of film clips from the manufacturer were shown to the cadets. One of the
instructors stated, right before he started showing the film clips, “Watch how a Taser can
change attitudes.” While this may have been meant as harmless bantering, the film clips
tended to reinforce the idea that use of the CED was appropriate when an officer needed to
“change the attitude” of the suspect as opposed to being required to overcome the active
resistance by the suspect. For instance, one of the clips showed a CED deployed against a
man in a jail cell who was being verbally disruptive. The man was in the cell by himself and
did not appear to present any danger to the officers. Additionally, the film depicted the CED
deployed multiple times with the prisoner being seen hitting his head while the officers did
nothing to prevent the potential injury. Another clip showed multiple cycles used without the
suspect given the opportunity to comply with the orders that were given.
Almost no emphasis was given to the unknown or potential dangers of the use of the CED.
The one discussion of custody deaths was sped through right before lunch. In fact, more
time was spent showing and commenting on a film clip of a CED being used on a dog by an
officer than the possible dangers of the CED.
There was some discussion regarding the possibility of adding on of cameras to CEDs. The
instructor stated there were problems with CED cameras because the cameras do not
record any information prior to the CED being turned on by the officer. The cadet’s concern
was that the activity prior to the activation of the CED camera was not recorded and
therefore not available for review to justify the CED deployment. 22
After lunch, cadets were shown a videotape of Chief Hurtt. In the first part of the videotape,
the Chief stated, in essence, that after introduction of CEDs in Phoenix that officer
involved shootings went down by one-half and that it was an alternative to deadly
force. During his remarks he made no mention of whether officers should deploy multiple
cycles or use the CED on the mentally ill. He did state that the CED should not be used on
someone who is only passively resisting such as refusing to get out of a vehicle or passively
protesting.

22

The negative reaction to the idea of cameras being placed on the CEDs also was reflected in the Focus Groups
conducted by MFR. While some officers in the Focus Groups appeared to be concerned about the cameras being
another item on their tool belt that would add weight and might get broken, the main concern seemed to be that
camera use would be detrimental to the officer. One of the issues appeared to be that the camera would capture only
part of the situation since it would not be activated until the CED was turned on immediately prior to deployment. The
other issue was that the microphone on the camera would capture inadvertent, inappropriate language being used by
the officers. The issue on the officers’ part appears to be more centered on a fear of the images and language being
used than the technical reliability of the equipment.

140

The students were also shown a videotape of Assistant Chief McClelland in which he
answered the most frequently asked questions regarding the CED use. He also
emphasized that CEDs were not to be used on passive suspects. Additionally, he cautioned
officers not to use the CED when a suspect had placed drugs in his mouth.
Another part of the classroom instruction consisted of the students being required to drive
stun the classmate on each side of them.23 This resulted in a lot of jumping in reaction to the
stun and yelling. Unfortunately, it may be that it gave a false sense of how to apply a drive
stun and the level of pain experienced by a suspect when a drive stun is properly applied.
During the classroom instruction, one student asked, “If the Taser doesn’t work, what should
I do then? Shoot him?” The instructor responded by telling the student to make an
assessment based on the use-of-force criteria. Unfortunately, that is all that was said and it
appeared that an opportunity to discuss what to do in such a situation was missed. Cadets
as well as in service officers should be provided with training which more closely resembles
actual situations faced by officers and incorporating different levels of use-of-force and how
to transition from one to another. During other parts of the training, scenarios were used
and it was intended that cadets will have to transition from one use of force to another.
However, there were not specific scenarios regarding situations where the CED use was
ineffective.
Hands On Instruction
In the afternoon, the students were taken into the gym and the volunteers were instructed to
remove their uniform shirts and protective vests so as to only have their undershirts on their
upper bodies. Students were then held up over protective pads by two of their fellow
students while the CED was deployed into their backs. In all cases, the students
demonstrated incapacitation and were lowered to the mats by their classmates. While it is
understandable that there are limitations as to what is proper to do in hands on training, the
method chosen seemed to ignore some of the realities of deploying a CED in real situations
and the consequences. For instance:
•

•

•

•
•

•

All of the students were standing still at the optimal distance with their backs turned
to the instructor. In actual incident reports, often the suspect is fleeing or struggling
and does not present an easy target.
None of the students were running when they were hit by the darts. This may have
resulted in the students, perhaps, not appreciating the force at which a running
suspect would fall to the ground in the case of sudden incapacitation.
None of the students were given orders during the cycle so that they and their fellow
students could appreciate the inability of the suspect to comply with orders during the
cycle and, perhaps, to even comprehend orders given during the cycle.
None of the students were subjected to multiple cycles.
There was no training, during this section, on how to go about cuffing the suspect
once the CED resulted in incapacitation. Specifically, there was no mention of the
PERF recommendation that a restraint technique be used that does not restrict
respiration.
There was no training, during this section, on how to transition from CED use to
another type of use-of-force if the CED proved ineffective.

23

As noted previously, a “drive stun” or “dry stun” is when the CED is deployed on a person without the use of the
probes.

141

After the demonstration on the cadets, all of the cadets were given two cartridges. The first
cartridge was to be deployed into a standing silhouette target. The student was instructed to
then discard that cartridge, load a second cartridge, and deploy the second cartridge at a
silhouette in a prone position. The student was to then discard the second cartridge and
drive stun a dummy. The students were then divided into two lines and quickly put through
this portion of the training.
Several problems were noted during the hands on training. First, a true demonstration of
proficiency was not required. Students were required to go through the steps, but they were
not required to perform them correctly. For instance, approximately ten percent of the
students missed the first silhouette with one or both of the darts. More disturbing, clearly
40% of the students did not perform the drive stun correctly. The most common error in the
drive stun was that the cadet placed the CED against the dummy but did not pull the dummy
towards them to overcome any resistance. Second, although it was explained to the cadets
that in order to have the darts hit a prone target that they would have to cant the weapon, it
was not explained when it would be appropriate to deploy the CED against someone who
was in a prone position.
It should be noted that the observer was only present in the training for the one day. It
should also be noted that the Academy staff appears to be under pressure to get a large
number of students through training in a very short period of time which does not
necessarily lend itself to stopping each student when the task was performed incorrectly and
correcting their technique. Additionally, the cartridges are expensive which limit the ability to
have the students deploy multiple cartridges. That being said, it may be that cadets are
leaving the Academy with some mixed messages.
•

•

•

•

•

They were told that the CED is only to be used on persons who are actively resisting,
but they are shown a film clip in which the CED was used on a suspect whose
provocation was limited to verbal. In addition, they were told that CEDs “change
attitudes.”
They were told that the CED was not alternative to deadly force and should not be
relied on if the suspect had a knife, but they were shown a videotape in which Chief
Hurtt talked about the reduction in officer involved shootings in Phoenix after the
introduction of the CED.
Little mention and almost no emphasis were put on the potential danger of the CED
to the suspect. In fact, cadets were told CEDs have never killed anyone and only
passing mention was made of excited delirium and individuals under the influence of
drugs.
The hands on training during which cadets were subjected to having the CED
deployed was not in a realistic atmosphere with no time taken to explain how it was
not reflective of real life situations.
The students were not required to demonstrate proficiency with the CED; only go
through the steps.

