Nov. 2017, CCJP 2017-2

 

The Las Vegas Body-Worn Camera Experiment:
Research Summary
By Anthony A. Braga,a James R. Coldren, Jr.,b William H. Sousa,c Denise Rodriguez,b and Omer Alper b
Introduction
The past several years have seen an increase in
body-worn cameras (BWCs) on police as agencies
around the U.S. have acquired the technology for
their officers. BWCs on officers enjoy a great deal
of public support and are generally believed to be a
mechanism for reducing police use of force,
reducing complaints of officer misconduct, and
enhancing the transparency of policing. However,
while BWCs offer several potential advantages for
both police and the public, empirical research into
their value is still limited. Questions remain in terms
of the impact of BWCs on police practice and their
influence on police-citizen relationships.
In 2014, the Las Vegas Metropolitan Police
Department (LVMPD) became one of the first large
police agencies in the United States to begin
equipping its officers with BWCs. The National
Institute of Justice (NIJ) provided funding for the
first 200 of these BWCs as part of a research study
designed to determine the impact of the technology
on a variety of outcomes, including officer use of
force, complaints of officer misconduct, and officer
discretionary activities. The study also provided a
cost-benefit analysis associated with the utilization
of BWCs.
This Research in Brief summarizes information
from three products of the Las Vegas BWC
Experiment: the study’s final report presented to
NIJ (Braga, Coldren, Sousa, Rodriguez, and Alper,
2017); an article that discusses the study’s
methodology and implementation challenges
(Sousa, Coldren, Rodriguez, and Braga, 2016); and
an article that addresses key elements of the
impact evaluation (Braga, Coldren, Sousa, and

• Officers with body-worn cameras had fewer
complaints of misconduct than their control
group counterparts. The percent of officers with
body-worn cameras that generated at least one
complaint decreased from 54.6% to 38.1%. By
comparison, the percent of officers in the
control group that generated at least one
complaint decreased from 48.0% to 45.5%.
Officers with body-worn cameras had fewer
incidents of uses of force than their control
group counterparts. The percent of officers with
body-worn cameras that generated at least one
report for use of force decreased from 31.2% to
19.7%. By comparison, the percent of officers
in the control group that generated at least one
report for use of force increased from 26.3% to
27.3%.
Although there were few differences between
body-worn camera officers and control group
officers in terms of responses to dispatched
calls, officer-initiated stops, or responses to
crime incidents, officers with body-worn
cameras issued slightly more citations and
made slightly more arrests.
The costs of body-worn cameras were
estimated to be between $828 and $1,097 per
user per year. The cost savings as the result of
fewer complaints of misconduct and fewer
resources spent on misconduct investigations
were estimated at $4,006 per user per year.
The net savings associated with body-worn
cameras were therefore estimated to be
between $2,909 and $3,178 per user per year.

a

School of Criminology and Criminal Justice, Northeastern University
The CNA Corporation
c
Center for Crime and Justice Policy, University of Nevada Las Vegas
b

© 2017 University of Nevada, Las Vegas 

  

 

Rodriguez, In Press).
Research Objectives
In addition to examining the BWC implementation
process in Las Vegas, the research for the Las
Vegas BWC Experiment had two principle
objectives: conduct an impact evaluation using a
“randomized controlled trial” (RCT) with a sample of
LVMPD officers, and conduct an evaluation of the
costs and benefits of implementing BWCs in Las
Vegas. The impact evaluation was specifically
concerned with whether BWCs influenced citizen
complaints of officer misconduct, officer use of
force incidents, and officer discretionary activities
such as citations issued and arrests made. The
primary research question for the cost-benefit
analysis asked how the possible financial benefits
of BWCs (i.e., potentially lower labor costs
associated with investigating fewer complaints of
misconduct) compared to the financial costs of
acquiring and equipping officers with the
technology.
Impact Evaluation – Design
The design for the impact evaluation involved an
RCT that called for LVMPD patrol officers to be
randomly assigned into one of two groups: an
“experimental” group comprised of officers who
wore BWCs, and a “control” group comprised of
officers who did not wear BWCs but who served as
comparisons for those in the experimental group.
According to the principles of random assignment,
RCTs produce two equivalent groups prior to the
introduction of the experimental condition (in this
case, BWCs). Therefore, any differences at the end
of the experiment between the two groups on the
key outcome variables can be attributed to the
BWCs worn by the experimental group officers.
Recruitment for the sample of officers began in
February 2014. Obtaining a sufficient sample of
volunteers to accommodate the research design
involved overcoming a number of technical,
administrative, and personnel challenges.

