Measuring The Effects Of Video Surveillance On Crime In Los Angeles (PDF) 210 Cameron

User Manual: Measuring the Effects of Video Surveillance on Crime in Los Angeles (PDF) Security Systems & Surveillance Equipment | JMAC Supply

Open the PDF directly: View PDF .
Page Count: 91

Measuring the Effects of Video

Surveillance on Crime

in Los Angeles

Prepared for the California Research Bureau

CRB-08-007

May 5, 2008

Aundreia Cameron

Elke Kolodinski

Heather May

Nicholas Williams

CONTENTS

EXECUTIVE SUMMARY ...........................................................................................................4

THE RISE OF CCTV SURVEILLANCE IN CALIFORNIA...................................................6

The Prevalence of Video Surveillance in California ...........................................................7

California Crime Rates, Video Surveillance and Spending.................................................8

Privacy, Efficacy and Public Opinion................................................................................10

Arguments against CCTV on Privacy Grounds…………………………………. 10

Arguments for CCTV on Efficacy Grounds…………………………………….. 11

Privacy versus Efficacy in California…………………………………………... 13

META-ANALYSIS OF EXISTING EMPIRICAL WORK ....................................................14

Crime Deterrence...............................................................................................................14

Crime Detection, Mitigation and Prosecution ...................................................................17

Local Characteristics of CCTV Implementation ...............................................................18

CCTV, CRIME AND POLICING IN LOS ANGELES...........................................................18

Crime and Policing in Los Angeles...................................................................................19

Gang-related Activity.............................................................................................20

The Advent of Video Surveillance ....................................................................................20

RESEARCH QUESTIONS, CASE STUDIES AND METHODOLOGY ..............................21

Research Questions............................................................................................................22

Case Selection: Hollywood Boulevard and Jordan Downs Housing Project ....................22

Methodology......................................................................................................................25

Statistical Analysis of Crime Data.....................................................................................25

Geographic Areas: Target, Buffer, Control...........................................................25

Time Period of Analysis.........................................................................................27

Measures of Crime.................................................................................................28

Statistical Tests ......................................................................................................28

Qualitative Interviews........................................................................................................29

STATISTICAL FINDINGS........................................................................................................29

Hollywood Boulevard Statistical Findings........................................................................30

Pre-Post Crime Analysis........................................................................................32

Crime Displacement Effects...................................................................................33

Pre-Post Arrest Analysis........................................................................................34

Jordan Downs Statistical Findings.....................................................................................35

Pre-Post Crime Analysis........................................................................................36

Crime Displacement Effects...................................................................................38

Pre-Post Arrest Analysis........................................................................................38

Discussion of Findings..................................................................................................................39

Data Limitations ....................................................................................................39

Disparities between the Target and Control..........................................................40

Nature of Crimes....................................................................................................41

Capacity Constraints .............................................................................................42

Increased Detection, Mitigation and Resolution of Crime....................................42

IMPLEMENTATION AND EFFECTIVENESS......................................................................43

- 1 -

Funding & Private Public Partnerships..................................................................44

Hollywood..................................................................................................44

Jordan Downs............................................................................................45

Types of Partnerships ................................................................................45

Legacy Issues and Maintenance ................................................................45

Additional Public Resources......................................................................46

Simultaneous Crime Reduction Strategies.............................................................46

Surveillance Coverage Area and Camera Placement.............................................47

Hollywood..................................................................................................47

Jordan Downs............................................................................................47

Technology and Monitoring ..................................................................................48

Camera Technology...................................................................................48

Head End Monitoring Systems...................................................................48

Future Implementation...............................................................................49

Operations, Training and Turnover........................................................................49

Camera Configuration...............................................................................49

Leadership, Training and Turnover...........................................................49

Privacy Considerations and Community Involvement ..........................................50

Station Practices........................................................................................50

Protecting Privacy Rights..........................................................................51

CONCLUSIONS FOR POLICYMAKERS...............................................................................53

REFERENCES.............................................................................................................................55

APPENDICES..............................................................................................................................64

APPENDIX A. Evidence of CCTV Crime Prevention......................................................61

APPENDIX B. Sorting the Data........................................................................................73

APPENDIX C. Statistical Analysis ...................................................................................77

APPENDIX D. Video Camera Specifications ..................................................................90

LIST OF FIGURES AND TABLES

Table 1. CA Cities with Significant Surveillance Systems..............................................................7

Table 2. CA Jurisdictions with Actively Monitored Systems..........................................................7

Figure 1. Crime in California 1996-2005 .......................................................................................9

Table 3. Causal Mechanisms of CCTV Crime Curtailment..........................................................12

Table 4. Summary of Meta-analysis Results .................................................................................16

Figure 2. Crime in Los Angeles 1997-2007...................................................................................19

Figure 3. Case Study Locations .....................................................................................................22

Table 5. Hollywood Boulevard Characteristics.............................................................................23

Figure 4. Hollywood Demographic Change..................................................................................23

Table 6. Jordan Downs Characteristics..........................................................................................24

Table 7. Methodology....................................................................................................................25

Figure 5. Hollywood Camera Locations........................................................................................26

Figure 6. Jordan Downs Camera Locations...................................................................................27

Table 8. Jordan Downs and Nickerson Gardens Demographics....................................................27

Table 9. Hollywood Average Monthly Crime Rates (pre-implementation)..................................30

Table 10. Hollywood Target and Control Area Arrests January 2003 to February 2005 (pre-

implementation).............................................................................................................................31

- 2 -

Table 11. Hollywood Monthly Crime Rates..................................................................................32

Figure 7. Hollywood Monthly Crime ............................................................................................33

Figure 8. Percentage Change in Hollywood Monthly Crime Rates Following CCTV .................34

Table 12. Hollywood Monthly Arrest Rates..................................................................................34

Table 13. Jordan Downs/Nickerson Gardens Average Monthly Crime Rates

(pre-implementation) .....................................................................................................................35

Table 14. Jordan Downs Target and Control Arrests January 2003 to September 2006 (pre-

implementation).............................................................................................................................36

Table 15. Monthly Crime Rates in Jordan Downs and Nickerson Gardens..................................37

Figure 9. Jordan Downs Monthly Crime .......................................................................................38

Figure 10. Percentage Change in Crime in Jordan Downs Study Areas .......................................38

Table 16. Jordan Downs Monthly Arrests Rates ...........................................................................39

Table 17. Implementation Successes and Challenges....................................................................43

- 3 -

Acknowledgements: Our project would not have come together without the guidance, assistance

and enthusiasm of the following individuals: Dr. Juliet Musso, Charlene Wear Simmons and

Dean Misczynski of the California Research Bureau, Commander Louis H. Gray Jr., Deputy

Chief Charlie Beck, Lieutenant Jeff Godown, Douglas Wagner, Nathan Ong, Sergeant Daniel

Gomez, Sergeant Don Graham, Officer Manuel Hernandez, Officer Jude Washington, Captain

Clay Farrell, Lieutenant Ruben De La Torre, Kerry Morrison, Sarah MacPherson, Tracey

Yasunaga, Peter Bibring, Romy Ganschow, Tamika Taylor, Pastor Mike Cummings and Chris

Weare. And special thanks to Toasted Head Wines and Nabisco Corp.

EXECUTIVE SUMMARY

This study reviews existing literature of video surveillance and measures the effectiveness of

surveillance to deter crime in select locations within Los Angeles, California. Aided by

additional funding from the federal government, private donations and the dropping cost of video

surveillance equipment, a growing number of local law-enforcement departments across

California are employing fixed video surveillance, also called closed-circuit television (CCTV),

of public space. Proponents of video surveillance argue that it may deter criminal behavior by

increasing the probability of detection and apprehension, but evidence pertaining to its

effectiveness is mixed and concerns about privacy infringement persist.

Building upon previous research of CCTV by the California Research Bureau, we first examine

44 video surveillance evaluations, primarily conducted in the United Kingdom, that investigate

the technology’s deterrent effect on crime. Following deployment of CCTV, 41 percent of these

evaluations reported a statistically significant reduction in crime, 43 percent reported no

statistically significant effect (increase or decrease) on crime, and 16 percent showed a

significant crime increase. Importantly, none of the five evaluations of CCTV conducted in the

United States found a statistically significant crime reduction. In addition, few empirical studies

have investigated the utility of video surveillance in enhancing law enforcement’s ability to

detect crimes or mitigate harm after it occurs. Nor did our meta-analysis uncover studies that

examined the relationship between the presence of CCTV and the ability of police to elicit

confessions from suspects or raise the likelihood of obtaining a guilty verdict in court.

The City of Los Angeles is one municipality in the state in which law enforcement monitors

video surveillance cameras in “real time.” L.A. has financed many of its cameras through

innovative public-private partnerships and federal grants. To further inform state and local

policymakers as to the effectiveness of CCTV in deterring crime, we employ a quasi-

experimental research design to analyze the effectiveness of video surveillance in two locations

in Los Angeles: Hollywood Boulevard’s “Walk of Fame” and Jordan Downs public housing

development.

Our statistical analysis of crime and arrest data before and after implementation found:

• Neither cameras in Jordan Downs nor Hollywood Boulevard had any statistically

significant effect in reducing the overall monthly crime rates within the target areas;

• The monthly rate of violent crimes fell in both the Jordan Downs and Hollywood target

areas but the results were not statistically significant;

• The monthly rate of property crimes decreased in Hollywood and increased in Jordan

Downs, but the results were not statistically significant;

• The evidence on the displacement of crime is mixed; in both locations, some crimes

increased at a faster rate in buffer areas (between 500 and 1000 feet), while other crimes

decreased at a faster rate in these same areas; however, the results were not statistically

significant;

- 4 -

• CCTV had no statistically significant effect on monthly arrest rates for misdemeanor

“quality of life” infractions in either Jordan Downs or Hollywood Boulevard.

Because local implementation and operations were found to be critical to CCTV effectiveness in

our meta-analysis, we complement our statistical findings through interviews with the Los

Angeles Police Department (LAPD), community groups and business interests as well as through

related documents and media reports. In so doing, we find that the types of crimes being targeted

by the Los Angeles Police Department, and the dynamism of the areas under study, may limit the

generalizability of our results.

We also isolate specific aspects of deployment in Los Angeles that may present challenges or

opportunities to policymakers and law enforcement considering CCTV. These include funding

and public-private partnerships; the presence of simultaneous crime reduction strategies; the

importance of camera placement and coverage area; the technology and monitoring capabilities

of the cameras; the influence of training, turnover and leadership in CCTV operations; and

privacy considerations and community involvement.

Our study led to the development of the following six overarching lessons for policymakers:

1. CCTV is a tool for law enforcement, not a panacea;

2. Public-private partnerships save costs, but raise new policy questions;

3. Effective and sustainable CCTV systems require adequate training, leadership and

resources;

4. Explicit guidance on storage and use of video surveillance might help to allay privacy

concerns;

5. Deterrence and enforcement are strongly intertwined;

6. Additional research is needed into local program operations, as well as the detection,

apprehension and prosecution of criminal suspects.

- 5 -

THE RISE OF VIDEO SURVEILLANCE IN CALIFORNIA

This study examines the effectiveness of video surveillance, also called closed-circuit television

(CCTV) systems, in reducing and deterring crime. While privately monitored CCTV systems

have long been a feature of crime prevention in shopping malls, convenience stores, parking

garages, airports and hospitals, video surveillance systems that are monitored by law

enforcement are relatively new to the U.S. A nationwide survey of law enforcement agencies

conducted by RAND in 2000 found that only 3 percent of local departments and 7 percent of

state police reported extensive use of “fixed-site video surveillance cameras (Schwabe et al.

2001).1

In contrast, public CCTV systems have been employed extensively in cities throughout Europe

for more than a decade. In Great Britain, for example, CCTV is so prevalent that some residents

can expect to be captured by a camera at least 300 times a day (Fussey 2007). With more than

10,000 cameras operating in London alone, at a total cost of nearly $400 million, the security-

camera cordon surrounding the city has earned the nickname the “Ring of Steel” (Davenport

2007).2

However, an amalgamation of technological improvements, cost reductions in video-monitoring

equipment and federal grants to aid local law enforcement in fighting terrorism have contributed

to an appreciable increase in CCTV deployment in the United States. Since the terrorist attacks

of September 11, 2001, for instance, the Department of Homeland Security’s Grant Program

(HSGP) has awarded $23 billion to state and local governments — funds that can be used by

local law enforcement for a variety of security enhancements, including video surveillance

(Center for Democracy and Technology 2008). While the Department of Homeland Security will

not divulge the funding allocated specifically for video surveillance, a Boston Globe

investigation suggests that a “large number of new surveillance systems, costing at least tens and

probably hundreds of millions of dollars, are being simultaneously installed around the country

as part of homeland security grants” (Savage 2007). Our own analysis of news coverage suggests

the following cities received federal grants to aid in their CCTV deployment: Chicago, New

York, St. Paul, Madison, Pittsburgh, New Orleans, Baltimore, San Francisco and Los Angeles.3

Chicago, which now has 560 cameras in “high-crime areas,” received $34 million to help link up

existing camera systems to one centralized location (Moore 2005). New York, for its part,

recently unveiled the Lower Manhattan Security Initiative, a plan that includes adding more than

3,000 surveillance cameras to the 250 cameras currently in place (Viahos 2008). Other cities

such as Philadelphia, Washington D.C. and New Orleans have also invested heavily in CCTV

surveillance (Bulwa and Stannard 2007).4

1 The RAND survey further found that only 1 percent of state and local police nationally use “mobile video

surveillance cameras, [which] might be used in a stakeout or hostage negotiation situation” (Schwabe et al. 2001).