142

COMMENTS ON UH CCP TEAM STATISTICAL ANALYSIS OF THE USE OF
CEDS
Introduction
Often the results of studies and/or audits are viewed in terms of the numbers that result from
them. The comments on the results of the statistical analysis were an attempt to answer the
“whys”. Why was it that if the CED is effective in bringing a suspect under control that there had
not been a significant reduction in the number of injuries to officers? Since CEDs were viewed
by a majority of the public as an alternative to the use of deadly force by officers, why hadn’t the
number of officer involved shootings gone down? Were there circumstances where officers
were deploying a CED to control a suspect where they would have been legally justified in using
deadly force? If there were disparities in the way officers deployed CEDs, what might be at the
root of the decisions being made by officers?
Injury Data Comments
While the CED tends to be effective in temporarily incapacitating the suspect, in order to take
the suspect into custody, the officer is still required to “put hands on.” By this it is meant that,
even if the CED was effective at stopping an aggressive action or a fleeing suspect, the officer
still must physically engage the suspect to handcuff the suspect and transport the suspect to
jail. A great number of the incident reports indicated that the suspect began to again resist once
the initial cycle was finished.
The results of the study by UH CPP did not indicate a reduction in injuries which resulted in
workers’ compensation claims since the introduction of the CED that could be attributed to the
use of CEDs at this time. There was inadequate data to determine whether there was a
reduction in minor injuries to officers which would not result in a claim being filed.
Additionally, the data was not available as to injuries to suspects. While there are dangers to
the suspect as a result of the CED, there are also dangers of injury to the suspect when other
types of force are used. Deadly force is the most obvious extreme example, but the use of a
baton is probably more appropriate for analysis of reduction of injuries to suspects. While in
most cases it appears that the effects of the CED deployment dissipate shortly after the cycle is
over, the effects of being struck by a baton, particularly with enough force to knock the suspect
to the ground, continue after the use of the baton has ceased.
Substitution Data Comments
The results of the study by the UH CPP did not indicate a reduction in the number of officer
involved shootings since the introduction of the CED. This is not surprising as the opportunities
to use a CED instead of a firearm are very limited. For example, no one would expect an officer
to rely on a CED to stop a suspect who was armed with a firearm. For one, the officer must be
a maximum of 21’ from the suspect in order to successfully deploy the CED.24 At such a
distance, the officer would definitely be within range of a firearm.

24

The wires on the CED are 21’ long. Additionally, the optimal distance is 7’ to 15’.

143

As far as knives or other items which might cause serious bodily harm or death25 to an officer,
the range of the CED also presents a challenge. Studies have shown that a suspect can cover
21’ (the maximum distance from which a CED can be deployed) in the time it takes an officer to
perceive the need to react, draw a weapon, and fire. That is why it was emphasized in the
training that a CED is not the proper weapon to use if the suspect has a knife unless the officer
has cover and there are at least two officers on the scene. In that situation, one of the officers
deploys the CED while the other officer stands ready to use a firearm if the CED is not effective
and the suspect’s actions warrant the use of the firearm.
That being said, there were incident reports which indicated that officers would have been
legally justified in using deadly force and officers chose, instead, to deploy the CED. What was
not clear from most of the reports was whether another officer was present who had drawn a
firearm and was prepared to use it if the danger to the officer deploying the CED escalated. The
following are examples of when it appeared that the situation might escalate, or had escalated
to the point where an officer would have been justified in using deadly force, but chose to use
the CED instead. It should be noted that just because the officer would have been legally
justified in using deadly force, it does not mean that the officer would definitely have chosen this
option. However, a review of the reports indicated situations in which other intermediate
weapons were unlikely to have been used due to officer safety concerns; thus, if the CED had
not been available, the use of deadly force would have been more likely. Some examples of
these types of incidents are:
•

•
•

•

•

•
•

2005-Suspect who was wanted on six felony warrants fled on foot. While running he
was yelling that he was going to kill the officers and was reaching into his pocket. The
officer deployed the CED which knocked the suspect to the ground and he was secured.
2005-Assault suspect charged after the complainant was attacked with a machete. The
officer deployed the CED twice and the suspect was secured.
2005-Mental health suspect had knives taped to both hands. Subject told officers he
wanted the officers to shoot him. Officer deployed CED which caused the subject to go
to ground where officers were able to disarm and secure the subject.
2005-Aggravated robbery suspect fled after robbing a restaurant with a gun. Officer
deployed CED to knock suspect to ground and to keep him incapacitated until backup
officers arrived.
2005-Juvenile suspect who was reported to have knives refused to remove hands from
pocket and told officers she had a gun. Officer deployed CED one time and suspect was
secured.
2006-Criminal trespass suspect fled on foot. When caught, he pulled out a knife. The
officer deployed the CED one time and was able to disarm him.
2006-Mental health subject was yelling at the officers and waving a pipe and a tire iron
at the officers. The first officer deployed the CED, but missed. The sergeant deployed
the CED and the subject was subsequently secured.

25

Under both the Texas Penal Code and United States Supreme Court cases, the criteria that must be met before an
officer may legally use deadly force is that the suspect’s actions pose a significant risk of serious bodily injury or
death to the officer or others.

144

•

•

2006-Suspect dropped a gun during a chase. When caught, he looked towards the gun
and appeared to be getting ready to reach for it. Officer deployed a CED one time and
the suspect was secured.
2007-Assault suspect who was intoxicated came at officers with a 15” pipe in each hand
telling the officers they would have to kill him to stop him. Officer deployed CED one
time and the suspect was secured.

Incidence Data Comments
The results of the incidence analysis indicate that African American suspects are more likely to
have a CED deployed on them. This is particularly true when the officer deploying the CED is a
Anglo male. As reflected in the UH CPP report, more analysis of the reason for this disparity is
appropriate.
One possible hypothesis is the phenomenon coined by sociologist Jerome Skolnick as
“symbolic assailants.”26 The idea is that the use-of-force, particularly excessive use-of-force, is
an outcome of the informal subculture of law enforcement. According to the theory, the
subculture is a function of the danger to which officers are exposed and the authority they are
given to use force. This phenomenon may be heightened when the persons that law
enforcement encounters are seen as “defiant, threatening, disrespectful, and disorderly
individuals, many of whom possess the low socioeconomic and minority status attributes of the
marginalized.”27
Correspondingly, studies indicate that minorities, particularly African Americans, perceive law
enforcement more negatively than Anglos. Some studies have shown that minorities fear law
enforcement.
Often the reports reflect that the suspect, when confronted by officers, remarks that the officers
have somehow singled the suspect out unfairly. This response appears to often be taken by the
officer as a further warning sign of impending aggression by the suspect. It may be that one
explanation for the disparity noted of Anglo male officers deploying CEDs disproportionately on
African Americans is that these officers perceive these suspects as “symbolic assailants” who
present a threat to their safety and the suspects perceive the Anglo male officers as being more
likely to treat them unfairly. Just as the Anglo male officers may react to the African American
suspect differently, the African American suspects may be reacting to the Anglo male officer
differently than they react to female officers of all races or African American male officers.
Further analysis of reports with an emphasis on this possibility is recommended.