2   B W C

Experiment

 

Nevertheless, by September 2014 the final sample
had been obtained, consisting of 416 randomly
assigned officers: 218 in the experimental group and
198 in the control group.1
Table 1 presents information on the sample.
Importantly, the experimental group did not differ
significantly from the control group on key
demographic variables or on prior complaints of
misconduct. This suggests that the randomization
procedure did indeed produce two balanced groups.
Also, Table 1 suggests that the sample of officers in
the study was generally representative of the larger
population of LVMPD patrol officers. The main
difference between the study participants and the
remainder of the patrol division related to officer
rank, where there was a slight overrepresentation of
sergeants in the sample – the result of a recruitment
strategy that encouraged sergeants to volunteer for
the study in an effort to lead by example.2
Officers in both groups were monitored for a oneyear period on the key outcome variables, including
use of force, complaints of misconduct, arrests
made, and citations issued. Two potential concerns
were also monitored during this time. The first
related to attrition – the idea that officers might drop
out of the study for various reasons, thereby
resulting in unbalanced groups and / or impacting
the statistical power of the study. The second
potential concern was the possibility of
contamination – the idea that officers in the
experimental group might interact with control group
officers and influence their decisions, such as when
a BWC officer and a control officer respond to the
same call for service. Analyses ultimately
determined that both attrition and contamination had
minimal impacts on the results of the study.3
Impact Evaluation – Results
The impact evaluation first considered whether
BWCs had an impact on citizen complaints of police
misconduct.

  

TABLE 1: Randomized Controlled Trial Sample
Experimental
n=218

Control
n=198

Study Participants
n=416

Remainder of Patrol
n=955

Percent Male
Percent Female

92.1%
7.9%

91.5%
8.5%

91.8%
8.2%

90.1%
9.9%

Percent White
Percent Hispanic
Percent Black
Percent Asian / Other

73.2%
12.6%
7.9%
6.3%

69.7%
15.4%
9.6%
5.3%

71.5%
14.0%
8.7%
5.8%

71.3%
14.7%
6.4%
7.6%

Average Age
Average Years on Job

36.3
8.7

37.7
9.9

36.9
9.3

35.8
8.6

Percent Patrol Officer
Percent Sergeant

86.4%
13.6%

83.0%
17.0%

84.7%
15.3%

93.0%
7.0%

.94

.86

.89

.96

Average Complaints
* Adapted from Braga et al. (2017)

DIAGRAM 1: Percent of Officers with at Least One Misconduct Complaint / Use of Force Incident
60%

55%
48%

50%

pre-intervention

46%

post-intervention

38%

40%

31%
26%

30%

27%

20%

20%
10%
0%

experimental

control

COMPLAINTS OF MISCONDUCT

 

experimental

control

USE OF FORCE INCIDENTS

 B W C

Experiment

   3  

  

 

DIAGRAM 2: Average Number of Citations Issued / Arrests Made Per Officer Per Month
12

11.08

10.53

10.27

10.65

pre-intervention

10

post-intervention

8

6.87

7.30

7.04

7.12

6
4
2
0

experimental

control

CITATIONS ISSUED

Diagram 1 compares the experimental group and the
control group in terms of complaints of police
misconduct. The percent of officers in the experimental
group that generated at least one complaint decreased
from 54.6% pre-intervention to 38.1% during the
intervention – a difference of 16.5% and an overall
percent reduction of 30.2%. By comparison, the
percent of officers in the control group that generated
at least one complaint decreased from 48.0% preintervention to 45.5% during the intervention – a
difference of only 2.5% and an overall percent
reduction of only 5.2%. These results suggest a
significant difference between officers with BWCs and
those without BWCs in terms of citizen complaints of
misconduct (Braga et al., In Press; Braga et al., 2017).
Diagram 1 also compares the experimental group and
the control group in terms of use of force incidents. The
percent of officers in the experimental group that
generated at least one use of force report decreased
from 31.2% pre-intervention to 19.7% during the
intervention – a difference of 11.5% and an overall
percent reduction of 36.9%. The percent of officers in
the control group, however, that generated at least one
use of force report actually increased from 26.3% preintervention to 27.3% during the intervention – a
difference of 1.0% and an overall percent increase of