2 Costs reported in pound sterling by London’s Evening Standard as £200 million. Converted into dollars April 11,

2008, at a rate of 1.969.

3 The Globe reports the following grants from DHS: St. Paul, MN, received a $1.2 million grant for 60 cameras;

Madison, WI, received a $388,000 grant for 32 cameras; Pittsburgh received $2.58 million for 83 cameras.

- 6 -

4 In 2007, U.S. Senator Joe Lieberman (I-CT), Chairman of the Homeland Security and Government Affairs

Committee, proposed an amendment requiring the Homeland Security Department to develop a “national strategy”

for surveillance-camera deployment in the United States to fight terrorism (Office of U.S. Senator Joseph Lieberman

2007).

The burgeoning federal involvement in CCTV funding and

deployment has likely contributed to the growth of public video

surveillance by law enforcement in California — a trend that is

likely to continue. In this section, we discuss the prevalence of

CCTV in California and give a general snapshot of crime trends in

the state. We also examine some of the privacy concerns raised by

opponents of public video surveillance and delineate the

predominant theories as to why video surveillance may prove an

effective crime fighting tool.

The Prevalence of Video Surveillance in California

The use of CCTV in California is on the rise. A 2007 American

Civil Liberties Union (ACLU) report found that, of the 131 cities in

northern and southern California surveyed, at least 37 have some

sort of public video surveillance program in place. Of these cities,

18 have “significant” surveillance programs for public streets and

plazas (Schlosberg and Ozer 2007). Cities and jurisdictions with

“significant” surveillance systems are listed in Table 1; cities and

jurisdictions with “actively monitored” systems in which local

police departments can monitor the cameras in real time are listed in

Table 2.5

Table 1.

California Cities with

Significant Surveillance

Systems

Table 2.

California Cities with

Actively Monitored

Systems

Antioch Ripon Brentwood Sacramento

Brentwood Sacramento Clovis Santa Rosa

Clovis San Francisco Fairfield Stockton

Fairfield Stockton Fresno Fullerton

Fresno Long Beach Lindsay Long Beach

Lakeport Palm Springs Pinole Los Angeles

Los Angeles Riverside Pittsburg Oxnard

Manteca Santa Barbara Ripon Riverside

Pittsburg South Gate Santa Monica

Schlosberg and Ozer (2007). “Public Records Survey Summary Findings”;

available at:

www.aclunc.org/docs/Government_Surveillance/report_spreadsheet_for_website.

pdf.

5 Schlosberg and Ozer (2007) do not indicate whether these municipalities view the cameras in real time, nor do they

specify what constituted a “significant” CCTV system.

- 7 -

Active vs. Passive

Monitoring

Two basic modes of

deployment currently

enable CCTV

surveillance in public

spaces: actively

monitored systems and

passively monitored, or

recording-only,

systems.

Actively monitored

systems require

personnel to monitor

television screens in

real time. Depending

on the hardware

installed, actively

monitored systems can

allow for recording as

well, in addition to

remote movement of

cameras and even

audible communication

with subjects in the

surveilled areas. Such

systems enable law

enforcement to respond

to, and potentially halt,

crime as it occurs.

Passively monitored,

recording-only

systems, on the other

hand, provide a record

of criminal activity that

can be used as evidence

later, but these systems

cannot be viewed in

real time.

Particularly notable cases of video surveillance programs in California include the following city

programs:6

San Francisco. The city has installed 68 CCTV recording cameras since 2005. The San

Francisco Chronicle reported that, as of March 2008, the city had spent $900,000 and

had budgeted an additional $200,000 for 25 more cameras to target both violent gang

activity and property crime (Knight 2008).7

Fresno. The city has 47 video surveillance cameras currently in place. The Fresno Bee

reports that the city plans to deploy 130 more cameras by October 2008, followed by 250

more cameras by 2009 (Guy 2008).8 The Fresno Police Chief has indicated that the $3

million project will be actively monitored.

El Cerrito. This California city in Contra Costa County enacted the Surveillance Act of

2007. The law mandates police-accessible video surveillance cameras in certain types of

businesses such as liquor stores, banks, check-cashing businesses, gas stations,

convenience stores and pawn shops (El Cerrito City Ordinance 2007–06).

San Jose. In June 2006, the San Jose City Council appropriated $96,727 to install

surveillance cameras in the city’s downtown areas (San Jose City Council 2006). The San

Jose Redevelopment Agency raised $97,000 for the purchase of the video surveillance

cameras (Herhold 2007).9

Los Angeles. In conjunction with private businesses, federal grants and business-

improvement districts, the Los Angeles Police Department has purchased and deployed

80 actively monitored cameras in various locations since 2002, including the famous

Hollywood “Walk of Fame”; MacArthur Park, a municipal park in downtown; Santee

Alley in downtown; and the Jordan Downs public housing projects in Watts, among

others. Private parties have given the LAPD direct access to data collected by

approximately 30 additional cameras (Email correspondence, Gomez April 14, 2008).

California Crime Rates, Video Surveillance and Spending

As the utilization of video surveillance by law enforcement has increased, overall crime in

California has remained relatively stable, and, in some cases, declined. In 2006, the state had 518

violent crimes and 1,889 property crimes per 100,000 residents (California Department of Justice

2006). Compared with other states, California ranked 25th in the rate of crime per 100,000

residents (United States Department of Justice 2005). While the violent crime rate in California

6 For brief descriptions of additional CCTV systems in California, including Palm Springs, San Diego and Vallejo,

see California Research Bureau (Nieto et al. 2002) “Public and Private Applications of Video Surveillance and

Biometric Technologies,” p. 17-18.

7 Cameras are located at 24th and Harrison streets in the Mission District, at the corner of Eddy and Laguna streets,

outside the Yerba Buena Plaza East housing project, Pacific Avenue and Stockton Street in Chinatown, and Market

Street and Sunnyvale Avenue at the request of residents.

8 The article notes that “cameras are already in place in Roeding Park, near the FAX bus transfer station in

Manchester Center and along Parkway Drive” and will also be used to monitor activity near public schools.

- 8 -

9 Note that because of the timing of the ACLU report, neither El Cerrito nor San Jose was included in its survey.

increased 1.2 percent in 2006, this was the first such increase in 13 years (California Department

of Justice 2006). In contrast, the property crime rate has steadily increased since 1999 (ibid).10

Rates for violent and property-related crime per 100,000 residents in California from 1996 to

2005 are shown in Figure 1.

Figure 1. Crime in California 1996-2005

COMPSTAT Historical Fact Sheet (2008).

Despite these general declines, crime and crime prevention loom large in both the minds of

residents and the state budget. A 2006 poll conducted by the Ralph and Goldy Lewis Center for

Regional Policy Studies at the University of California, Los Angeles (UCLA), indicated that

crime was the second most important concern among southern Californian residents that year

(SCS Fact Sheet 2006). Another poll, conducted in 2005, indicated that 56 percent of

Californians were “extremely concerned” about crime and law enforcement (California Opinion

Index 2005). Clearly, public safety is important to Californians, and the use of video surveillance

has increasingly become a part of the crime prevention toolkit used by police departments.

In 2002, California spent about $18.1 billion to fight crime, a figure that includes the costs of

police, prosecution, courts and corrections (California Department of Justice 2002).11 Additional

grants from the federal government to California cities and jurisdictions for new security

measures such as video surveillance represent an important source of new revenue for law

enforcement in the state.

10 The crime rate represents the number of crimes per 100,000 people in California. It should be noted that official

crime statistics imperfectly measure actual crime. Official statistics are based on reported crimes and, therefore, do

not account for crimes that were not reported to police or were otherwise not detected by law enforcement officials.

- 9 -

11 This figure is not adjusted for inflation and represents the last year in which the cost of crime was estimated by the

California Department of Justice.

Furthermore, the sale of video surveillance equipment has exploded in recent years. While

surveillance sales trends in California were not available, one market research firm estimates that

global surveillance sales will have grown at a compound annual growth rate of 12.4 percent from

2005 to 2008 (RNCOS, 2007). Another market research firm anticipates revenue for video

surveillance equipment worldwide will expand from $13.5 billion in 2006 to $46 billion in 2013

(ABI Research Press Release, 2008). “This is a modern version of the California gold rush

except that people are bringing cameras instead of pickaxes and shovels,” ABI Research Vice

President and Research Director Stan Schatt stated (ibid).

The general fear of crime, the availability of federal grants and an ever-expanding video

surveillance marketplace have certainly played a role in the growth of CCTV in California and

elsewhere. But the expansion has been accompanied by lingering questions about the

effectiveness of CCTV systems and concerns about the protection of individuals’ civil liberties.

We next consider these underlying concerns in the context of general support for video

surveillance.

Privacy, Efficacy and Public Opinion

The ubiquity of public video surveillance cameras nationwide is experiencing widespread public

support, including in California. An ABC News/Washington Post Poll (2007), for example,

reveals that 71 percent of Americans approve of additional video surveillance. This has not

changed since a CBS News opinion poll conducted in 2002 (CBS) found that 60 percent of

respondents polled thought video surveillance would help reduce crime.

Nevertheless, these programs do face some opposition on privacy grounds, a problem that

California policymakers will eventually be called on to weigh against the systems’ purported

value as a law enforcement tool (discussed in “Arguments for CCTV on Efficacy Grounds” on p.

11). Growing concern regarding the value of video surveillance as a crime prevention strategy is

starting to make headlines in California (Knight 2008). The debate generally hinges upon two

main issues: privacy and effectiveness.

Arguments Against CCTV on Privacy Grounds

- 10 -

Many privacy advocates and constitutional scholars have taken issue with CCTV surveillance,

both nationally and in California. The Fourth Amendment, which ensures “… the right of the

people to be secure in their persons…” has served as the primary legal grounds for opposition to

CCTV surveillance. Privacy advocates suggest that the use of video surveillance in public places

violates the Fourth Amendment wherever a person’s reasonable expectation of privacy is

intruded (Schlosberg and Ozer 2007). The U.S. Supreme Court has battled over whether video

surveillance is constitutional and should be used as evidence in criminal proceedings. In

California, local chapters of the ACLU have further decried the possible “chilling effect” (Nieto

1997) video surveillance might have on First Amendment rights, such as the freedom of

assembly and the freedom of speech. Additionally, California is unique in that voters

overwhelmingly approved a “Right to Privacy” amendment to the California Constitution in

1972. Privacy advocates argue that unrestricted or unregulated utilization of video surveillance

may infringe upon this constitutional provision (Schlosberg and Ozer 2007).

To address the privacy issues that emerged after the deployment of CCTV in businesses, the

courts established the rule that “whatever the public may see from a public place cannot be

private”12 (Paton-Simpson 2000). In 1997, the standard was tightened slightly to include only

what could be seen with “the naked eye”13 in response to the intrusive nature of some video

surveillance technologies. Privacy advocates note that the confluence of sophisticated

technologies — easily available personal data; GIS-mapping software; free street-mapping

software with satellite and, in some cases, street views; facial-recognition software; radio-

frequency identification; and high-resolution video surveillance equipment — dramatically

widens the scope of information that can be obtained about an individual within the view of “the

naked eye” (Nieto 2002).14

Of primary importance to privacy advocates is an apparent lack of guidance on the protection of

civil liberties regarding the storage of video imagery and how actively monitored systems should

be used, if at all. The ACLU notes that “[c]ities throughout California have approved and

implemented camera systems without guidelines to guard against abuse and, in most

circumstances, with little or no public debate” (Schlosberg and Ozer 2007). They argue that

without such guidance, stored video footage may be obtained by random citizens through routine

public records requests. Moreover, video operators may inappropriately profile certain groups or

target individuals for personal amusement. In the United Kingdom, one privacy watchdog

suggests that up to “90 percent of CCTV installations fail to comply with the Information

Commissioner’s U.K. CCTV code of practice, and many installations are operated illegally”

(Davenport 2007). A review of more than 600 hours of data from three different surveilled areas

in the United Kingdom found that 9 out of 10 targets by CCTV operators were men and almost

half were teenagers (Norris 2007). The study also found that Britons of African descent were

between 1.5 and 2.5 times more likely to be targeted by CCTV operators than one would expect

when compared to their percentage in the overall population (ibid).15

Arguments For CCTV on Efficacy Grounds

Supporters of video surveillance typically adhere to “rational choice” theories or environmental

criminology, both of which emphasize place-based constraints on criminal behavior. A rational-

choice model assumes that criminals are rational actors that make decisions as to where, when

and how they will commit a crime. For environmental criminologists, the spatial dimensions of

crime are most important, and, therefore, offer “place-based” strategies for prevention

(Brantingham and Brantingham 1981). Taken together, Ronald Clarke (1983) calls these

“situational crime prevention” strategies, which he characterizes as comprising measures that:

“(1) target specific forms of crime, and (2) involve management, design, or manipulation of the

immediate environment … so as to increase its risks as perceived by a wide range of offenders.”