26

rd

Skolnick, J. (1994). Justice without trial: Law enforcement in a democratic society (3 ed.). New York: Macmillan.
Micucci, A.J., & Gomme, I.M. (2005). American police and subcultural support for the use of excessive force.
Journal of Criminal Justice, 33, 487-500.
27

145

Another important issue, assuming no disproportionate or discriminatory effect, is what the
officer should do instead of deploying the CED. Assuming the officer is legally justified in
detaining the suspect and the suspect is not yielding to verbal commands, the officer is placed
in a position of escalating the level of force to one where the officer goes “hands on” with the
suspect. With physical contact comes the risk of injury to both the officer and the suspect. This
is particularly true when the subject is under the influence of drugs and/or alcohol or
experiencing a mental health crisis. It may be that the CED is the most effective way to get the
person under control. Perhaps the emphasis should be on what officers should do once the
person is under control so as to lessen the likelihood of further injury to the suspect.
One particular type of incident noted was that the CED was used because the suspect had fled
the officer on foot when the officer attempted to detain the suspect due to a suspicion of drug
activity. The report often described the suspect as reaching into his pocket, groin area, or
waistband while running. In a number of these cases, the suspect discarded some thing as he
fled. In Texas, Resisting Arrest, Search, or Transportation is a Class A misdemeanor.28 Evading
Arrest is a Class B misdemeanor.29 Possession of cocaine or methamphetamine is a felony.30 It
can be assumed that a majority of the suspects hope to elude capture by the officer. It also
appears, at the very least, that the suspects may be fleeing in hopes of being able to discard
their drugs before they are captured. If the suspect is successful in discarding the drugs and
they are not retrieved by the officer, the most serious charge the suspect will face as a result of
fleeing is a misdemeanor. If the suspect does not flee and is in possession of drugs such as
cocaine or methamphetamine, they will face felony charges.

28

A Class A misdemeanor is punishable by confinement in jail for up to one year.
A Class B misdemeanor is punishable by confinement in jail for up to 180 days.
30
Whether the possession will be a state jail felony (punishable by confined in a state jail for up to two years) or a
higher level of felony punishable by confinement in the Institutional Division of the Texas Department of Criminal
Justice depends on the drug involved and the amount of the drug possessed.
29

146

RECOMMENDATIONS
Policy
HPD’s policy on the use of CEDs is well written and in line with the best practices of the law
enforcement profession. The problems with its implementation may be the result of training
and supervision issues as opposed to the wording of the policy. However, it is
recommended that HPD consider altering the policy so as to:
•

•

•

•

•

•
•

Require the CED download information to be incorporated into the initial report so
that it can easily be obtained and reviewed. This should be in an electronic format.
The current system does not allow for this recommendation to be implemented; it is
suggested that this option be explored when designing the new system.
Stress in training that officers should limit the number of cycles deployed in light of
the potential danger to the suspect of multiple cycles and the apparent
ineffectiveness of multiple cycles. The review of the reports indicated that the CED
was seldom successful on a subject when it was used more than five cycles. It is
recommended that there be an automatic detailed analysis of any incident in which
the CED is used in excess of five cycles to determine compliance with policy.
HPD GOs require medical screening at the jail of any subject against whom a CED is
deployed. This practice should be emphasized as, while the available research
which has been conducted on the physical effects of being shocked by a CED has
not definitely concluded that deploying a CED on a person under the influence of
alcohol and/or drugs has an adverse effect, the majority of the studies suggest that it
is desirable to have the subject checked out by medical personnel.
Require immediate medical screening and transportation to a medical facility of the
subject if a CED is deployed and the subject appears to be experiencing excited
delirium. While the research which has been done on the physical effects of being
shocked by a CED has not definitely concluded that deploying a CED on a person
suffering excited delirium has an adverse effect, a majority of the studies, which has
addressed the issue, strongly suggest that this particular group is the one most likely
to be adversely effected. Given the sudden onset of the effects of excited delirium, it
is suggested that a subject should be transported to the hospital immediately for
medical care.
Clarify when it is appropriate to use a CED on a subject who is fleeing from an
officer. This should require active aggression on the part of the subject, separate
and apart from fleeing in itself. The HPD GOs set out the policy clearly, but the focus
groups indicated there may be a lack of understanding in the implementation on the
part of some officers.
Provide for tracking of CED deployment and the number of cycles used in the HPD
EIS.
Continue to prohibit use of the CED by multiple officers at the same time. The HPD
GOs set out the policy clearly; however, it should be emphasized in roll call training.

147

Training
While it is recognized that the Academy has been called upon to train a large number of new
cadets and to provide in services training to the entire force, it is recommended that the
CED training continued to be reviewed. One of the tools that HPD has used that appeared
to be particularly effective was the use of training bulletins and roll call training to emphasize
certain aspects of the GOs. It is recommended that the training be reviewed to:
•

•

•

•

•

•

•

•

•

Determine whether additional scenarios should be included in the Field Problems
Program or Simulations which specifically involves the use of CEDs under situations
in which the CED appears to be less effective.
Continue to emphasize when a suspect’s behavior is actively aggressive so as to
warrant CED use. The calls for service/incident reports reflect a lack of
understanding by a small number of officers of the difference between passively
resisting and aggressively resisting and the alternatives available to overcome the
resistance.
Continue to emphasize when it is appropriate to use a CED on a fleeing subject.
Part of the training should be to demonstrate, or, at least, explain the risk to the
subject of sustaining injury due to falling after being suddenly incapacitated. Also,
training should emphasize the difficulty of actually hitting a fleeing subject with both
darts.
Demonstrate how a subject has difficulty in complying with orders given by an officer
while being subjected to a CED deployment. For instance, have the volunteers who
are subjected to a CED deployment, attempt to comply with common orders such as
placing their hands behind their backs.
Emphasize how to use the initial incapacitation period as an opportunity to gain
control of the subject. For instance, demonstrate how to assume the proper position
for handcuffing the subject.
Continue to emphasize the risk to subjects of being placed in a position which
impairs respiration after deployment of the CED. During the initial CED training,
instructors should demonstrate handcuffing the subject in a manner which does not
impair respiration and have cadets demonstrate proficiency. While it is understood
that this is emphasized later in the cadet training, consideration should be given to
demonstrating it during the initial training.
Continue to emphasize how to accurately report the circumstances which warrant
CED use including describing the actions of the suspect and any warnings given to
the suspect.
Consider providing additional training which closely resembles actual situations
faced by officers in the use of CEDs and incorporates different levels of use-of-force
and how to transition from one type of use-of-force to another. For instance, add
specific scenarios in which the excited delirium may be present and/or the CED is
ineffective in controlling the suspect’s behavior. It is recommended that the actual
calls for service/incident reports and/or complaints be used as the basis for designing
the scenarios.
Continue to emphasize the signs of excited delirium and the proper steps to be taken
to lessen the likelihood of serious injury if a CED is used. The training should
continue to emphasize the potential danger to a suspect if the suspect is
experiencing excited delirium and the need for immediate medical treatment.

148

•

Continue to train on the various roles of officers during a call for service/incident
where the CED is deployed. It is suggested that this be part of the scenario based
training and that the various roles be given a designation so that officers are able to
interchange the roles in the field. The current training films should be reviewed with
an emphasis towards moving towards consistency in the message being delivered
by HPD policy and training. It is recommended that the time currently being used to
explain the workings and history of the CED would better be utilized to address some
of the concerns noted in this report.