4   B W C

Experiment

 

experimental

control
ARRESTS MADE

3.8%. Similar to the analysis of citizen complaints
of misconduct, these results suggest a significant
difference between officers with BWCs and those
without BWCs in terms of use of force incidents
(Braga et al., In Press; Braga et al., 2017).
The impact evaluation was also interested in
whether BWCs impact officer activity levels.
Analyses revealed few differences between
experimental and control officers in terms of
responses to dispatched calls, officer-initiated
stops, or responses to crime incidents. There
were differences, however, in terms of citations
issued and arrests made.
Diagram 2 compares the experimental group and
the control group in terms of citations issued and
arrests made. The number of citations issued by
officers in the experimental group increased from
an average of 10.27 per officer per month before
the intervention to an average of 11.08 per officer
per month during the intervention – an increase of
7.9%. The number of citations issued by officers
in the control group increased from an average of
10.53 per officer per month before the
intervention to an average of 10.65 per officer per
month during the intervention – an increase of

  

just 1.1%. This suggests a significant difference
between officers with BWCs and those without BWCs in
terms of issuing citations (Braga et al., In Press; Braga
et al., 2017).
Similarly, the number of arrests made by officers in the
experimental group increased from an average of 6.87
per officer per month before the intervention to an
average of 7.30 per officer per month during the
intervention – an increase of 6.3%. By comparison, the
number of arrests by officers in the control group
increased from an average of 7.04 per officer per
month before the intervention to an average of 7.12 per
officer per month during the intervention – an increase
of only 1.1%. This also suggests a significant
difference between officers with BWCs and those
without BWCs in terms of arrest activities (Braga et al.,
In Press; Braga et al., 2017).
Cost-Benefit Analysis
The cost-benefit analysis considered the annual
financial costs and benefits per BWC user. The costs
were estimated using FY 2014 amounts for BWC
installation, training, operation, maintenance, FOIA
requests, and video storage costs, including one-time
and recurring costs. Using these data, the total cost
incurred for BWC implementation was $1,097 per BWC
user per year.4 More recent invoices, however,
suggested a slight decrease in costs for BWCs and
storage, bringing the amount down to $828 per user
per year. Considering this adjustment, the analysis
therefore estimated that BWCs cost between $828 and
$1,097 per user per year.
The financial benefits of BWCs were derived primarily
from the estimated decrease in complaints of
misconduct as the result of BWCs, the reduced cost to
investigate these complaints, and the reduced amount
of time it takes to resolve complaints when video
evidence is available.
There may be additional financial benefits, such as

 

fewer and / or lower court settlements arising
from citizen complaints of misconduct, but the
inclusion of these data were not feasible given
the timeframe of the study and the length of time
required to resolve court proceedings.
Table 2 estimates the labor costs associated with
investigating an average complaint of officer
misconduct with and without BWC video information.
When considering the investigator’s modified hourly
wage5 and hours spent investigating a complaint of
misconduct, considerable cost savings are realized
when BWC video is available. Rather than a
combined 91 hours of investigative time costing
$6,776 without BWCs, the estimate is slightly over 7
hours of investigative time costing $554, for a
difference of over $6,200 per complaint of
misconduct.
The average number of complaints of misconduct
during the intervention phase of the impact
evaluation was 0.59 for the BWC users. Based on
difference-in-difference estimates, had the BWC
users experienced the same post-intervention
change as the control group, their average would
have been 0.84 complaints per officer. Table 3
considers these averages along with the estimated
costs of investigating misconduct complaints with
and without BWCs. BWCs result in fewer complaints
of misconduct – and when complaints do occur with
BWCs, the average costs of the investigations are
much less. As a result, BWCs are associated with
substantially less investigative costs per officer per
year.
Given the cost of BWCs to be between $828 and
$1,097 per user per year – and given the estimated
savings of $4,006 per officer per year as the result
of fewer misconduct complaints / less expense
associated with misconduct investigations – the net
annual savings generated by BWCs is estimated to
be between $2,909 and $3,178 per user per year.