12 N.O.C., Inc. v. Schaefer, 484 A.2d 729, 730 (N.J. Super. 1984).

13 Vega-Rodriguez v. Puerto Rico Telephone Company, 110 F.3d 174 (1st Cir. 1997).

14 For additional information on these technologies, see California Research Bureau (Nieto, et al. 2002) “Public and

Private Applications of Video Surveillance and Biometric Technologies.”

- 11 -

15 It is worth noting that 40% of people of the people that the police targeted in the study were picked out “for no

obvious reason.”

Clarke and Homel (1997) classify CCTV as a type of “formal surveillance” that grew from Oscar

Newman’s (1972) “defensible space” approach, which proposed designing buildings and

communities to increase visibility in vulnerable public spaces. Newman’s was among the first

situational crime techniques and a significant catalyst for formalized surveillance.16 It has been

suggested that CCTV surveillance acts as a “force multiplier” that can prevent crime by

enhancing law enforcement’s ability to detect and apprehend criminals, increasing the public’s

awareness of criminal activity, and elevating the perceived risk of apprehension to criminals.

Table 3 provides a matrix of crime-prevention mechanisms.

Table 3. Causal Mechanisms of CCTV Crime Curtailment

Caught in the act Perpetrators will be detected and possibly removed or deterred.

You’ve been

framed

CCTV deters potential offenders who perceive an elevated risk of

apprehension.

Nosy parker

CCTV may lead more people to feel able to frequent the surveilled

places. This will increase the extent of natural surveillance by

newcomers, which may deter potential offenders.

Effective

deployment

CCTV directs security personnel to ambiguous situations, which may

head off their translation into crime.

Publicity CCTV could symbolize efforts to take crime seriously, and the

perception of those efforts may energize law-abiding citizens and/or

deter crime.

Time for crime CCTV may be perceived as reducing the time available to commit crime,

preventing those crimes that require extended time and effort.

Memory jogging The presence of CCTV may induce people to take elementary security

precautions, such as locking their car, by jogging their memory.

Anticipated

shaming

The presence of CCTV may induce people to take elementary security

precautions for fear that they will be shamed by being shown on CCTV.

Appeal to the

cautious

Cautious people migrate to the areas with CCTV to shop, leave their

cars, and so on. Their caution and security mindedness reduce the risk.

Armitage, Smyth, and Pease (1999). Burnley CCTV Evaluation. In N. Tilley (ed.), Surveillance of Public

Space: CCTV, Street Lighting and Crime Prevention, pp. 226-227

These theories function in contrast to traditional criminological theories that investigate the “root

causes” of crime, that is, the psychological and social forces that lead to criminal behavior

- 12 -

16 Nevertheless, Newman was much more interested in “natural surveillance” and never explicitly advocated for

formal video surveillance. In addition, his own defensible space research was widely criticized because of

methodological errors.

(Clarke 1983). From this perspective, because situational strategies do not address the underlying

causes of crime, CCTV surveillance simply displaces criminal activity to areas that are not being

watched (Brantingham and Brantingham 1981). Moreover, if the deterrent effects of CCTV

surveillance rest on perpetrators being rational actors who minimize the chance of being caught,

it may not be effective in preventing certain types of crimes, such as crimes of passion or those

where post-operation apprehension does not apply, such as suicide bombing.

Privacy versus Efficacy in California

The tensions between liberty and security in relation to video surveillance are playing out, in

particular, in two northern California cities. However, as the use of CCTV to monitor public

spaces grows, cities in other parts of California are likely to face similar deliberations.

The 68 publicly funded cameras in San Francisco, spearheaded by Mayor Gavin Newsom are

passively monitored, a configuration that was decided upon to address privacy concerns. San

Francisco city law prohibits police officers from viewing video surveillance footage in real time,

and mandates that the cameras be turned off during planned protests to protect political speech.

Instead, footage may be viewed after a crime occurs, but is automatically erased every 72 hours.

However, Kevin Ryan, the newly appointed director of the mayor’s own Office of Criminal

Justice, argued that without an actively monitored surveillance system like that in Chicago or

Los Angeles, the power of video surveillance as a law enforcement tool cannot be realized

(Bulwa, 2008). Ryan and others argue that actively monitored systems provide police with an

opportunity to respond to crimes as they occur. San Francisco Police Commissioner Joe Alioto-

Veronese, too, believes that the passively monitored configuration — along with their low

resolution and slow frame rate (two to four frames per second) — limits the system’s usefulness

to police. Some of the police commissioners note that only one arrest has been made in

connection with a murder after more than two years of video surveillance. San Francisco Police

Chief Heather Fong, however, argues that, despite the difficulty in quantifying criminal

deterrence, the level of comfort video surveillance brings to citizens should not be discounted.

In neighboring Oakland, then-mayor and current California State Attorney General Jerry Brown

rejected the use of video surveillance in the city in 1999 (Schlosberg and Ozer 2007). Brown

argued that the police department and the community working together will make the city safer;

video surveillance, while becoming increasingly intrusive, will not. Mr. Brown’s past opposition

to video surveillance and his current position as the State Attorney General at the California

Department of Justice make him a potentially significant source of opposition to further

deployment of CCTV systems in California. In October 2005, however, Oakland Councilman

Ignacio De La Fuente circumvented the City Council, attracted funding through a public-private

partnership and had three cameras installed in his district, giving full utilization of them to the

police. Interestingly, De La Fuente had eloquently stated his opposition to the installation of

video surveillance cameras in the city in 1999, but apparently changed his mind (Ozer 2006) in

an effort to combat drug dealing outside a liquor store in his district (De La Fuente 2005).

- 13 -

In many cases, these arguments play out ideologically because empirical study of the

effectiveness of CCTV in California is lacking. Currently, researchers at UC Berkeley are

evaluating San Francisco’s CCTV program, but according to Peter Bibring, contributing writer

on the ACLU’s “Watchful Eye” study, and Romi Ganschow, policy department coordinator for

ACLU, Northern California, very little research is currently underway in other parts of the state

to determine the true efficacy of video surveillance in crime prevention (Interviews, November

14, 2007).17 The goal of this study is to provide a better empirical foundation for considering

CCTV policy in California through both existing research and original case studies. We first

summarize empirical findings from studies of video surveillance systems in other jurisdictions.

We then use these studies to inform our evaluation of the effectiveness of video surveillance in

select locations within the City of Los Angeles.

META-ANALYSIS OF EXISTING EMPIRICAL WORK

A number of studies from around the world have examined the effectiveness of video

surveillance as a law enforcement tool. This meta-analysis builds upon these studies to detail the

apparent effects of video surveillance on various types of crime both in the United States and

internationally. It also reviews the effects of video surveillance by the environment in which the

surveillance cameras operate. Finally, the analysis provides a framework for understanding local

programmatic characteristics that may impact CCTV, describing qualitative differences between

the prevention of crime, the mitigation of crime as it occurs, and the apprehension and

prosecution of criminal subjects.

Crime Deterrence

In 2002, Welsh and Farrington published a meta-analysis of the research on the effectiveness of

CCTV done to date for the United Kingdom’s Home Office, the government department

responsible for protecting the public from crime and terrorism. They found that half (11) of the

22 evaluations analyzed reported statistically significant reductions in crime, 27 percent (6)

reported no statistically significant results, and 23 percent (5) reported a statistically significant

increase in crime. We built upon their analysis and used their framework to include additional

evaluations that have been produced since the meta-analysis was published. Our results are

similar, with 41 percent of the evaluations reporting a statistically significant decrease in crime,

43 percent reporting no statistically significant or uncertain results, and 16 percent reporting a

statistically significant increase in crime.

The criteria for including a CCTV evaluation, as specified in Welsh and Farrington’s (2002: 3)

and used in our analysis, are:

1. CCTV intervention is the focus of the study;

2. Outcome measures of crime are reported;

3. The study used a minimum research design of before-and-after measures of crime in

experimental and control areas;

4. The total number of crimes in each area prior to the CCTV intervention was at least

20.

- 14 -

17 University of California researchers released preliminary findings from as study on the effectiveness of video

cameras in San Francisco in March of 2008.

We searched online academic databases and the bibliographies of CCTV reports to identify

additional studies. We found another 22 CCTV evaluations in addition to the 22 evaluations

from the initial analysis for a total of 44 assessments.18 Of these evaluations, 79.5 percent (35)

were conducted in the UK, 11 percent (5) in the U.S., and the remaining 4 percent were

conducted in Canada, Japan, the Netherlands or Norway. The majority 57 percent (25) of the

CCTV systems we evaluated monitored city streets, 20.5 percent (9) were in residential areas, 11

percent (5) were in public transit stations, and 11 percent (5) were in public parking lots.

Summaries of the evaluations methodologies, locations, and results included in our analysis can

be found in Appendix A (p. 65).

We then classified the evaluations into one of the categories defined by Welsh and Farrington

(2002: 7):19

1. Desirable effect: Significant decrease in crime;

2. Undesirable effect: Significant increase in crime;

3. Null effect: No statistically significant effect on crime;

4. Uncertain effect: Unclear evidence of an effect on crime.

Of the 44 evaluations included in our analysis, 43 percent reported the cameras had no or an

uncertain effect on reducing crime, 41 percent reported a statistically significant reduction in

crime, and 15.9 percent reported some undesirable effect (i.e. a statistically significant crime

increase). Within the 19 evaluations that found no statistically significant effect on crime or were

uncertain as to CCTV’s effect, 36.8 percent (7) reported a reduction in crime, 52.6 percent (10)

reported an increase in crime, and 10.5 percent (2) reported no change or a very small change in

crime.

Importantly, none of the five evaluations that were conducted on CCTV systems in the United

States showed any significant decrease in crime from the presence of cameras. This raises

interesting questions as to whether cultural or geographic differences may influence CCTV’s

deterrent effect. Another review by Welsh and Farrington (2004) suggests that video surveillance

was “found to be far more effective in reducing crime in Britain than in America.” These

differences, they argue, may be attributable to a number of factors, including the absence of

interventions such as ”improved street lighting or police patrols” in the American CCTV

programs they evaluated, as well as cultural differences, such as stronger public support of

CCTV in the United Kingdom (ibid).

The CCTV systems in parking lots had the greatest success rate, with four of the five systems

evaluated demonstrating a desirable effect. In contrast, none of the nine CCTV systems in

residential areas were classified as having a desirable effect, with the majority (seven)

demonstrating no or an uncertain effect on crime. The results for the systems in city streets and

public transit stations were mixed; about half (12) of the systems in city streets were associated

with a desirable effect, 10 exhibited no or an uncertain effect on crime, and three were associated

18 The number of assessments is the total number of quantitative CCTV evaluations, not the number of studies.

Some reports included evaluations of CCTV systems in multiple locations, so the number of evaluations exceeds the

number of studies.

- 15 -

19 The categories are based on the results of the statistical analyses reported in the evaluations and do not necessarily

indicate that the CCTV system caused the observed change in crime.

with increased crime compared to the control. Similarly, two of the CCTV systems in public

transit stations fell in the desirable effect category, two were classified as having a null effect and

one experienced an undesirable effect. Refer to Table 4 for a summary of the results of our meta-

analysis.

Table 4. Summary of Meta-Analysis Results

Effect Evaluations Location (#

evaluations) CCTV Environment

(# evaluations)

Desirable Effect 18 UK (16)

The Netherlands (1)

Japan (1)

City Streets (12)

Parking lot (4)

Public Transit (2)

Undesirable Effect 7 UK (7)

City Streets (3)

Residential (2)

Parking lot (1)

Public Transit (1)

Uncertain/Null

Effect 19

UK (12)

US (5)

Canada (1)

Norway (1)

City Streets (10)

Residential (7)

Public Transit (2)

Some of the studies also analyzed changes in certain types of crime. Overall, CCTV had the

greatest impact on vehicle and other types of property crimes, with 10 evaluations reporting

decreases in vehicle crime and 12 evaluations reporting decreases in property crime.20 The

CCTV systems had a lesser effect upon violent crimes; four assessments reported a decrease in

violent crime and two evaluations reported increases in violent crime (although the authors

suggest that the increase in violent crime may be due to increased detection). Similarly, in

California, University of California, Berkeley researchers released preliminary findings of their

evaluation of San Francisco’s CCTV system. They found “that nonviolent thefts dropped by 22

percent within 100 feet of the cameras, but the devices had no effect on burglaries or car theft”

(Knight, 2008). The study also found that San Francisco’s cameras had “no effect” on violent

crimes (ibid).

These differences between the apparent ineffectiveness of cameras to deter violent crimes, by

comparison to their apparent success in deterring property crimes, may relate back to the

underlying theory that some crimes, such as property-related offenses like larceny, are often

opportunistic (Fabrikant 1979). In contrast, many violent crimes may be motivated by passions

that make individuals less rational, more impulsive, and therefore less influenced by the risk of

detection or apprehension.