Reports
The reports are generally well written, but there are areas of deficiency. There were concerns
expressed during the Focus Groups related to reporting requirements surrounding CED use. It
is recommended that the report form and process be revised so as to:
•
•

•

Continue to emphasize that the report should reflect the behavior that warranted the
deployment of the CED.
Provide a drop down menu where appropriate. For example the menu could include
a place to indicate whether a verbal warning was given before the initial and
subsequent CED use and whether the subject voluntarily complied.
Once the system allows, require the CED download information (a history of spark
testing and deployments since the last download) be incorporated into the initial call
for service/incident report.

149

PART V
PROTOTYPE, FUSION & MODELING, LLC
GRAPHICAL DEPLOYMENT ANALYSIS

PART V

PROTOTYPE, FUSION & MODELING, LLC
GRAPHICAL DEPLOYMENT ANALYSIS

TABLE OF CONTENTS

Page
Introduction
Background
Recommendation

150
150
156

CEDView Operating Manual
1. Main Window
2. Menus
3. Panel Menu
4. GIS Panel
5. Query Panel
6. Timeline Panel
7. Info Panel
8. Example

~?[0J

.....ft>nPC...-

~-..MI:

158
161
163
164
165
169
171
172

INTRODUCTION
Prototype, Fusion, & Modeling, LLC (PFM) was invited to take part in the Houston Police
Department (HPD) Conducted Energy Device (CED) Program Performance Audit in December of
2007. PFM’s role was to assist the audit team by creating a dynamic, interactive, threedimensional visualization of HPD CED deployments across the greater Houston area. As a
result, PFM created the CEDView software to meet these requirements.
Using principles found in the Department of Defense 5000 Series Directive for Concept
Exploration the CEDView software was developed both as an analysis tool and presentation aid.
The tool has very low system requirements. It will run on machines with Microsoft Windows
operating systems dating back to the year 2000. Its interface was designed to be
understandable to both computer experts and novices alike. The target audience for the
application ranges from government officials to the general public.

BACKGROUND
CED deployments took place in all City Council Districts during the Scope period. The chart
below summarized the total number of CED deployment by Council District for the periods
November and December 2004, 2005, 2006 and January through June 2007.

Scope Period
Council
District
A
B
C
D
E
F
G
H
I
Total

November and December
2004
3
5
2
7
4
3
3
4
4
35

2005
31
101
48
123
31
29
20
85
62
530

2006
25
85
49
93
20
37
24
82
64
479

January through June
2007
18
42
21
37
10
16
13
26
22
205

Total
77
233
120
260
65
85
60
197
152
1,249

Thirty five of the total 1,284 deployments were not included in the above summary because they
did not relate to a suspect. Seventeen of the 35 were coded in HPD’s electronic system as a
duplicate deployment (ie a deployment by a second officer); another 17 deployments related to
animals (i.e. dogs) and one was classified as other.
For graphical representation of the data summarized above see the following:

~?[0J

.....ft>nPC...-

~-..MI:

150

November and December 2004 CED Deployments By Council District
•

~:[jI,r~

CEDV" .. 1 Dd

FrVnl E.... ~r,~ZU041

1M...,. iJ

o NlglILOI!IC8mbM31 ilt)O..t
01&1 E"WI1'" 2'
K&yn'I,¥,.-gJS I.J10~

Ic""o."","

I Zope....

ZlpCodo7?006 1.17ll"o:
CatInO! Ol.lnd 0 6. il.W"

r
r

M... RQo:f:I.

r

Mo,orAoodl.otJt1t

ZlpCDdIoLMHII.

""--

r;;- HlghwGyLobcb

po

HPOSIr*lm

rHPDS~LMIs.

r

HO~piIM

rH.o.plU!lll8bllll

0",..:.0 -..:..L:J YO"...:.l..:J

151

2005 CED Deployments By Council District
•

~:[jI,r~

CEDY" .. 1 Dd

FrVnl OcY.JO/Il.aY.l.2005

1M...,. iJ

o NIgIILOl!lcembM312tOS

Ic""o."","

01&1 E"WI11 !5~O
K&yn'I'¥'.-gJSIlI"8%

ZlpCodo7?1XJEi

I Zope....

18,33~

r
r

CatJnOlOl.lncsO' 12'... U '9IiK

ZlpCDdIoLMHII.

M... RQO:f:I.

r

Mo,orAoodl.otJt1t

r;;-

HlghwGyLobcb

po

HPOSIr*lm

""--

rHPDS~LMIs.

r

HO~piIM

rH.o.plU!lll8bllll

Dey..:.LJ -

..:.LJ

Y.~..:.LJ

152

2006 CED Deployments By Council District
•

~:[jI,r~

CEDY" .. 1 Dd

FrVnl OcY.JO/Il.aY.l.2006
o NlglILOl!lcambM312006
OI&IE"WI1I.o1DI
K&yn'I'¥'.-gJ&2..04Z%

ZIP Codo 7?006

1M...,. iJ

Ic""o."","

e. Ui6~

I Zope....

CatJnOlOl.lncsD8&'18)(Joq;

r
r

M... RQo:f:I.

r

Mo,orAoodl.otJt1t

ZlpCDdIoLMHII.

""--

r;;- HlghwGyLobcb

po

HPOSIr*lm

rHPDS~LMIs.

r

HO~piIM

rH.o.plU!lll8bllll

Dey..:.LJ -

..:.LJ

Y.~..:.LJ

153

January through June 2007 CED Deployments By Council District

MOO

From OCfy.Jonuory.l,to01

I"""eo,"""

o o..;.Juna.3o.2'OO1
QlllJE""WrtJ20G
1(6)IM~-49JS'Q.OOO%

Zip Codo 11OC16, t 1 4611;
LAMlnolOllll'loO'36.Tt48l'O

CMl"'"

3

r

z.,eoo..

r

ZlPCocaLMHIl1

I

MOfOIRoot:b

r

MoJorRoodLob*It:

FHi'17 HIl)hwoylobeb

FHPO_

r
r
r

Oey•..kInlt,)O.2001

154

MPO$IMQI'Il./ItMol..

HO!lpkeh
MOIPIlClil Lebel,

November 2004 through June 2007 CED Deployments by Council Districts

From

e.... ~.I,Z5.:D04

MOO

I"""eo,"""

o o..;.Juna.3G.2'OO'
QlllJE""WrtJ 12504
1(6)IM~-49JS

11.08t%

ZiP CodO 'i'0CJ6, 30. 2.39'"4

LAMlnoi OIll1'loO' 2S4,.2fl26K

CMl"'"

3

r

z.,eoo..

r

ZlPCocaLMHIl1

I

MOfOIRoot:b

r

MoJorRoodLob*It:

FHi'17 HIl)hwoylobeb

FHPO_

r
r
r

155

MPO$IMIOf'Il./ItMol..