 B W C

Experiment

   5  

  

 

TABLE 2: Labor Costs per Misconduct Complaint Investigation

Detective
Sergeant
Lieutenant

Modified Hourly
Wage

Hours Spent on
Complaint with
BWC

Investigation
Cost with
BWC

Hours Spent on
Complaint
without BWC

Investigation
Cost without
BWC

$71.52
$89.40
$107.28

6
1
0.33

$429
$89
$35

80
7
4

$5,721
$626
$429

7.33

$554

91

$6,776

 

 

 

 

Total
 

 

* Adapted from Braga et al. (2017)

TABLE 3: Investigation of Misconduct Complaint Costs, With and Without BWCs

Annual complaint investigations per officer
Average cost per investigation
Percentage cleared based on BWC alone
6
Annual investigation costs per officer

With BWC

Without BWC

0.59
$554
66%
$1,686

0.84
$6,776
0%
$5,692

$4,006

 

 

 

 

Cost Avoidance per user
per year

* Adapted from Braga et al. (2017)

Discussion
The results of the impact evaluation’s randomized
controlled trial indicate that BWCs were associated
with substantial reductions in complaints of officer
misconduct and police use of force incidents. These
results are consistent with the perceived benefits of
the BWC technology and support the notion that
BWCs can help to improve relations between police
and communities.7 When comparing BWC officers to
their control counterparts, however, it also appears
that BWC officers generated more citations and
arrests. Further research is needed to determine the
reasons for this increase in enforcement activity – or
whether this increase in activity could impact policecitizen relations.8
The results of the cost-benefit analysis suggest
substantial cost savings associated with BWCs. These
cost savings are primarily the result of the impact of
BWCs on complaints of officer misconduct. First, as
demonstrated in the impact evaluation, BWCs are
associated with fewer complaints of officer
misconduct. Second, when complaints of misconduct

6   B W C

Experiment

 

do occur, investigations into the complaints are
much shorter and require fewer resources due to
the presence of video evidence. While there are
expenses related to BWC implementation, training,
operation, and video storage and maintenance, the
benefits of the technology appear to offset these
costs.
Overall, the results of Las Vegas BWC Experiment
suggest the benefits of the technology. BWCs were
associated with substantially fewer complaints of
officer misconduct, fewer incidents of police use of
force, and a net cost savings. Several of the
findings – such as indications that BWCs are
associated with slightly more citations issued and
arrests made by officers – warrant further
investigation. Nevertheless, the results presented
here demonstrate the potential value of BWCs to
police and citizens.
Endnotes
1. The randomization procedure allowed for slightly
more officers in the experimental group in

  
anticipation of higher levels of attrition from that group. aware of the BWCs, are less likely to make false
allegations of police misconduct or engage
2. Due to the original technical infrastructure, four of
officers in ways that could result in use of force.
the then eight patrol area commands were better
The extent to which either of these possibilities
equipped to handle a large number of officers with
(or both) influenced the results is not known.
BWCs. Therefore, although officers in the sample
came from all eight area commands, Bolden,
8. It may be, for example, that BWC officers
Enterprise, Northeast, and Northwest area commands increase their discretionary enforcement activity
were overrepresented in the sample.
because they are more confident that video
evidence will hold offenders accountable. It may
3. Analyses revealed low levels of attrition and very
also be, however, that officers are concerned that
modest levels of contamination. Attrition (for reasons
their supervisors will review videos and hold them
such as retirement, promotion, changes of
accountable for discretionary decisions.
assignment, medical leave, and voluntary withdrawal
from the study) was approximately 12% from the
References
experimental group and 8% from the control group.
A.A. Braga, J.R. Coldren, Jr., W.H. Sousa, D.
Contamination, which was measured by the
Rodriguez, and O. Alper (2017). “The Benefits of
percentage of calls for service that involved one or
Body-Worn Cameras: New Findings from a
more experimental and control group officers,
Randomized Controlled Trial at the Las Vegas
averaged approximately 19% per month over the
Metropolitan Police Department.” Final report to
course of the study. Although additional analyses
suggest that the impacts of attrition and contamination the NIJ, 2013-IJ-CX-0016, September 2017.
were minimal, they remain potential limitations of the
A.A. Braga, W.H. Sousa, J.R. Coldren, and D.
study design (see Braga et al., 2017).
Rodriguez (In Press). “The effects of body worn
cameras on police activity and police-citizen
4. For an itemized list of BWC implementation costs
encounters: A randomized controlled trial.”
and the specific procedure to estimate the cost per
Journal of Criminal Law and Criminology, 108 (3).
user per year, see Braga et al. (2017).
5. Modified hourly wage considers wages, holiday and
leave costs, taxes and fringe benefits.
6. Annual investigation costs per BWC officer was
determined via the following formula: a = (0.59
“annual complaint investigations per BWC officer” x
$554 “average cost per video review investigation” x
66% “percent cleared based on BWC video alone”) +
b = (0.59 x ($554 + $6,776) “average cost per video
review investigation + usual investigation cost” x 34%
“percent not cleared based on BWC video”).
7. This study does not explore why BWCs are
associated with fewer complaints of misconduct or
fewer uses of force. One possibility is that officers are
altering their behavior knowing that their actions are
being recorded. Another possibility is that citizens,