- 16 -

20 The evaluations were carried out in several countries, each with their own different system of categorizing and

defining crime. Here, vehicle crime generally refers to vehicle theft and theft from vehicles. Other types of property

crime include burglary, robbery, theft, larceny and “street crimes.”

The findings from the University of California, Berkeley study regarding possible differences in

deterrent effects based on the distance from the cameras also raise concerns about possible

displacement of crime to other areas — an issue raised by critics of situational crime prevention

strategies.

Crime Detection, Mitigation and Prosecution

Although an inherent part of deterring crime is the detection and apprehension of criminals,

operational differences between CCTV systems affect law enforcement’s ability to mitigate

crime as it is happening and the prosecution of a crime after it has been committed.

As discussed briefly in “Privacy versus Efficacy in California” (p. 13) proponents of actively

monitored video surveillance argue that while not all areas may show a statistically significant

net reduction in crime, such monitoring can be used to reduce the escalation of crime as it is

happening. For instance Shepherd (et al. 2005) argues that CCTV surveillance helps to explain

conflicting reports in England and Wales, between an increase in the number of violent crimes

recorded by police, and a decrease in the number of injuries resulting from violent crimes.

Shepherd’s analysis suggests that surveillance has increased detection of violent crimes while

“facilitating a faster police response to arguments and assaults, which limits their duration and

reduces the likelihood and seriousness of injury” (Laurence 2005).

Law enforcement and other officials further argue that video evidence provides a critical

prosecutorial tool, but no studies in our review examine their effectiveness in the apprehension

and prosecution of criminal suspects. Anecdotal evidence suggests that a crime captured on

video may move an undecided jury to reach a guilty verdict or a judge to impose a more severe

sentence on an assailant. For instance, the California Research Bureau cites the importance of

video evidence in prosecuting the perpetrators of the first World Trade Center Bombing (Nieto

1997). According to Deputy Chief Charlie Beck of the Los Angeles Police Department, recorded

evidence of a crime is also instrumental in gaining a confession from a suspect before court

proceedings begin (Interview, Beck, February 5, 2008). A recent homicide investigation in

Fresno provides further evidence for this assertion: Two individuals captured on video shooting a

young man in southwest Fresno were apprehended by police. The suspects originally pleaded

self-defense but confessed to the crime after being told of the video evidence (Guy 2008).

Some studies have questioned the admissibility of CCTV and other surveillance-related

technologies in a courtroom setting (Murphy 1999). As evidence, CCTV footage is considered

circumstantial, not conclusive (Schlosberg and Ozer 2007). Moreover, to be considered

admissible as evidence, the video footage must add new factual elements to the prosecution’s

case; merely being persuasive does not make it pertinent to a case in court (Greenfield 1991).

- 17 -

Local Characteristics of CCTV Implementation

This discussion of the effectiveness of CCTV in reducing crime, and its operational use as a

crime mitigation and prosecution tool, suggest that local programmatic characteristics heavily

influence the success of surveillance programs. It raises issues around the specific tactical

elements of deployment, for example, whether the deployment of different hardware systems in

different locations causes technical problems for a police department, how the resolution and

frame rate of surveillance cameras affect CCTV’s usefulness, and whether sporadic monitoring

of systems designed to be actively monitored contributes to public safety (Interview, Beck,

February 5, 2008).

These varying results give rise to five operational elements that researchers have identified

which may enhance the deterrent effect of video on criminal activity:

 Comprehensive crime-prevention programs that employ other strategies to complement

CCTV programs (Gill and Spriggs 2005),

 More cameras and greater surveillance coverage of an area (Gill and Spriggs 2005),

 Continuous surveillance areas (e.g., parking lots, commercial areas, and parks) (Gill and

Spriggs 2005),

 Small and well-defined surveillance areas (Ratcliffe 2006),

 Police monitoring and operation of the CCTV system (Ratcliffe 2006).

These individual operational elements may not play an equal role in CCTV program

effectiveness in every case, but they do provide an initial framework for understanding potential

differences. By studying a city in California, additional characteristics that may influence CCTV

effectiveness may arise. The City of Los Angeles’ targeted deployment of CCTV, made possible

largely through public-private partnerships and donations, may also offer new insight to

California policymakers and law enforcement. In the next sections, we provide some context for

our examination of the effectiveness of video surveillance in two locations within Los Angeles.

CCTV, CRIME AND POLICING IN LOS ANGELES

Our meta-analysis identified that, despite the growing popularity of CCTV systems in California

and the mixed results of previous studies, only San Francisco has undertaken a serious evaluation

of its surveillance program — an assessment conducted only recently. Given our finding that

local program characteristics likely affect the overall success of CCTV programs — which are

often very different from one other — in reducing and preventing crime, it is critical that other

cities’ CCTV programs are evaluated and that local characteristics and context are described.

- 18 -

To inform policymakers and law enforcement officials who currently have or are considering

employing CCTV systems as a crime-prevention and crime-reduction tool, we evaluate two

actively monitored public video surveillance programs in Los Angeles. As the largest

municipality in California, Los Angeles may have the greatest demand for future deployment of

CCTV systems and an early assessment of its successes and failures may have broader

implications than those from a smaller city. Moreover, the unique partnerships and actively

monitored systems in Los Angeles provide a framework for comparing programmatic

characteristics to San Francisco’s passively monitored, publicly funded video surveillance

program.

As previously mentioned, working through various public-private partnerships, the Los Angeles

Police Department actively monitors about 80 cameras within its city limits.21 Locations of these

cameras include MacArthur Park, major thoroughfares in Hollywood, the Jordan Downs housing

project in Watts, and a street in downtown Los Angeles’ fashion district named Santee Alley.

We first discuss crime in Los Angeles, the policing situation in the City, and the successful

deployment of CCTV in MacArthur Park. We then provide an overview of our two chosen case

studies or target areas: Jordan Downs Housing Development in Watts and Hollywood Boulevard

between La Brea and Vine in Hollywood. We also describe our control or comparison areas for

our statistical analysis before describing our research questions and methodological approach.

Crime and Policing in Los Angeles

With more than 4.2 million residents, Los Angeles is the second most populous city in the United

States and the largest city in California. Overall, crime rates in Los Angeles have steadily

declined over the years. As illustrated in Figure 2, both property crimes and violent crimes (per

10,000 residents) are the lowest in 10 years. In 2007, the number of crimes reported was fewer

than the previous year in every major (Part I) crime category, including homicide, rape, robbery,

aggravated assault, larceny, auto theft and burglary. Although the crime rate in Los Angeles

remains high when compared to the largest city in the country, New York City, L.A.’s crime rate

(per 10,000 residents) in 2006 was lower than other major California cities, such as San

Francisco and Oakland. (Both San Francisco and Oakland’s crime rates per capita increased in

various crime categories over the last two years.) Still, homicide rates are higher in Los Angeles

than in San Jose, San Diego, Santa Ana, Fresno, Sacramento and many other California cities.

Figure 2. Crime in Los Angeles 1997-2007 Spanning 469 square

miles, Los Angeles is

also one of the largest

cities in the United

States. Despite its

significant size and

population, Los

Angeles has far fewer

police officers per

resident and per square

mile than other large

American cities

(Wagers 2007). As of

- 19 -

21 It should be noted that “actively monitored” video feeds are not watched constantly. The LAPD uses knowledge

of an area and daily crime statistics to determine during which hours of the day to monitor cameras and how the

cameras should be oriented while merely recording (Interview, Graham, March 26, 2008.)

2006, the City of Los Angeles employed 2.4 police officers for every 100,000 residents, as

compared to 4.9 in New York City and 4.8 in Chicago (ibid). Similarly, Los Angeles’ 19.7

officers per square mile are dwarfed by New York’s 128.8 or Chicago’s 61.9 (ibid).

Gang-related Activity

Another important element of Los Angeles crime and policing, particularly in Southeast Los

Angeles, is the prevalence of gang-related activity and the focus of law enforcement on the

suppression of such crimes. A recent report by the Advancement Project (2006) names Los

Angeles “the gang capital of the world,” noting the 40,000 gang members and more than 700

gangs in the city; an astounding 75 percent of youth gang homicides in California take place in

Los Angeles. (In the beginning of 2008, Los Angeles experienced a sharp upswing in gang-

related homicides: As of March 19, 2008, 93 people had been killed in Los Angeles, compared

with 69 during the same period last year — a nearly 35 percent increase [Rubin 2008]).

The Advent of Video Surveillance

William Bratton took over as chief of the LAPD in October 2002. Prior to his arrival, crime had

increased by 54 percent over the previous three years (The Economist 2007). Chief Bratton

brought to Los Angeles many of the same crime-reduction strategies he had previously applied in

other large metropolitan areas cities, among them New York and Boston. These strategies largely

focus on his belief in a “broken windows” theory of policing, wherein officers are encouraged to

address minor criminal offenses in an effort to deter more serious crimes later. One such tactic to

capture and deter these minor offenses was the deployment of video surveillance cameras.

The first police-coordinated deployment of video surveillance in Los Angeles took place at

MacArthur Park, a 40-acre park in the Rampart police district, west of downtown Los Angeles.

While the park was once a tourist hot spot surrounded by an affluent neighborhood and upscale

hotels, after World War II the area transformed into a high-density immigrant community.

Beginning in the 1980s, the park experienced a substantial increase in gang activity, prostitution

and drug-related crimes, and eventually earned the unfortunate distinction of having one of the

highest homicide rates in the city (Leovy 2006). The park was home to five rival gangs — the

Mara Savatrucha, 18th Street, Temple Street, the Playboys and the Drifters — who turned the

park into an open-air black marketplace where drugs, fake ID cards, and stolen goods could be

readily bought and sold. The park became notorious as a place where “anything and everything”

illegal could be obtained (Interview, Beck, February 5, 2008).

- 20 -

A seven-camera actively monitored system was deployed in MacArthur Park in late 2003 with

the help of a federal grant and a partnership with GE Hamilton at a cost of $100,000. (Blankstein

and Ari B. Bloomekatz 2008; Interview, Beck, February 5, 2008). The cameras were one part of

the Alvarado Corridor Initiative, a multifaceted crime prevention and reduction program that

included both tactical deployments and cooperation with other city departments. For example,

the Department of Recreation and Parks added recreational after-school programs in the area,

maintenance crews began regularly cleaning up litter and graffiti in the park, and the Department

of Water and Power doubled the park’s lighting. In addition, the LAPD added standardized

patrols wherein the same police officers

would patrol the same beats. According

to Deputy Chief Charles Beck, the

cameras were used predominately to

target drug dealing and document

forgeries taking place in the park

(Interview, Beck, February 5, 2008).

Senior training officers within the LAPD

at the Rampart Police Station control the

cameras and can call units in to make a

arrest. “At first we made so many arrests

that it overloaded our system — 50 to 6

buyers within a fe

s within the LAPD

at the Rampart Police Station control the

cameras and can call units in to make a

arrest. “At first we made so many arrests

that it overloaded our system — 50 to 6

buyers within a fe

w hours”(ibid).

While no formal evaluation of the video

surveillance cameras was conducted, Sousa and Kelling (2006) report that Part I offenses fell

from 38 per week before the Alavardo Corridor Initiative was put into place to 30 after it

began.22 These early successes of the MacArthur Park video surveillance system were referenced

to support subsequent deployments in Los Angeles, including along Hollywood Boulevard and

in Jordan Downs Housing Project.

While no formal evaluation of the video

surveillance cameras was conducted, Sousa and Kelling (2006) report that Part I offenses fell

from 38 per week before the Alavardo Corridor Initiative was put into place to 30 after it

began.

22 These early successes of the MacArthur Park video surveillance system were referenced

to support subsequent deployments in Los Angeles, including along Hollywood Boulevard and

in Jordan Downs Housing Project.

Hollywood Boulevard cameras were initially deployed in March 2005, focusing on reducing

property crimes, for sting operations, mitigating criminal behavior outside local nightclubs and

retail establishments, and thwarting potential terrorist attacks. By contrast, cameras in Jordan

Downs, a residential public housing project in Watts, were installed in October 2006 to target

gang activity, and drug-related and violent crime.23

Hollywood Boulevard cameras were initially deployed in March 2005, focusing on reducing

property crimes, for sting operations, mitigating criminal behavior outside local nightclubs and

retail establishments, and thwarting potential terrorist attacks. By contrast, cameras in Jordan

Downs, a residential public housing project in Watts, were installed in October 2006 to target

gang activity, and drug-related and violent crime.23

The following section explores our research questions, case studies, methodology and data

collection.

The following section explores our research questions, case studies, methodology and data

collection.