HO!lpkeh
MOIPIlClil Lebel'

The CEDView software includes a database of 1,254 CED events recorded by HPD. The
events were selected because they contained geographic information such as key map
address, zip code, and/or City Council District. The CED event database is stored in an
encrypted format to preserve the sensitive nature of the records. Events in the current CED
event database range in time from November 25, 2004 to June 30, 2007. The software can
readily incorporate any new additional CED events occurring between January 1, 2000 and
December 30, 2031.
Data, not directly generated by PFM, was obtained from a variety of sources. The HPD CED
incident data was supplied, under full confidentiality, by the University of Houston Center for
Public Policy. Data for the GIS component of the software was obtained from the Harris County
Public Infrastructure Department1. The Key Map coordinate system utilized by HPD, and found
in the CED incident reports, is proprietary data that is defined and maintained by Key Maps, Inc.2

RECOMMENDATION
One issue noted by the PFM team during our analysis of the data was contradictory geographic
information appearing in a few of the CED incident reports. For example, the key map address
and zip code listed for a single CED deployment in the CED incident report might, in actuality,
not coincide geographically. In fact, in a few cases, they were very far apart. Today, software
tools capable of resolving geo-coordinates (latitude, longitude) from address information are
readily available and are portable such as hand-held GPS devices that determine geopositioning in real-time.
We recommend that the new system has the functionality of generating geo-coordinates from
addresses. Not only could geo-coordinates more accurately pin-point the location of an
important scene, the geo-coordinates would also fit into the current scheme HPD uses for
locating an incident. A geo-coordinate pair uniquely maps to one key map address, one zip
code and one City Council District.

1
2

http://www.eng.hctx.net/gis/htm
http://www.keymaps.com

~?[0J

.....ft>nPC...-

~-..MI:

156

CEDVIEW OPERATING MANUAL

~?[0J

.....ft>nPC...-

~-..MI:

1. MAIN WINDOW
The Main Window is the primary display area for the CEDView application. An image of the
Main Window, with all of its default contents, is shown in Figure 1 below. A key feature of the
Main Window is the interactive map. The map presents the major features of Harris County and
the greater City of Houston metropolitan area. Although it appears similar to a traditional twodimensional map, it is actually a three-dimensional geometric model. Because of the threedimensional capabilities, CEDView supports a variety of features not possible with traditional
flat-map software.
Figure 1 CEDView Main Window
•

~.!51/~.

C£DVI,w I Do

iJ

.".,..,.

(Ballllmllp

r

Z,.Codot

I Zip Codao LobOl~
p" MtI/Or RootdJ
rt.1Q\or~lIIdllttl.11

I-Hi,....,..
r;' HlrJI'1Mr;lab811

r_-..

r

HPOS'Io'01l.cbtll"

r_

r_....""""'•

• Of

iJ 1AU.
~ 1AU.

1AU.

Ooy

.:..l:J

""'''~.:..l:J

V...

.:..l:J

iJ
iJ

Ooy .....,..,..t!OIlO

2

iJ

1AU.

iJ
iJ

1AU.

Aoco

0-000

1AU.

_.
(kindtr

iJ
Aoco

iJ

Ofndtr

1AU.
,.".

1AU.
Ro....

1AU.

iJ
iJ
iJ
iJ

1.1 Navigating the CEDView Main Window
There are three primary methods of interacting with the CEDView program:
1. Mouse Commands used in CEDView are typical of those found in other, well-known,
software packages. These commands include: Clicking the mouse buttons, clicking
mouse buttons with keyboard keys pressed, and dragging the mouse. Dragging the
mouse involves moving the mouse with one of the mouse buttons held down.
2. Pop-up Panels are sub-windows that appear inside of the Main Window. They can be
relocated or hidden from view by the user. These panels serve a multitude of purposes.
They can: Change the way the map looks, determine which types of CED events are
displayed on the map, adjust the timeline to show CED activity between specific dates,
and present detailed numerical information to users that require it.
3. The Menu Bar appears in the upper left-hand side of the Main Window, directly under
the title bar. Items under each menu heading can be revealed by clicking the left mouse
button on the heading. To pick a revealed item, click it with the left mouse button.
1.2 Map viewing
Map viewing is accomplished through the use of a “virtual camera” which, in many ways, is
analogous to a film or video camera. Associated with the camera is a “focal point” at which it
always looks. When the application starts, the focal point is at the center of the map. The
camera can be positioned and oriented through four primary camera controls as follows:
1. Aiming changes the location of the focal point. To aim the camera, place the mouse
over the desired location on the map. Next, click the right mouse button with the control
key pressed. The focal point will automatically update to the new location.
2. Zooming allows the camera to move closer to, or further away from, the focal point. To
adjust the zoom (range), drag the mouse forwards and backwards with the right mouse
button pressed and the shift key held down.
3. Panning allows the camera to rotate horizontally (azimuth) about the focal point. To pan
the camera, drag the mouse to the left and right with the right mouse button pressed.
4. Tilting lets the camera rotate vertically (elevation) about the focal point. To tilt the
camera, drag the mouse forwards and backwards with the right mouse button pressed.

~?[0J

.....ft>nPC...-

~-..MI:

3

1.3 Key map cells
The map imagery is superimposed on top of a grid of tiles called key map cells. Each cell
represents a ¾ mile by ¾ mile area. The Harris County region covers a grid of approximately
84 (east - west) by 66 (north - south) of these cells.
1.4 CED incidents
Centered on each cell is a three-dimensional bar, analogous to those seen in two-dimensional
bar charts. The grid, collectively, represents a bar graph spatially distributed across Harris
County. The height of each bar is determined by the number of CED incidents that occurred
geographically inside the cell. Each bar has one or more tick-marks along its sides. These
marks indicate the current number of incidents tallied for that cell.
1.5 Selection
Users can also get detailed information about CED incidents in a cell via Selection. To make a
Selection, point the mouse at the top of a bar on the map. Then, click the left mouse button
while pressing the control key on the keyboard. Key map cells are the smallest selectable area
on the map.
1.6 Quick reference guide
To access an on-line quick-reference guide for viewing and selection controls, choose About
from the Help menu.

~?[0J

.....ft>nPC...-

~-..MI:

4

2. MENUS
2.1 File Menu
The File Menu allows the user to perform CED database input and output related tasks. Figure
2 has an image of the File Menu which has six items:
Figure 2 CED File Menu

Load Data..
Generate Data
Save Snapshot..
ExportCSV..
Print..
Quit

2.2 Load Data
Selecting Load Data brings up an Open File dialog box that allows the user to select a CED
database to load. The default file, ced_0208.ced, can be found in the data directory of the
standard distribution folder. This file is packed and encrypted in a secure format that can only
be read by the CEDView software.
2.3 Generate Data
Should a CED database be unavailable, the Generate Data item can be selected to have the
application randomly fabricate a set of representative CED events. This item allows for a
demonstration of the software capabilities without requiring the disclosure of sensitive HPD
information.
2.4 Save Snapshot
The user can capture the current contents of the Main Window to a Windows Bitmap image file
(BMP) by selecting Save Snapshot. In addition to saving the map, all open panels are saved in
the capture as well if the video display driver supports it. A panel not needed in the output
image file can be closed and hidden from display. The saved bitmap image can be printed,
shared via e-mail, or used with other software for further processing/analysis.
2.5 Export CSV
Export CSV allows the user to save a Comma Separated Value (CSV) file containing an
unencrypted, human-readable, report of all records currently loaded into the program. It will
export reports for databases loaded with the Load Data menu item as well as databases
generated with the Generate Data menu item.