 

W.H. Sousa, J.R. Coldren, D.Rodriguez, and A.A.
Braga (2016). “Research on body worn cameras:
Meeting the challenges of police operations,
program implementation, and randomized
controlled trial designs.” Police Quarterly, 19 (3).
Acknowledgments
This project was supported by Award No. 2013IJCX-0016, awarded by the National Institute of
Justice, U.S. Department of Justice. The
opinions, findings, and conclusions or
recommendations expressed in this publication
are those of the authors and do not necessarily
reflect those of the DOJ or of the LVMPD.
The authors would like to thank LVMPD for its
assistance with this project.

 B W C

Experiment

   7  

  

 

CENTER FOR CRIME AND JUSTICE POLICY
RESEARCH IN BRIEF SERIES
 
The Research in Brief series is produced by the
Center for Crime and Justice Policy at the University
of Nevada, Las Vegas. The Center is housed in the
Department of Criminal Justice, which is located in the
Greenspun College of Urban Affairs. Research in
Briefs are modeled after the Bureau of Justice
Statistics' Special Reports and Bulletins. The Briefs
provide summaries of various criminal justice system
practices in Nevada over time, and highlight
differences between Nevada and other states. These
reports cover all aspects of the criminal justice
system, including trends in crime and arrests, police
practices, prosecution, pretrial activities, adjudication,
sentencing, and corrections. Although Research in
Briefs typically focus on criminal justice issues within
Nevada, these reports may focus on national issues
as well.
Research in Briefs are designed to provide members
of the general public, local officials, community
organizations, and media outlets a concise and
objective profile of current crime and criminal justice
trends in Nevada and elsewhere that may serve as a
foundation for informed discussions of future crime
control policies and practices.

Questions or comments about the Research in Briefs,
information contained in this current report, or other
resources available related to this topic should be
addressed to:
William H. Sousa, Ph.D.
Director, Center for Crime and Justice Policy
University of Nevada, Las Vegas
4505 Maryland Parkway - Box 5009
Las Vegas, NV 89154-5009
Phone: 702-895-0247; Fax: 702-895-0252
Email: william.sousa@unlv.edu
Website: https://www.unlv.edu/ccjp 

8   B W C

Experiment

 

Recent Research in Briefs
(Available at www.unlv.edu/ccjp/briefs)
Aerial Drones, Domestic Surveillance, and
Public Opinion of Adults in the U.S.
Body Worn Cameras on Police: Results
from a National Survey of Public Attitudes
A Comparison of Online Sampling
Approaches for Generating Samples
The Impact of Foreclosures on
Neighborhood Crime in Nevada, 2006-2009
Nevada vs. U.S. Residents’ Attitudes toward
Surveillance Using Aerial Drones
School Violence Prevention in Nevada
Perceptions of Disorder: Results from Two
Las Vegas Tourist Locations
Prisoner Reentry in Nevada
Public Attitudes about Aerial Drone
Activities: Results of a National Survey
Regulation of the Gaming Industry across
Time and Place