RESEARCH QUESTIONS, CASE STUDIES AND METHODOLOGYRESEARCH QUESTIONS, CASE STUDIES AND METHODOLOGY

This study seeks to fill a critical information gap regarding the effectiveness of closed-circuit

television (CCTV) programs in California in deterring crime. It also seeks to identify

programmatic features that enhanced or inhibited the successful implementation of CCTV. We

employ both quantitative and qualitative methods to determine the efficacy of CCTV systems in

two Los Angeles locations: Jordan Downs Housing Project and Hollywood Boulevard. Both

locations have had CCTV systems for at least one year and provide an important context to

compare CCTV systems in different environments. Although each location has unique social,

political and economic conditions, the problems of disorderly conduct, potential terrorist acts,

22 It is important to note that this apparent reduction could not be directly attributed to the presence of the cameras.

Moreover, the study looked at crime reduction for both the reporting district (RD) that contained MacArthur Park

and the surrounding area. Finally, no comparison or control area was used in the Sousa and Kelling study.

23 Other ongoing surveillance in Los Angeles, including a partnership with the Motion Picture Association of

America to combat bootlegged DVD sales in Santee Alley in downtown Los Angeles, and cameras in MacArthur

Park that target drugs, gangs and prostitution, are not included in this analysis but remain important areas for future

study.

- 21 -

gang violence, property crime and theft, drugs and prostitution are by no means unique to these

locations or to Los Angeles, and the lessons learned here may be applicable to many other areas

throughout California.

Research Questions

Our analysis addresses the following questions:

1) What effect does CCTV surveillance have on crime rates in different environments?

• Does video surveillance have different deterrent effects on different types of crime? For

instance, has it been more effective in deterring property crimes rather than violent

crimes?

• Does CCTV displace crime to adjacent areas outside of the view of the cameras, or does

it have a more robust deterrent effect on the larger community?

• How has CCTV deployment affected arrest rates for quality of life infractions? Has it

increased the detection of these violations that would otherwise go unreported?

2) What factors have contributed to the success of, or present challenges to, CCTV

implementation?

• Which technical aspects of implementation might other localities adopt that may

contribute to success?

• Are certain kinds of environments and arrangements, such as public housing projects or

heavily touristed retail areas, well-suited or ill-suited to CCTV surveillance?

While other questions, such as the effectiveness of CCTV in the identification and prosecution of

suspected criminals, remain important, we viewed an in-depth legal case analysis to be beyond

the scope of this report.

Case Selection: Hollywood Boulevard and Jordan Downs Housing Project

Figure 3. Case Study Locations

To investigate the potentially

deterrent effects of CCTV

surveillance on the different types of

crimes that happen in different

locales, we juxtaposed research on a

retail environment using a matched

pairs analysis and a housing

development using a suitable control.

Arguably one of the most famous

streets in the world, Hollywood

Boulevard is home to the Walk of

Fame, Grauman’s Chinese Theatre,

and the Hollywood and Highland

shopping center. More than 10

million tourists visit various

- 22 -

locations along Hollywood Boulevard each year (Economic Research Associates 2004).

Table. 5 Hollywood Boulevard Characteristics

Site: Retail Area

Dated Installed: February/March 2005

Number of

Cameras: 5 studied; 14 total

Cost: $103,000 in 2004 + $100,000 in

2005

Hrs of Surveillance: 10-12 hours per day; 90% of time at

night

Type of

Surveillance: Active monitoring

Funding Source: Private donation

Implementation: Cooperative with police and the

Business Improvement District

Crimes Targeted: Narcotics, property crimes, public

nuisance, robbery, and terrorism

Figure 4. Hollywood Demographic Change

In addition to the large number of

visitors to the Hollywood area, which

encompasses roughly 20 square miles, it

is also home to more than 230,000

residents (Economic Research

Associates 2004).24 As of 2003, these

residents were slightly older (36.4), had

a lower median household income

($41,026) and were more likely to be

unemployed than Californians as a

whole (Economic Research Associates

2004; American Community Survey

Data Profile Highlights: California

2003). Figure 4 shows demographic

changes in total population and total households between 2003 and 2008. Over the next five

years, more than 4,000 new units of high-end housing will become available in and around the

24 Note that the area size includes both Griffith Park and Los Feliz; pursuant to CA Assembly Bill 588.

- 23 -

A 2003 LA Inc. (formerly Los

Angeles Convention &

Visitors Bureau) study

reported that the Hollywood

Walk of Fame was the fifth

most visited attraction in Los

Angeles; Grauman’s Chinese

Theatre was also in the top 10.

Together with the 38 area

hotels and motels, 53

nightclubs and 110 restaurants

on or adjacent to Hollywood

Boulevard, Hollywood is

trafficked by pedestrians day

and night. Hollywood

Boulevard is also home to the

Kodak Theatre where the

Academy Awards are held and

a Metrolink subway station,

and is frequently the site of

movie premiers, political

rallies and protest marches,

and festivals.

Hollywood Entertainment District, representing a development investment of more than $1

billion dollars (Economic Research Associates 2004).

Jordan Downs, a 700-unit public housing development in the Watts neighborhood in South Los

Angeles, is the second largest public-housing project run by the city’s Housing Authority. The

housing project was originally constructed during World War II to house the influx of Southern

African American laborers who came to fill jobs left vacant by departing soldiers serving

overseas (McGrath 2005). Beginning in the 1960s, the area suffered as the number of

manufacturing jobs steadily declined. By the late 1970s, street gangs formed, and Jordan Downs

became a stronghold for the Grape Street Crips (Winton and Vives 2008). While Jordan Downs

was almost exclusively African American from the early 1940s until the 1980s, the number of

Latino tenants has increased steadily to about 50 percent (Interview, Graham, March 26, 2008),

which has increased social tensions in the area (McGrath 2005). The poverty rate for the ZIP

code containing Jordan Downs was 38.1 percent in 2000 compared to 18.3 percent for the city of

Los Angeles as a whole. In addition, Jordan Downs Housing Complex has been the central point

in a gang war since 2005.

Table. 6 Jordan Downs Characteristics

Site: Public Housing Project

Dated Installed: October 2006; December 2006

Number of Cameras: 6 studied

Cost: $1.2 million

Hrs of Surveillance: Before and after school

Type of Surveillance: Active monitoring

Funding Source: Private grant, Federal grant

Implementation: Cooperative with police and

community groups

Crimes Targeted Gang violence, property

crimes, homicides, robberies

- 24 -

Methodology

Table 7 below summarizes our research methods, data and data sources.

Table 7. Methodology

How does CCTV affect crime rates?

Methodology Data

Statistical analysis of monthly crime rates

before and after CCTV intervention in target,

buffer, and control/matched pair sites

Part I and Part II crime records; arrest

records for quality of life violations

What factors of CCTV implementation contribute to success or present

challenges?

Methodology Data

Semi-structured interviews; document and

media report analysis

Qualitative accounts of the CCTV

implementation process and other

crime prevention interventions in study

areas

Statistical Analysis of Crime Data

We employ a quasi-experimental research design to examine monthly crime data before and after

the introduction of CCTV along the Hollywood Boulevard Walk of Fame between Vine Street

and La Brea Avenue and at the Jordan Downs public housing development in Los Angeles to

measure the effect of CCTV on crime. Descriptive statistics for overall crime and different

categories of crime are analyzed in target, buffer and control areas to determine if there are any

significant changes in crime rates and if these changes can be attributed to the CCTV system.

Our geographic areas of analysis, the time frame of the analysis, measures of crime and

statistical tests are detailed below.

Geographic Areas: Target, Buffer, Control

We delineated target, buffer and control areas that form the geographical basis of our analysis.

The target areas consist of the public space that is directly under surveillance, which we

estimated as the 500 foot radius around each camera. The buffer area is the space 500 to 1000

feet away from the cameras that is not directly under surveillance but is the most susceptible to

the diffusion of benefits or the displacement of crime from the nearby CCTV systems.25 The

- 25 -

25 We base the geographic estimate on previous studies that have looked at possible displacement effects; while the

cameras at each location may be capable of viewing a crime beyond 500 feet in any direction, we felt it reasonable

to assume that not all crimes occurring within sight range could be viewed.

control/comparison areas serve as a point of comparison for any change in crime or arrest rates;

they are geographically and socio-economically similar and do not have CCTV systems.

For the Hollywood site, the target and buffer locations run along the approximately one-mile

stretch of Hollywood Boulevard between La Brea Avenue and Vine Street. Determining a

control area proved to be difficult, however, due to the unique combination of establishments,

tourism, armed private security and a Metrolink station along Hollywood Boulevard. Upon the

advisement of the LAPD, we selected the five surrounding police reporting districts to serve as a

comparison or matched pair.26 This area is commonly referred to as “the Box” by the local police

since it contains a high concentration of restaurants and nightclubs that clearly delineate it from

the surrounding neighborhoods and is the focus of Hollywood Area law enforcement. Figure 5

depicts the Hollywood target, buffer and comparison areas.

Figure 5. Hollywood Camera Locations

For the Jordan Downs location, the target and buffer areas are on the grounds of the Jordan

Downs public housing development. Since the movements of the residents of Jordan Downs are

tightly bounded by the territorial boundaries of the Grape Street Crips gang that operates in the

complex, the buffer areas for the southernmost cameras on 103rd Street have been shortened to

conform to the gang’s turf boundary delineated by 103rd Street.

- 26 -

26 A reporting district is the smallest statistical unit regularly reported by the LAPD.

Figure 6. Jordan Downs Camera Locations A nearby housing project

called Nickerson Gardens, the

largest public housing

development in Los Angeles,

was used as a control. As seen

in Table 8, the inhabitants of

both Jordan Downs and

Nickerson Gardens have

similar demographic

composition and incomes.

Both projects house powerful

gangs to which police

attribute the majority of

crimes — the Grape Street

Crips in Jordan Downs and

the Bounty Hunter Bloods in

Nickerson Gardens. In

addition, Jordan Downs’ two

main policing strategies aside

from video surveillance — a gang injunction and task force — have been active at Nickerson

Gardens for the duration of our study period. Figure 6 depicts the Jordan Downs study areas.

Table 8. Jordan Downs and Nickerson Gardens Demographics

Jordan Downs Nickerson Gardens

White alone 8% 13%

Black or African American alone 61% 47%

Not Hispanic or Latino 60% 47%

Hispanic or Latino 40% 53%

Median Household Income in 1999 $13,306 $12,053

Time Period of Analysis

- 27 -

We collected crime data from January 2003 — the earliest date detailed crime records were

available — through March of 2008. In Hollywood, five cameras were first installed during

February of 2005, followed by the deployment of three additional cameras in June of 2006. To

control for the potentially confounding effect of the addition of cameras in the buffer and

matched pair, we evaluated the period 25 months before and 14 months after the initial

implementation, but prior to the subsequent deployment. Six cameras in Jordan Downs began

installation in December 2006, so the pre-deployment period is 45 months and the post-

deployment period is 16 months. In both cases, a two-month implementation period was

excluded from the analysis to account for the difference between when the cameras first went up

and when they went “live.” (Note: The exact dates from when surveillance poles first went up to

all cameras’ going live could not be precisely determined.) These time frames are within the

range employed by studies with similar aims and will allow us to account for seasonal and

longer-term crime trends.

Measures of Crime

We collected two different measures of crime: arrest records and reported crime records. We

looked at both Part I crimes— homicide, robbery, aggravated assault, theft and motor vehicle

theft — and Part II crimes — including battery, bomb threats and disorderly conduct. While

LAPD reported that Part II crimes are less of a priority for the Department, we felt that in the

case of Hollywood, at least, many of these crimes may be deterred by the presence of the

cameras. Types of crime that cannot reasonably be affected by CCTV, such as fraud, were not

included in our analysis.

In addition, we collected data on arrests for minor offenses, such as narcotics, prostitution and

drunkenness, to measure CCTV’s effect in detecting crime. These “victimless” violations are not

reported to police as crimes; they are only recorded when police actually observe the criminal

behavior and make an arrest. Therefore, these crimes are under-reported, and it is hypothesized

that CCTV’s “caught in the act” mechanism described earlier in “Privacy, Efficacy, and Public

Opinion” (p. 10) will lead to an increase in detection and arrests for these quality-of-life crimes.

Statistical Tests

To determine if the changes in crime rates and arrests are real as opposed to the result of random

fluctuation, we performed a relative effect size (RES) statistical test. This test allows us to

compare before-and-after crime levels in two different areas. We compared changes in monthly

crime rates between the target and control areas as well as between the buffer and control areas.

If the RES is statistically significant, meaning that there is less than a 5 percent probability that

the observed change in crime is due to random variance, then we can conclude that crime

changed at a different rate in the experimental and control areas. It is important to note that a

statistically significant RES does not prove causation. Moreover, the absence of a statistically

significant result does not prove that the cameras were ineffective. For further information on the

statistical test, refer to Appendix C on p. 77.