5

2.6 Print
The Print menu item captures exactly the same information as Save Snapshot but sends the
image directly to a printer instead of to a file. The software will automatically scale the image to
the printer resolution. In most cases, however, printing in landscape mode results in a better
quality hardcopy.
2.7 Quit
The program can be terminated at any time via the Quit menu item.
Location, Time, and Query Menus
Figure 3 is an illustration of the Location Menu, Time Menu, and Query Menu. The intent of the
three menus is primarily to provide support for a user who is using the tool as a presentation
platform. The idea is to have several events of interest ready to be called upon in rapid
succession, assisting in situations where a presenter must provide a great deal of information in
a short time.
Figure 3 Location Menu, Time Menu, and Query Menu

2.8 Location menu
The Location Menu allows a user to set the “virtual camera” focal point to a predefined location
on the map. Selecting Default Location restores the view to the center of Harris County. Saved
Location is a temporary placeholder for the first user-defined location (currently undefined).
2.9 Time menu
The Time Menu gives a user the ability to skip forwards or backwards to preset times in the
timeline. Choosing Default Time resets the current time to the first time in the currently loaded
CED incident database. Saved Time is a temporary placeholder for the first user-defined time
(currently undefined).
2.10 Query menu
The Query Menu allows a user to rapidly set the fields of the Query Panel to a predetermined
query. Picking Default Query sets each of the fields in the Query Panel to the value ALL, thus
returning it to its default state. Saved Query is a temporary placeholder for the first user-defined
query (currently undefined).

6

3. PANEL MENU
The Panel Menu allows the user to show panels that are currently hidden from view. Figure 4 is
an image of the Panel Menu which has four items.
Figure 4 Panel Menu

Show GIS Panel
Show Query Panel
Show Timeline Panel
Show Info Panel

By default, all panels except for the Info Panel are visible when the program starts. To hide a
panel, click the small red box ( ) in the upper right hand corner of the panel.
3.1 Show GIS Panel
To reopen the GIS Panel select Show GIS Panel from the Panel Menu.
3.2 Show Query Panel
To reopen the Query Panel select Show Query Panel from the Panel Menu.
3.3 Show Timeline Panel
To reopen the Timeline Panel select Show Timeline Panel from the Panel Menu.
3.4 Show Info Panel
To reopen the Info Panel select Show Info Panel from the Panel Menu, or make a Selection with
the mouse.
Each panel can be moved around inside the Main Window. In many cases, panels must be
moved to uncover parts of the map they obscure. To drag a panel, place the mouse over the
panel name and move the mouse while holding down the left mouse button. For a layout that
works well, place GIS Panel in the upper right, the Query Panel in the lower right, and the
Timeline Panel in the lower left of the Main Window. Panels will remember their positions when
hidden.

7

4. GIS PANEL
The GIS Panel allows the user to modify the map displayed in the Main Window. Figure 5 is an
image of the GIS Panel which consists of three major components: The Overlay check-box
group, the Map drop-down menu button, and the Selection drop-down menu button. By default,
the panel is configured to show the base-map with major roads, highways, and highway labels
displayed.
Figure 5 GIS Panel
GIS Panel

~
Mop

iJ

IBasemap

-Overlay-

r
r
P-

r
PP-

r
r
r
r

Zip Codes
Zip Code Labels
Major Roads
Major Road Labels
Highways
Highway Labels
HPO Stations
HPO Station Labels
Hospitals
Hospital Labels
Selection

iJ

IKeymap

4.1 Overlay check-box group
As seen above, the Overlay check-box group allows the user to add additional map information
to either the base-map or City Council District map. The user can add a feature to the map by
clicking the left mouse button on its check-box. Currently supported features include: ZIP codes
and ZIP code labels, major roads and their labels, highways and their labels, HPD station
locations and their labels, and finally, hospital locations and their labels. The term “label” is
synonymous with “name”.
4.2 Map drop-down menu button
The Map drop-down menu button, shown in Figure 6, allows the user to select between a basemap and a City Council District map. The base-map shows the major topographic features of
Harris County. These features include: the county boundary, major bodies of water, and the
urban areas of the City of Houston. The City Council District map contains the same features as
the base-map, but also includes labeled and color coded Houston City Council Districts.
Figure 6 Map drop-down menu button
Map
ISasemap

'4141$111.,".
[City Council

8

iJ

I

4.3 Selection drop-down menu button
The Selection drop-down menu button allows the user to choose the type of highlighting they
want to see on the map when a selection is made. The drop-down menu button has three
options: Key map, City Council District, and ZIP code. Figure 7 shows the Selection drop-down
menu button. When a key map cell, City Council District or ZIP code is selected, it will appear
darker on the map.
Figure 7 Selection drop-down menu button
Selection

::iJ

IKeymap
City Council
Zi Code

4.4 Keymap Selection
A Key Map selection highlights only the key map cell selected by the user. Key map is the
default highlighting method.
4.5 City Council Selection
City Council District selections highlight all of the key map cells in the selected City Council
District.
4.6 Zip Code Selection
ZIP Code selections highlight all of the key map cells in the selected Zip Code.

5. QUERY PANEL
The Query Panel provides the user with options for selecting which CED events are displayed
on the map. An image of the Query Panel is illustrated in Figure 8. The Query Panel has four
major sections: The Shift and Division drop-down menu buttons, the Outcome group, the Officer
group, and the Suspect group.
Figure 8 Query Panel
~

Query Panel
Officer

Shift
IALL

IALL

Division

#01

:.:OJ

IALL

Race

:.:OJ

IALL

,suspect

#01

:.:OJ

:.:OJ

IALL

IALL

Gender

:.:OJ

IALL

IALL

IALL

:.:OJ

IALL

IALL

9

:.:OJ

~ Reoson

Response

:.:OJ

:.:OJ
Age

Successful

~lnJu'Y

:.:OJ

IALL

Gender

ro~~.~

:.:OJ
Race

:.:OJ

IALL

:.:OJ

There are several special values than can be selected in the drop-down menu button items of
the query panel. A field with value “ALL” will match any value for that field in the database of
CED incidents. Its use is to indicate that the user has no particular preference for a field. It is
also the default value for all fields in the panel. The value “Dog” is used when a dog or other
animal is involved in a CED incident. Fields with a value “Duplicate” imply that a CED event
was recorded more than once in the database. This is due to the fact that some events involve
multiple officers reporting the deployment of a CED. Finally, the value “Officer” is used when an
officer is involved as the suspect in a CED incident.
5.1 Shift drop-down menu button
The Shift drop-down menu button, shown in Figure 9, allows the user to select CED events
based on the shift in which they occurred. It has three possible values: Day, evening and night.
Figure 9 Shift drop-down menu button
Shift

5.2 Division drop-down menu button
The Division drop-down menu button, shown in Figure 10, lets the user restrict displayed events
to only those involving officers from one of 19 HPD divisions.
Figure 10 Division drop-down menu button
Division

Airport
Central
Clear Lake
Travis
Eastside
Fondren
Kingwood
Narcotics
North Division
Northeast
Northwest
South Ce ntral
Southeast
Southwest
SpecialOps
SWAT
Traffic
Westside
X-Job