- 28 -

Qualitative Interviews

To supplement and aid in the interpretation of the statistical analysis, we also collected

qualitative data on the implementation and operation of the CCTV systems and other policing

strategies. We triangulated media reports, official documents and semi-structured interviews with

Los Angeles Police Department personnel in the Hollywood and Southeast stations (which

monitor our test areas), COMPSTAT (the statistical analysis unit of the LAPD), and others in the

South Bureau for background information on the original implementation of video surveillance

in the Hollywood and Jordan Downs vicinities. Additionally, we interviewed community leaders

in the Jordan Downs area and business leaders in Hollywood. Through these interviews, we

obtained descriptions of the CCTV-system implementation process from many different

perspectives, as well as information on ongoing implementation practices, and the integration of

the technology with other policing and community strategies. This information allowed us to

identify factors of the CCTV implementation process that have been successful, what the biggest

challenges have been and how those challenges have been addressed. Given the variability in the

effectiveness of CCTV systems in other states and countries, this qualitative aspect of the

evaluation is imperative to the interpretation of the quantitative results and highlights a number

of programmatic issues that affect how the CCTV systems operate. A complete list of

interviewees is listed in the references.

STATISTICAL FINDINGS

In this section we test the three prevailing hypotheses on closed-circuit television’s (CCTV)

effects on crime. First, if CCTV effectively deterred criminal activity, we would expect that any

decrease in crime within the test area would be greater than any decrease in the matched pair or

control following CCTV deployment. In particular, we would expect property crime rates to be

more affected by CCTV than violent crimes, based on previous studies. Second, if the presence

of cameras displaced crime from the target area to the surrounding areas (the buffers), we would

expect that the crime rate in the buffer areas would increase or decline less than in the matched

pair or control area. Lastly, we would expect the arrest rate for minor violations would increase

in the target areas relative to the matched pair or control after the implementation of CCTV.

This latter hypothesis is due to the fact that, unlike property and violent crimes that are reported

to police by the victims, many minor, quality-of-life violations are “victimless,” and, therefore,

are only recorded if the police themselves detect the criminal behavior and make an arrest. We

would expect that CCTV would allow the police to better detect these minor infractions as they

are happening, leading to an increase in the arrest rate. (A lowering of crimes accompanied by an

increase in arrests occurred in MacArthur Park, where the implementation of an actively

monitored CCTV system led to a sharp increase in narcotics arrests [Sousa 2006]).

Our statistical analysis of crime and arrest data before-and-after implementation in both Jordan

Downs and Hollywood found:

- 29 -

• Neither cameras in Jordan Downs nor Hollywood Boulevard had any significant effect in

reducing violent or property crime rates within the target areas;

• The monthly rate of violent crimes fell in both the Jordan Downs and Hollywood target

areas; however, the Nickerson Gardens control site and the Hollywood Box matched pair

experienced similar reductions and the results were not statistically significant;

• The monthly rate of property crimes decreased in Hollywood, and increased in Jordan

Downs, but the results were not statistically significant in either case;

• The evidence on the displacement of crime is mixed; in both locations, some crimes

increased at a faster rate in the adjacent areas, indicating that CCTV may displace crime,

while other types of crimes decreased relatively more in the buffer areas, though results

were not statistically significant;

• CCTV had no statistically significant effect on arrest rates for misdemeanor quality-of-

life infractions in Jordan Downs or on Hollywood Boulevard.

Our statistical findings in each location are further explored below. We first examine the crime

profile of each site prior to CCTV deployment, disaggregating the monthly rate of different types

of arrests and suppressible crimes; we then present the results of the before-and-after analysis of

monthly crime and arrest rates. Following our presentation of the statistical findings, we discuss

the limitations of the analysis and provide some possible alternative explanations that may have

contributed to the uncertainty of the statistical effects of our study.

Hollywood Boulevard Statistical Findings

Over the 25 months observed prior to the implementation of CCTV along Hollywood Boulevard

(target area), there were 1,951 crimes we deemed as suppressible; that is, crimes that occurred

outside on the street or sidewalk in areas that might have been observed by the cameras. Of these

crimes, 44.3 percent were violent, 54 percent were property-related, and the remaining 1.6

percent were “other” crimes (including both violent and non-violent offenses such as shots fired,

bomb threats and pandering).

- 30 -

Table 9. Hollywood Average Monthly Crime Rates

January 2003 to February 2005 (pre-implementation)

Hollywood Blvd

Target Area Hollywood Box

Matched Pair

Crime Type % of Total Avg per Month % of Total Avg per Month

All Crimes 100.0% 78.0 100.0% 248.0

Violent 44.3% 34.6 39.3% 97.4

Homicide 0.1% 0.04 0.1% 0.4

Aggravated Assault 11.3% 8.8 11.1% 27.6

Robbery 7.6% 5.9 8.6% 21.2

Battery 25.4% 19.8 19.4% 48.2

Property 54.0% 42.2 59.6% 147.8

Vandalism 7.1% 5.6 10.1% 25.0

Car Theft 7.2% 5.6 10.4% 25.8

BTFV 13.0% 10.2 19.0% 47.1

Theft 26.7% 20.8 20.1% 49.9

Other Crimes 1.6% 1.3 1.1% 2.8

As shown in Table 9, theft — including purse snatching, petty theft, and grand theft person —

represented the most prevalent crime in the area (26.7 percent), followed closely by battery (25.4

percent). On an average monthly basis, there were 20.8 thefts reported, 19.8 batteries, 10.2

burglaries or thefts from vehicles (BTFV), 8.8 aggravated assaults, 5.9 robberies, 5.6 car thefts

and vandalism-related crimes, 1.3 other crimes and 0.04 homicides. (There was only one

homicide in the target area over the 25-month pre-implementation period).

In the remainder of the Hollywood “Box,” the matched pair comparison area, 6,200 crimes took

place during our pre-implementation period. A higher percentage of total crimes in the Box were

related to property crimes (59.6 percent). In particular, a higher proportion of cars were stolen

(10.4 percent of all crimes) or broken into (19.0 percent of all crimes) in the Box, though the

target area had a relatively higher percentage of thefts. In contrast, the percentage of violent

crimes was lower in the Box than in the target area with a considerably lower proportion of

batteries and slightly lower percentage of aggravated assaults than the target area; however,

robberies and homicides represented a slightly larger share of crimes in the Box.

In addition to violent and property crime, we also examined the incidence of quality-of-life

infractions, which are recorded as arrests rather than as crimes. In the 25 month period before the

cameras were installed, there were 2,439 arrests in the Hollywood target area for misdemeanor

offenses. As portrayed in Table 10, the most frequently occurring arrest categories were

narcotics (35.9 percent of all arrests), drunkenness (25.3 percent of all arrests) and “other” arrests

(37.3 percent of all arrests), which include Part II crimes that may not rise to level of being

categorized into a specific category. On an average monthly basis, there were 98.9 total arrests,

35.5 arrests for narcotics, 25 arrests for drunkenness, 36.9 other arrests, and less than one arrest

for liquor laws, prostitution and weapon possession.

Table 10. Hollywood Target and Control Area Arrests

January 2003 to February 2005 (pre-implementation)

Hollywood Blvd

Target Hollywood Box

Arrest Group % of

Total Avg per

Month % of

Total Avg per

Month

Total Arrests 100.0% 97.6 100.0% 261.5

Drunkenness 25.1% 24.5 22.1% 57.9

Liquor Laws 0.4% 0.4 1.0% 2.6

Narcotics 36.7% 35.8 35.3% 92.3

Prostitution 0.2% 0.2 4.5% 11.8

Weapon

Possession 0.7% 0.7 1.1% 3.0

Other 36.8% 35.9 35.9% 94.0

In comparison, there were 6,537 misdemeanor arrests in the Box. Table 10 shows that the

proportion of arrests were very similar in the Box and the target area; however, in the Box, a

slightly higher percentage of arrests were for liquor laws (1 percent), prostitution (4.5 percent)

and weapon possession (1.1 percent) than in the target area.

- 31 -

Taken together, the crime and arrest summaries from the pre-implementation period indicate that

the Box provided an adequate matched pair comparison for the Hollywood Boulevard target

area.

Pre-Post Crime Analysis

As shown in Table 11, we found that the average monthly crime rate on Hollywood Boulevard

decreased 10.3 percent following CCTV deployment, which is a slightly smaller reduction in

crime than seen in the matched pair (11.1 percent).

Table 11. Hollywood Monthly Crime Rates

Hollywood Blvd

Target Area Hollywood Box

Matched Pair

Crime Type Pre Post % Change Pre Post % Change

All Crimes 78.04 70.00 -10.3% 248.00 220.50 -11.1%

Violent Crimes 34.60 33.93 -1.9% 97.36 93.86 -3.6%

Homicide 0.04 0.07 78.6% 0.36 0.14 -60.3%

Agg. assault 8.84 7.57 -14.4% 27.60 21.57 -21.8%

Robbery 5.92 5.64 -4.7% 21.24 19.36 -8.9%

Battery 19.80 20.64 4.3% 48.16 52.79 9.6%

Property Crimes 42.16 34.64 -17.8% 147.80 123.50 -16.4%

Vandalism 5.56 6.71 20.8% 25.00 24.64 -1.4%

Car Theft 5.60 4.36 -22.2% 25.80 19.57 -24.1%

BTFV 10.16 6.00 -40.9% 47.12 34.36 -27.1%

Theft 20.84 17.57 -15.7% 49.88 44.93 -9.9%

Other Crimes 1.28 1.43 11.6% 2.84 3.14 10.7%

Violent crime decreased slightly less in the target area (1.9 percent) compared to the matched

pair (3.6 percent) after CCTV implementation. Of the violent crimes, aggravated assault declined

the most — 14.4 percent — along Hollywood Boulevard; however, the aggravated assault rate

fell to a greater degree (21.8 percent) in the Box. Similarly, while robbery decreased by 4.7

percent in the target area, the Box saw an 8.9 percent reduction in robberies during same time

period. Battery, the most prevalent violent crime, increased in both the target and matched pair

during the post-implementation period, but Hollywood Boulevard had a relatively smaller

increase (4.3 percent) compared to the Box (9.6 percent). Although the homicide rates posted

large percentage changes from the pre- to post-implementation periods (78.6 percent and -60.3

percent in the target and matched pair, respectively), the number of homicides is so low that they

have little effect on the overall violent crime rate.

- 32 -

Property crimes decreased substantially more than violent crimes in both the target and matched

pair; however, in contrast to violent crimes, property crimes decreased more in the target area

(17.8 percent) than in the matched pair (16.4 percent). Both areas saw the greatest decline in

vehicle crimes, but Hollywood Boulevard experienced a relatively greater decrease in theft from

vehicles (40.9 percent versus 27.1 percent in the matched pair), while the Box saw a relatively

greater reduction in the number of stolen vehicles (24.1 percent versus 22.2 percent in the target

area). Thefts also decreased considerably in both areas, though the reduction was greater in

Hollywood Boulevard (15.7 percent) than the Box (9.9 percent). The starkest contrast was the

change in vandalism rates following CCTV deployment; along Hollywood Boulevard vandalism

increased 20.8 percent, while the vandalism rate decreased by 1.4 percent in the Box.

While this comparison seems to lend some credence to the hypothesis that CCTV deters property

crime, none of the observed divergences in crime rates between the target and matched pair were

statistically significant at the 95 percent confidence level. Therefore, we cannot distinguish to

what extent the differences were due to random fluctuation in crime or to an actual increase or

decrease in crime attributable to CCTV or other factors. As seen in Figure 7, crime fluctuates

widely from month to month. For more details on the statistical significance of our findings,

refer to Appendix C on p. 77.

Figure 7. Hollywood Monthly Crime

Crime Displacement Effects

We compared the before-and-after crime rates in buffer areas (i.e occurring between 500 and

1,000 feet of each camera) to the matched pair to test for the potential displacement of crime. In

particular, we were interested in examining displacement of battery, burglary or theft from

vehicle, and other thefts—the categories of crime where the target area experienced a relatively

greater decrease (or smaller increase) in crime than the matched pair.

- 33 -

As seen in Figure 8, the evidence is mixed. The buffer area experienced a smaller decline in

BTFV and a larger increase in battery when compared to the matched pair and the target area.

This suggests that these types of crimes may have been displaced rather than deterred by CCTV.

In addition, vandalism and other crimes increased relative to the matched pair, while robbery and

auto theft fell at a slower rate compared to the matched pair. Yet the buffer had a larger reduction

in aggravated assaults and thefts than the Box. However, the statistical analysis revealed that

none of these differences were significant. For additional information on the statistical

significance of our findings, refer to Appendix C on p. 77.

Figure 8. Percentage Change in Hollywood

Monthly Crime Rates Following CCTV

Pre-Post Arrest Analysis

The monthly arrest rate for minor violations rose by 7.6 percent in the target area after

implementation of CCTV, an increase that was less than the rate in the Box (12.3 percent). Both

the target area and the Box experienced the greatest increase in arrests for weapon possession;

however the increase was greater in than Box (61.9 percent) than along Hollywood Boulevard

(42.5 percent). In addition, both areas saw an increase in arrests for drunkenness and other

misdemeanor crimes but the target area had a relatively larger increase in drunkenness (36.8

percent versus 14.28 percent in the Box) and the matched pair had a relatively larger increase in

other arrests (25.6 percent versus 11.1 percent in the target area).

Table 12. Hollywood Monthly Arrest Rates

Hollywood Blvd Target Area Hollywood Box Matched Pair

Arrest Group Pre Post % Change Pre Post % Change

All Arrests 98.9 106.4 7.60% 261.5 293.6 12.30%

Drunkenness 25 34.2 36.80% 57.9 66.1 14.28%

Liquor Laws 0.4 0.4 -11.60% 2.6 1.8 -30.25%

Narcotics 35.5 29.5 -16.90% 92.3 95.5 3.44%

Prostitution 0.2 0.2 -19.00% 11.8 7.4 -37.44%

Weapon Possession 0.8 1.1 42.50% 3.0 4.9 61.90%

Other 36.9 41 11.10% 94.0 118.0 25.59%

- 34 -

Similarly, both the target area and the Box saw reductions in prostitution and liquor infractions,

but the Box experienced relatively larger declines in both categories. Further, narcotics arrests

decreased by 16.9 percent in the target area, compared to a 3.4 increase in the matched pair.

Although the target area did not see a relative increase (or smaller decrease) in arrest rates for

most categories as we would expect if CCTV effectively facilitated the police’s detection of

these crimes, none of the results were statistically significant at the 95 percent confidence level.

Jordan Downs Statistical Findings

In the 45 months prior to the implementation of CCTV in Jordan Downs, 535 suppressible

crimes (crimes that occurred outside on the street or sidewalk) were reported in the Jordan

Downs target area. Of these crimes, 54.0 percent were violent, 44.3 percent were property

related, and the remaining 1.7 percent were classified as “other crimes,” a category including

shots fired, throwing objects at vehicles and brandishing a weapon. As shown in Table 13,

among violent crimes, robbery (25.6 percent of all crimes) was the most prevalent, followed by

aggravated assault (15.3 percent of all crimes) and battery (12.3 percent of all crimes). Among

property crimes, burglary or theft from a vehicle (BTFV) (13.8 percent of all crimes) was the

most prevalent, followed by car theft (12.7 percent of all crimes) and vandalism (12.7 percent of

all crimes).

Table 13. Jordan Downs/Nickerson Gardens Avg. Monthly Crime Rates

January 2003 to September 2006 (pre-implementation)

Jordan Downs

Target Area Nickerson Gardens

Control

Crime Type % of Total Avg per Month % of Total Avg per Month

All Crimes 100.00% 11.89 100.00% 22.93

Violent 54.02% 6.42 50.10% 11.49

Homicide 0.75% 0.09 0.97% 0.22

Aggravated

Assault 15.33% 1.82 14.15% 3.24

Robbery 25.61% 3.04 20.64% 4.73

Battery 12.34% 1.47 14.34% 3.29

Property 44.30% 5.27 49.13% 11.27

Vandalism 12.71% 1.51 15.79% 3.62

Car Theft 12.71% 1.51 13.66% 3.13

BTFV 13.83% 1.64 13.86% 3.18

Theft 5.05% 0.60 5.81% 1.33

Other Crimes 1.68% 0.20 0.78% 0.18

- 35 -

Nickerson Gardens, the control area for Jordan Downs, experienced 1,032 suppressible crimes in

the pre-implementation observation period. Of these crimes, 50.1 percent were violent (a slightly

lower percentage than in Jordan Downs), 49.1 percent were property-related (a slightly higher

percentage than in Jordan Downs), and the remaining 0.8 percent were classified as “other

crimes.” As shown in Table 13 above, Nickerson Gardens had similar proportions of crimes to

Jordan Downs’ within the broader violent-crime category. Once again, robbery (20.6 percent of

all crimes) was the most prevalent type of crime; however, Nickerson Gardens had slightly more

batteries (14.3 percent of all crimes) and slightly less aggravated assaults (14.2 percent of all

crimes) than Jordan Downs. In both sites, homicides represented less than 1 percent of all crimes.

Among property crimes, vandalism (15.8 percent of all crimes) was the most prevalent, followed

by burglary or theft from a vehicle (13.9 percent of all crimes) car theft (13.7 percent of all

crimes) and theft (5.8 percent of all crimes). While the rank order of crime frequency differs

somewhat between Jordan Downs and Nickerson Gardens, the proportions are comparable

(within 5 percentage points in all cases), indicating that Nickerson Gardens is a suitable control.

As in Hollywood, there are several quality-of-life violations such as drunkenness and narcotics

that only appear in arrest records. In the 45-month pre-implementation period, there were 441

arrests of this nature in the target area. As shown on Table 14, the bulk of the arrests (67.8

percent) were for “other” misdemeanor violations, which includes arrests for Part II infractions

that may not rise to the level of any other specific category; one third were for drug offenses; 5.7

percent were for weapon possession; and drunkenness and liquor laws each accounted for less

than one percent of all arrests in the pre-implementation period.

Table 14. Jordan Downs Target and Control Arrests

January 2003 to September 2006 (pre-implementation)

Jordan Downs Nickerson Gardens

Arrest Group % of total Avg per

month % of total Avg per month

Total Arrests 100.00% 9.80 100.00% 40.7

Drunkenness 1.36% 0.13 3.28% 1.3

Liquor Laws 0.45% 0.04 0.00% 0

Narcotics 33.33% 3.27 42.24% 17.2

Weapon Possession 5.67% 0.56 1.69% 0.7

Other 67.82% 5.80 52.79% 21.5

During the same period, there were 1,830 quality-of-life arrests in the Nickerson Gardens control

area. The percentage breakdown of arrests by group resembles that of Jordan Downs, with the

majority (52.8 percent) of the arrests falling into the “other” category and followed by narcotics

(42.2 percent). There was a slightly higher percentage of arrests for drunkenness and a lower

percentage of weapon arrests in Nickerson Gardens, but the differences were small. In sum, the

crime and arrest profiles corroborate the LAPD’s assertion that Nickerson Gardens represents an

appropriate control for Jordan Downs.

Pre-Post Crime Analysis

- 36 -

As shown in Table 15 below, both the target and control sites experienced reductions in crime;

however, crime decreased more in Nickerson Gardens (25.1 percent) than in Jordan Downs (10.6

percent). In Jordan Downs, the reduction in overall crime was due to a decline in the violent

crime rate, which fell slightly more (20.2 percent) compared to the control (18.9 percent).

Robbery, the violent crime most prevalent in both the target and the control in the pre-test period,

fell to a relatively larger degree in Jordan Downs (50.7 percent) than in Nickerson Gardens (20.8

percent). Yet, the aggravated assault and homicide rate declined more in Nickerson Gardens

(49.9 percent and 43.8 percent, respectively) than in Jordan Downs (17.7 percent and 40.6

percent, respectively). Additionally, battery-related crimes grew faster in Jordan Downs (36.4

percent) than in Nickerson Gardens (15.9 percent).

Table 15. Monthly Crime Rates in Jordan Downs and Nickerson Gardens

Jordan Downs

Target Area Nickerson Gardens

Control

Crime Type

Pre Post % Change Pre Post % Change

All Crime 11.89 10.63 -10.63% 22.93 17.19 -25.05%

All Violent Crime 6.42 5.13 -20.20% 11.49 9.31 -18.94%

Homicide 0.09 0.13 40.63% 0.22 0.13 -43.75%

Aggravated

Assault 1.82 1.50 -17.68% 3.24 1.63 -49.91%

Robbery 3.04 1.50 -50.73% 4.73 3.75 -20.77%

Battery 1.47 2.00 36.36% 3.29 3.81 15.92%

All Property

Crime 5.27 5.38 2.06% 11.27 7.44 -33.99%

Vandalism 1.51 1.25 -17.28% 3.62 2.81 -22.35%

Car Theft 1.51 1.44 -4.87% 3.13 2.69 -14.23%

BTFV 1.64 1.69 2.62% 3.18 1.19 -62.63%

Theft 0.60 1.00 66.67% 1.33 0.75 -43.75%

Other Crime 0.20 0.13 -37.50% 0.18 0.44 146.09%

The disparity between the crime rates in the target and control were most apparent for property

crimes: Nickerson Gardens saw a relatively larger decline in every type of property crime

compared to Jordan Downs. As a result, property crime was reduced substantially in Nickerson

Gardens (34.0 percent), while it rose in the target area (2.1 percent). Although Jordan Downs did

exhibit reductions in vandalism (17.3 percent) and car theft (4.9 percent), these gains were

overshadowed by Nickerson Gardens’ 22.4 percent decrease in vandalism and 14.2 percent

decrease in theft. Moreover, theft and burglary-theft from vehicle (BTFV) increased in the target

area (by 66.7 percent and 2.6 percent, respectively) while the control experienced a reduction in

the same crimes (43.8 percent and 62.6 percent, respectively.)

Although the disparities between the crime rates in the target and control areas from the pre- to

post-implementation periods may seem considerable, they, too, failed to pass the relative effect

size test for statistical significance. The high degree of variance in month-to-month crime

(depicted in Figure 9) made it difficult to discern the extent to which any observed changes could

be attributable to CCTV or other interventions versus random variability. The results of the

relative effect size tests are detailed in Appendix C on p. 77.

- 37 -

Figure 9. Jordan Downs Monthly Crime

Crime

Displacement

Effects

As in Hollywood,

the evidence on the

displacement of

crime was

inconclusive. There

was some

indication that

certain types of

crimes may have

been displaced; as

depicted in Figure

10, the buffer areas

experienced a

smaller decrease in

the robbery rate

(6.3 percent) than

either the target

area (50.7 percent)

or the control area

(20.7 percent). In

addition, the rate of

theft grew at a

faster rate (321.9

percent) in the

buffer areas than in

Nickerson Gardens

(43.8 percent) or

the Jordan Downs target area (66.7 percent). However, several types of crime decreased faster or

grew more slowly in the buffer areas than in the control, including homicide, battery and car

theft. None of the results were statistically significant.

Pre-Post Arrest Analysis

Arrest rates for all categories, with the exception of weapon possession, increased relatively

more in the target area than in the control. As shown on Table 16, all arrests rose by 86.2 percent

in Jordan Downs compared to only a 14.2 percent increase in Nickerson Gardens. In particular,

the arrest rate in Jordan Downs for drunkenness, liquor law violations, narcotics and other arrests

outpaced those same arrest categories in Nickerson Gardens. Despite these ostensibly substantial

increases in arrests in the target area compared to Nickerson Gardens, the results were not

significant. (This may be due to the relatively low incidence of arrests and high degree of

- 38 -

Figure 10. Percentage Change in Crime in Jordan Downs Study

monthly variance in these areas.) For further information on the results of the relative effect size

test, refer to Appendix C on p. 77.

Table 16. Jordan Downs Monthly Arrest Rates

Jordan Downs Target Nickerson Gardens Control

Arrest Group Pre Post %

Change Pre Post %

Change

All Arrests 9.8 18.25 86.2% 40.7 46.4 14.2%

Drunkenness 0.1 1.0 650.0% 1.3 2.5 87.5%

Liquor Laws 0.04 0.3 603.1% 0 0 0.00%

Narcotics 3.3 6.1 87.5% 17.2 23.1 34.6%

Weapon Possession 0.6 0.2 -66.3% 0.7 0.4 -36.5%

Other 5.8 10.6 83.2% 21.5 20.4 -5.1%

Discussion of Findings

In sum, we had hypothesized that crime rates would decrease at a faster rate in our target areas

than in the controls. In addition, we expected that CCTV would lead to an increase in detection

of misdemeanor offenses, a fact manifested in higher arrest rates. The data did not support these

hypotheses. However, the lack of statistically significant results should not be taken as

conclusive evidence that the cameras had no effect on crime. There are several factors that may

confound our analysis or offer important alternative explanations, including limitations in the

data, disparities between the target and the comparison areas, the impulsive nature of certain

crimes, CCTV’s potential to enhance the detection of crimes, LAPD capacity constraints and the

scope of our analysis.

Data Limitations

Errors or ambiguities in the original crime and arrest reports may have affected our analysis. As

the head of LAPD’s COMPSTAT Unit explained “the data that we examine is only as good as

what is written up in a crime report” (Interview, Godown, February 28, 2008). For example, due

to missing or inaccurate addresses, approximately 2 percent of the crime and arrest records could

not be mapped in the geographical information system (GIS) software used in our analysis. In

addition, ambiguities in the premise codes used to identify crimes that occurred outside likely led

to Type I (false positive) and Type II (false negative) errors. There are several different codes

that could potentially be used to describe the location of any one crime, based on the subjective

judgment of the reporting officer. For example, a robbery that occurred on the sidewalk in front

of a gym may have been coded as occurring outside on the sidewalk, or coded as at a gym. If the

premise code was listed as “sidewalk” it would be included in our analysis; however, if it had

been listed as “gym” then it would have been excluded. It is difficult to determine the extent to

which these ambiguities and errors in the data affected the analysis.

- 39 -

Disparities between the Target and Control

There may be inherent differences between the target areas (Hollywood Boulevard and Jordan

Downs) and the matched pair (Hollywood Box) or control (Nickerson Gardens) that confounded

our findings.