10

5.3 The Outcome group
The Outcome group provides a method for filtering displayed events based on their outcome.
The Outcome group contains two drop-down menu buttons, depicted in Figure 11
Figure 11 Outcome group drop-down menu buttons
Injury

:=oJ

IALL

~.
~~s

:=oJ

IALL

Successful

I

None
Officer
Suspect
Both

5.4 Successful drop-down menu button
Incidents where a CED unit was successfully deployed, or where one was not, can be seen
using the Successful drop-down menu button.
5.5 Injury drop-down menu button
The Injury drop-down menu button provides a way to view instances where either an officer,
suspect, or both sustained injury due to the use of a CED.
5.6 The Officer group
The Officer group allows a user to display events based on the demographics of the officer(s)
involved. The Officer group contains four drop-down menu buttons, shown in Figure 12.
Figure 12 Officer group drop-down menu buttons
#Of
Race
Response
1

Gender

2
3
~

5
Duplicate

Asian
Black
Hispanic
White

IALL

:=oJ

Male
Female

5.7 # Of drop-down menu button
The # Of drop-down menu button selects incidents by the total number of officers present at a
scene where a CED was used.
5.8 Race drop-down menu button
The race of the officers(s) involved in a CED event can be selected via the Race drop-down
menu button.

11

5.9 Gender drop-down menu button
The Gender drop-down menu button filters events based on the gender of the officer(s).
5.10 Response drop-down menu button
CED events in which the officer(s) initiated action, was/were dispatched to the scene, or both
can be viewed by changing the Response dropdown.
5.11 The Suspect group
The Suspect group offers a way of choosing events to display based on the demographics of
the suspect(s) involved. The Suspect group consists of five drop-down menu buttons depicted
in Figure 13.
Figure 13 Suspect group drop-down menu buttons
#01

Age
Race

IALL

1
2
3

Asian
Black
Hispanic
White
Dog
Officer

~

5
9
Dog
Officer

Gender

(= 16
17-22
23-28

29-3~

Male
Female
Dog
Officer

35-~2

~3-~9

50-75
Unknown

Reason

Combat/Resist
Threat/Weapon
Verbaltphysical
Felony Suspect
Dog
Accident

The # Of, Race, and Gender drop-down menu buttons serve the same purpose for suspects as
they do for officers in the Officer group.
5.12 Age drop-down menu button
The Age drop-down menu button groups suspects involved in CED incidents into one of eight
age ranges.
5.13 Reason drop-down menu button
The Reason drop-down menu button filters incidents based on the reason provided by the
officer(s) for deploying CED devices on the suspect(s).

12

6. TIMELINE PANEL
The Timeline Panel allows users to dynamically display CED incidents that occur between two
specific points in time on the map. These two times are known as the “start time” and the
“current time”. By default, the start time is set to the first time present in the database. Figure
14 is an image of the Timeline Panel, which has three primary areas of interest: The Time/Date
Display, the Tape Deck, and Fine-Tuning Controls.
Figure 14 Timeline Panel
~

Timeline Panel

---.:.JII_}~I !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!~~

--.lJ..J

Shih -.:.L.:.J

Day -.:.L.:.J

Month -.:.L.:.J

Year-.:.L.:.J

Oay.January2.2000

6.1 Time/Date display
The Time/Date Display is visible in the lower right-hand corner of the Timeline Panel. It shows
the current date and time the software is using to select CED incidents for display on the map.
It has the following format: Shift, month, day, and year. The current date and time are updated
by user interaction with the mouse, Tape Deck or Fine-Tuning Controls. The Time/Date Display
will rollover if the user attempts to set it to anything beyond the night shift of December 31,
2031. It will also roll-under if the user attempts to set it earlier than the day shift of January 1,
2000.
The user can control forward and reverse playback with the mouse. Dragging the mouse to the
left, with the left mouse button down in the main window, controls reverse playback. To drive the
timeline forward, drag the mouse to the right in the main window with the left mouse button
down.
6.2 Tape deck
The Tape Deck derives its name from its similarity to an audio/video cassette player. It consists
of: a set of three radio-buttons, a Timeline Indicator bar, and a variable Playback Rate dropdown menu button.
6.3 Radio buttons
The three radio-buttons are mutually exclusive and selecting one will de-select either of the
other two. By default, the Pause Button ( ) is selected indicating that playback of CED
incidents is halted. Click the Forward Button ( ) to play events forwards from the current time.
Click the Reverse Button ( ) to play events backwards from the current time.

13

6.4 Timeline indicator bar
The Timeline Indicator bar depicts, for the user, where in the timeline the current time/date falls.
The user has reached the end of the timeline when the blue bar extends all of the way across its
gray container, as shown in Figure 15.
Figure 15 Timeline Indicator bar

6.5 Playback rate drop-down menu button
Playback will occur at the rate specified in the Playback Rate drop-down menu button depicted
in Figure 16. The drop-down has 4 possible rates: Shift, day, month, and year. The default
value is day, and can be changed on-the-fly while either forward or reverse playback is
occurring. The Playback Rate drop-down menu button also sets the playback rate for mousecontrolled playback.
Figure 16 Playback rate drop-down menu button
IDay

iJ

6.6 Fine-tuning controls
The Fine-Tuning Controls are provided so that users can make very specific adjustments to the
timeline. They consist of a set of: Increment Buttons ( ), Decrement Buttons (L.:J), and a Range
Button ( [.. ] I).
There are four separate increment/decrement button pairs, one for each component of the
current time/date. To move the shift, day, month or year forward by one unit, click on the
individual Increment Buttons. To move them backwards by one unit, click on the appropriate
Decrement Buttons.
Only CED incidents occurring between the start time and the current time are considered for
display on the map. The user can click the Range Button to set the start time to the current time
indicated in the Time/Date Display. When done, only CED events occurring on or after the new
start time are considered.
The Timeline Indicator bar will turn orange if the user sets the current time to a time earlier than
the newly selected start time. To reset the start time to the first time in the database, select
Default Time from the Time Menu.

14

7. INFO PANEL
The Info Panel provides detailed information about each selection made by the user. If the
panel is not visible when the user makes the selection, it will automatically be displayed. Figure
17 illustrates an image of the panel which has three components: The Text Area, the Clear
Button, and the Save Button. Information displayed in the Text Area depends on the time/date
chosen in the Timeline Panel and the specific query options enabled in the Query Panel.
Figure 17 Info Panel
~

Info Panel

Clear

Save

7.1 Text area
For each selection, a series of lines are printed into the Text Area:
1.
2.
3.
4.
5.

Two lines specifying the start time and current time of the query.
The total number of CED incidents occurring, for that query, Citywide.
The selected key map cell, along with its count and its percentage of the Citywide total.
Totals for ZIP codes contained in the key map cell, along with their percentages.
Tallies for City Council Districts present in the key map cell with their percentages.

Some key map cells contain multiple ZIP codes and/or multiple City Council Districts. As a
result, there may or may not be multiple listings for each.
Data from multiple selections are recorded in the Text Area in the order the selections were
made. When the visible portion fills with text, previous data will scroll out of view. A scrollbar
will then appear on the right-hand side of the Text Area. To see results from previous selections,
drag the scrollbar upwards with the mouse or use the arrow at the top of the scrollbar. Dragging
the scrollbar downwards, or clicking the arrow on the bottom of the scrollbar, will bring data from
more recent selections back into view. Making another selection will always cause the Text
Area to jump back to the most recently printed information.
7.2 Clear button
The Clear Button removes all selection information from the Text Area, returning it to its default
state.
7.3 Save button
The Save Button allows a user to save information currently recorded in the Text Area to a file
on disk. Both buttons can be used, in combination, to create a simple data file of CED incident
details. This file can be printed, shared via e-mail, or used with other software for further
processing/analysis.

15

8. EXAMPLE
Below is an example that is provided to illustrate some of the concepts and graphical user
interface components previously described in detail. As can be seen in Figure 18:
•
•
•
•
•
•
•
•
•

The user has taken a 3-point perspective, close-up view centered north of City Council
District D inside Loop-610.
The City Council District map is selected in the GIS Panel.
Highways and highway labels are the only overlay items in use.
Key map is the current selection mode.
The current query involves all shifts of the South Central division.
The user is looking for any events in which a Hispanic, female Officer has deployed a
CED on an Asian, male suspect 16 years old or younger.
The current timeline under consideration ranges from the evening of January 4, 2000 to
the evening of January 14, 2031.
Key map cell 533D has been selected which contains ZIP code 77004 and is part of City
Council District D.
There was only one CED event that matched all of these criteria indicated in the Info
Panel

~?[0J

.....ft>nPC...-

~-..MI:

16

Figure 18 Example of output from CEDView software
•

~ ;131 JX I

CEDVleoW 1 Oc:

::J

ICOy 0>",,,'
Ow,."

r

LpCodOG

r '" CoOIJ l.ebot~
r UoforRoooq
r Molor Acod t..obot~
r- Hoghwoy>
r;; HIgtrw#y 1.obctl,

r
r

HPOSlc.bon'
HPO StobOn L.obtk

rHaopitolt

r

$,"

..:..I.:.J ,,-,....:..I.:.J

VIer

El.:J

EV8rwtg.JenuOly.I "-2'ClJI

17

-Ol

::J
::J IAU
"IHrDPMIC
::J
::J
c.ndlilr
::J
1-·

1""-

::J
A.tpon..
-'"y
::J IAU
::J

I··a

""com.
S>.>ccn,...
Ooy

Ho'uprtl!lllobotl)

RoC(!

lkron

IM~'

1"-1.

a._
Ago

Flo.....

::J
::J
::J
::J
::J

PART VI

PART VI
VIEWS OF RESPONSIBLE OFFICIALS

1

CITY OF HOUSTON
INTER OFFICE CORRESPONDENCE

TO,

Annise Parker
City Controller

FRO",

Harold L. Hurtt
Chief of Police

OAT<,

August 28, 2008

SU""CT,

Conductive Energy Device Audit

, In the interest of public safety senior managem'ent at the Houston Police Department applaud you
for conducting an audit of the department's Conducted Energy Device (CED) program. The
review and evaluation of the audit is currently in progress and we will consider the
recommendations very seriously. The department will continue to follow the best practices in
the law enforcement as they relate to policies. procedures and training. Any major modifications

to our CED program in the future will be reported to the City Council Public Safety Commlttee.
In December 2004, the decision was made to implement Conducted Energy Devices in the
Houston Police Department. Although the department used older generation. devices prior to
December 2004: the current model is the X-26 manufactured by Taser Intemational.

The purpose of this less-lethal technology was to provide Houston Police Officers with
additional force options to: ossist officers in securing and controlling combative individuals,
reduce injuries to officers and 'suspects during physical confrontations, reduce the potential
financial impact of civil litigation in"use of force incidents and; in limited situations,'"provide an
alternative option prior to exercising deadly force. The use of CED's bas been extremely
successful in accomplishing our goals. AC,cording to page (5) of the Mir Fox & Rodriguez
executive summary, "HPD bas been effectively managing the CED program and was generally
in compliance with HPD' and the City's Policies and Procedures, as well 'os the related'
procurement Jaws."

The Mir Fox & Rodriguez audit examined CED deployments for calendar year 2005, 2006. and
January through June of 2007. During this specific audit period, officers from the Houston
Police Department responded to 2.8 million calls for polfcie service and arrested 272,885
individuals' for various criminal violations. The percentage of criminal suspects who were
arrested and involved in a CED event amounted to .47% of the total number of suspects arrested.
In other words, an individual being arrested by a Houston Police Officer bas a chance of being
involved in a CED event less than ~ of 1% of the time during an arrest.
Pbysical confrontations between poliCe officers and suspects occur frequently ,and potentially
pose a significant fmaricial burden on the department and community. In calendar year 2005, the
department's worker'~ compensation cOst was $1,152,195.00 as a result ·of injuries officers
received during physical confrontations 'with violent suspects. During calendar year 2006, that
amount decreased to $738.082.00. In calendar year 2007, the amount continued to. decline and·

Page 1 of2

Note:

In paragraph 3 above, page 5 was changed to page 2 during the report binding process.

174

Views of Responsible
Officials

Annise 'Parker, eity Controller

2

August 28, 2008

totaled $148,929.00. Since the .deployment of CEO devices. in the Houston Police Oepartinent,
worker's compensation expenditures have continued to decline. Furthermore, during the same
auditing period, no suspect has been seriously injured during an incident where the injury was
caused by the CEO device. The reductioll in injuries and the associated fmancial cost have been
a positive benefit to the department and community.
At the Houston Police Department we view any use of force by its officers as serious and have
instituted policies, procedures and training to ensure accountability. Houston Police Officers
have conducted themselves very responsibly exercising restramt and excellent judgement during
.53 CEO incidents where deadly force was clearly justified by stale law as well as department
policy. During these 53 incidents the suspects were armed with deadly weapons that included
but not limited to, frrearins, knives and other weapons. Officers successfully resolved these 53
incidents using a less lethal CED instead of deadly force. Furthennore, the department's policies
follow the best practices in law enforcement across the country, including the International
'Assoelation of Chiefs of Police (IACP) and the Police Executive Reseilrch Forum (PERF).
In an effort to address community concerns, management at HPD has been diligent in seeking
the physiological effects associated with Conducted Energy Devices. A mediCal review of all
. published physiological effects available at the time (2008), was gathered by medical doctors
from the University of San Diego and presented to the Houston PoHce Department and the City
Council Public Safety Committee. Senior management has also reviewed. similar studies
conducted by the Wake Forest University Baptist Medical Center study released in 2007 and the
National Institute of Justice (NIJ) study published in June 2008. In all studies an expert panel of
medical professionals found no conclusive evidence of high risk of death or serious injury from
the ·direct effects of Conducted Energy Devices used against individuals. The studies did
recognize high risk groups such as the elderly, children, pregnant women and individuals who
may be in a state of excited deliriUm. Department policy and traiJiing give officers the proper
guidance and discretion in dealing with high risk groups. According to the studies risk of
injuries from CED devices are usually low and result from secondary falls. These-injuries are
minor scrapes and bruises. We will continue to evaluate all relevant infonnation associated with
CEO technology. As stated byMir Fox & Rodriguez, HPD has been effectively managing the
CED program.

Views of Responsible
Officials

Chief of Police
hlh:ces
cc:

M~chael Moore, Mayor's Office

COP 108-30056

Page 2 of 2

175