Hollywood Boulevard is distinct from the surrounding areas that comprise the Box as it serves as

the hub of activity in the area. Its growth in recent years by comparison to other areas in the Box,

including considerably more foot traffic, more tourists and, therefore, more targets and criminals,

may have masked the cameras’ deterrent effects. For example, increased patronage of

Hollywood Boulevard’s bars and nightclubs, which has anecdotally increased 50 to 60 percent

over the past five years, may have artificially led to the absence of a statistically significant

reduction in crime (Interview, Farrell, March 27, 2008).27 According to Clay Farrell, Captain of

the Hollywood Area Community Police Station, “As people’s inhibitions are eased by alcohol,

their chance of getting into trouble, either by committing a crime or becoming the victim of a

crime, increases” (ibid). It is reasonable to posit that many of the crimes that occurred in the

target area began in area nightclubs and spilled onto Hollywood Boulevard.

In addition, the Hollywood Entertainment District Property Owners Alliance (HEDPOA), which

purchased the cameras for the area Business Improvement District, also employs private security

to patrol Hollywood Boulevard for quality-of-life and property-related offenses.28 It is possible

that policy changes on the part of the private security company — and not the LAPD — are

responsible for the increase in detection of and arrest rates for misdemeanor crimes in the target

area.29 Although HEDPOA could not provide historical data on the activities of their private

security, they reported that 1,492 individuals were arrested and brought to police custody in 2007

(Email correspondence, Morrison, April 21, 2008). Therefore, the observed 20.8 percent increase

in vandalism may also be due to increased detection and reporting by private security officers.30

Similarly, disparities in the level of gang activity and hostility in Jordan Downs and Nickerson

Gardens may have confounded the analysis. Both Sgt. Graham of the Southeast Station and

various stories in the media (LAPD Online 2008) indicate that during the study period a gang

war was under way, and that Jordan Downs’ Grape Street Crips were at the center of the

controversy. This war had been going on since 2005, beginning with a loose agreement between

the PJ Crips and the Bounty Hunter Bloods (of Nickerson Gardens) to band together for an attack

on the Grape Street Crips (LA Weekly 2007). Since Jordan Downs’ resident gang experienced an

27 Today, there are approximately 180 alcohol beverage-control (ABC) sites at or near Hollywood Boulevard with a

total nightclub capacity of more than 30,000 persons (Interview, Farrell, March 27, 2008).

28 The Hollywood Entertainment District’s initial purchase is discussed further in “Funding and Public-Private

Partnerships” p. 47.

29 In fact, the Hollywood Entertainment Property Owners Alliance, which manages the Hollywood Entertainment

District, hired a new security patrol, Andrews Security, in November of 2006.

30 However, HEDPOA told researchers that the private patrol primarily reported incidents of gang-related graffiti to

the police, while other types of graffiti are cleaned up by another company contracted by HEDPOA. Therefore, it

may be the case that vandalism is relatively under-reported in the target area, in which case, the incidence of

vandalism may have actually increased by more than the reported 20 percent.

- 40 -

escalation in tensions with two rival gangs, we may expect that crime in Jordan Downs would

increase or decrease to a lesser degree relative to Nickerson Gardens. Therefore, the fact that

there were no statistically significant differences in crime between Jordan Downs and Nickerson

Gardens may actually indicate that the cameras did deter some crime.

Another disparity between Jordan Downs and Nickerson Gardens that may have confounded our

results is the faster pace at which Jordan Downs has changed demographically. Police interviews

indicate that as the demographic balance between African American and Latino inhabitants

within housing developments shift from majority African American to majority Latino, tensions

between rival gangs living side-by-side in the complex may flare. U.S. Census data from 2000

shows Jordan Downs at 60.7 percent African American and 33 percent Latino; however, data

from the Housing Authority of the City of Los Angeles (HACLA), shows that by 2007, 63.4

percent of tenants were Latino and 35.9 percent of tenants were African American. By

comparison, Nickerson Gardens’ 2000 demographic breakdown of 47 percent African American

and 40 percent Latino changed to 55.2 percent Latino and 44 percent African American by 2007,

a substantially smaller shift in the balance between demographic groups and, therefore, relatively

less opportunity for racial tension and possibly crime.

Moreover, in both locations high crime rates served as the impetus for deploying the cameras,

which may indicate a certain degree of self-selection bias. It could be that the designated control

and matched pair areas did not receive CCTV because crime was not as much of a concern as in

the target areas.

Nature of Crimes

Our findings that CCTV surveillance in Hollywood, a hotbed of bars and nightclubs, has had

little effect on crime is consistent with Gill and Spriggs’ (et al. 2005) finding that “[t]hose

offences that are often considered impulsive and influenced by alcohol, for example public order

and violence against the person (VAP), were seen to increase in number more in the target areas

than in their respective control areas.”31 In Hollywood, the introduction of alcohol may not only

serve as a catalyst for certain types of impulsive crimes, such as battery, to occur, but may also

have reduced the chances that the criminals are cognizant that cameras are in the area. “There is

a degree of subjectivity to the cameras’ deterrent effects; individuals that may otherwise notice

the cameras when they are sober may not if they have been drinking” (Interview, Farrell, March

27, 2008).

Similarly, because of the entrenched gang rivalries in Jordan Downs, the cycles of violent

retaliatory attacks might be either unaffected by the presence of video surveillance or, possibly,

positively influenced as participants often seek recognition for their crimes (Heilbrun et al. 1978)

— including prison time, which becomes a mark of distinction for them (Nisperos 2008).

- 41 -

31 Criminologists suggest that the prevalence of violent crime, such as homicide, aggravated assault and battery,

among minority groups subject to prejudice is driven by frustration with deprivation, which results in the adoption

of values that include “quick resort to physical aggression as a sign of daring, courage or defense of status”

(Heilbrun et al.1978). This can lead perpetrators to act impulsively, without adequately weighing the risk of being

caught — by CCTV or any other crime prevention strategy — or the consequences of their actions.

Authorities in Los Angeles have acknowledged that gangs are a social phenomenon that cannot

be deterred by the threat of arrest and incarceration.32

Capacity Constraints

The resource constraints faced by the LAPD may be preventing the CCTV system from fulfilling

its potential as a crime mitigation and deterrence tool. Captain Farrell explained, “There is a

constant tug-of-war between resources. Do we put another cop on the street or get another

camera in place?” (Interview, Farrell March 27, 2008). Farrell further explained that even when

criminal activity is detected on the CCTV system, it is not always possible to mobilize units

quickly enough to respond if there are other, higher priority calls for service. Moreover, in

Hollywood, the cameras are not routinely monitored during the day, and when they are

monitored, there is generally one officer monitoring footage from 14 cameras; therefore, it is

likely that there are criminals who have had the experience of “getting away” with crime in the

monitored areas and may no longer feel deterred by the cameras.

In Jordan Downs, only six of the nine cameras donated by Motorola are in place.33 The resource

commitment required to properly monitor the cameras already in place, coupled with technical

difficulties such as not having sufficient bandwidth to accommodate the amount of data currently

flowing from Southeast’s cameras to the recording station in City Hall East, currently inhibit the

potential utility of the last three cameras even if they were in place (Interview, Graham, March

26, 2008).

Increased Detection, Mitigation, and Resolution of Crime

Finally, we assumed that CCTV would act as a deterrent for property and violent crime and serve

as a detection and mitigation tool in minor quality-of-life infractions that are not generally

reported; however, it may be the case that CCTV also deters some misdemeanors like

prostitution and narcotics-related offenses, and detects some violent crimes, such as aggravated

assault or battery, that would have otherwise gone unreported. These two CCTV functions

produce opposing influence on crime rates, and it is difficult to discern the degree to which any

observed changes reflect deterrence or detection.

Moreover, by just looking at the effect of CCTV on crime rates, we do not capture the effect

CCTV has on solving crime or prosecuting criminals. For example, a gang-related homicide in

the vicinity of Jordan Downs was solved using footage collected by the video surveillance

system. However, because the LAPD does not routinely track when CCTV is used to intervene

in crimes in progress, solve crimes and prosecute criminals, it is difficult to determine CCTV’s

full impact. We expand upon this discussion further in “Conclusions for Policymakers” (p. 53).

32 Jeff Carr, the Los Angeles deputy mayor for gang reduction and youth development, explains, “What we've really

had in the past is a mass incarceration strategy. We’ve locked a lot of people up, and we still have this epidemic

problem” (Mozingo 2008).

- 42 -

33 For more on the Motorola donation, see “Funding and Public-Private Partnerships” (p. 47).

IMPLEMENTATION AND EFFECTIVENESS

We now turn to a discussion of implementation strategy for CCTV use. Through semi-structured

interviews with LAPD personnel and community leaders, as well as document analysis and

participant observations in each area, we isolated several challenges to the adoption of CCTV

that lawmakers considering such systems will need to address. These factors are outlined in

Table 17, below.

Table 17. Implementation Successes and Challenges

Successes Challenges

Funding Private funding and federal grants

offered alternatives to spending

public monies.

Legacy systems and maintenance

contracts may limit future

interoperability and expansion.

Crime

Prevention

Strategy

CCTV can be an effective crime-

prevention and deterrence tool when

added to a cohesive suite of other

policing strategies targeted to a

specific area.

Even if CCTV were successful in

reducing crime in our test areas,

disaggregating its contribution would

prove difficult.

Camera

Placement

Cameras were strategically placed

with community support in high-

crime clusters to maximize their

impact.

Impediments to placing cameras in

strategic locations included technical

issues and working with local property

owners.

Technology

High-quality cameras with fast

frame rates, high resolution and

remote movement capabilities were

installed. Control room software

enabled easier monitoring of all

cameras' output in Jordan Downs.

Additional resources may be needed to

fully implement CCTV systems. The

configuration of control room

monitors may present challenges to

operators in Hollywood.

Operation

Systems in both test areas were

actively monitored in an effort to

mitigate crime as it occurred. A

specially trained subunit operated in

Jordan Downs.

Neither location actively monitored

CCTV 24 hours a day. Specialized

training was not apparent in

Hollywood.

Privacy and

Community

Involvement

Early buy-in of community

members initially allayed concerns

regarding the protection of

individual privacy. LAPD has

guidelines on use and storage of

camera data.

Transience in the community has

eroded some support for the cameras

in Jordan Downs. The specter of

private security officers’ gaining

access to cameras in Hollywood may

raise new privacy concerns.

- 43 -

Successfully addressing these implementation issues not only facilitates the initial deployment of

CCTV but can ultimately impact the system’s effectiveness in deterring crime, based on findings

in previous studies. For each factor, we compare and contrast systems in both locations to inform

policymakers and law enforcement on the qualitative differences that may impact CCTV’s

effectiveness. We summarize these issues and our statistical findings with some takeaways for

policymakers considering CCTV.

Funding and Public Private Partnerships

One feature worthy of exploration is the intrepid use of public-private partnerships to procure

video surveillance systems throughout Los Angeles. The City’s use of private donations for

CCTV hardware, installation and some maintenance costs marks a notable contrast to the 68

camera-system in San Francisco that has cost taxpayers nearly $900,000 (Bulwa 2008). Camera

systems in Jordan Downs and Hollywood, as well as cameras in Downtown’s Santee Alley and

MacArthur Park, have relied on donations from technology companies, business associations

and, in one case, the movie industry. The financial incentive for these partnerships is substantial:

A 1999 RAND survey found that 69 percent of local law enforcement agencies cited cost as a

major barrier to purchasing CCTV systems (Schwabe et al. 2001). Such partnerships allow a city

to create pilot projects and seek community support with a minimal amount of city resources.

Below we discuss the differences between the funding relationships at both sites and risks, such

as legacy and maintenance issues and the need for additional public resources. Despite these

issues, private funding of video surveillance seems to be gaining a firm foothold in police

departments statewide. Video surveillance is viewed by law enforcement as a potential “force-”

or “capability multiplier,” a tool that makes existing law-enforcement personnel more efficient

and productive. The fact that initial costs and maintenance contracts can be satisfied without a

large investment of public funds is seen as a boon to both local government and law enforcement

agencies. And, because camera technology advances quickly and prices are always coming

down, many, including the LAPD, believe funding to take care of these problems will not pose a

serious impediment to CCTV expansion.

Hollywood

Hardware, installation and initial maintenance for Hollywood Boulevard’s first five cameras

were made possible by a $103,000 donation from the Hollywood Entertainment District (HED),

the local business improvement district (BID), which is managed by the Hollywood Property

Owners Alliance (HEDPOA).34 Three additional cameras and a one-year service contract were

purchased by LAPD in June of 2006 for $100,000; money donated again by the BID. According

to HEDPOA, after observing the apparent successes of the cameras on Hollywood Boulevard,

surrounding business improvement districts located in the area donated additional cameras to the

LAPD (Interview, Morrison, April 14, 2008). Presently 14 cameras are operated from the

- 44 -

34 While HED is technically managed by a non-profit 501c6 organization, HPOA’s board of directors largely

comprises private property owners. Thus, we deem the donation of cameras as “private” donations.

Hollywood Community Police Station, most of which were purchased with non-city funds;35 and

the Los Angeles Times reports that the Hollywood area plans to deploy a total of 64 cameras

throughout the area in coming years (Garrison 2004).

Jordan Downs

Jordan Downs’ system was also purchased, in part, through private dollars. Motorola donated

$1.2 million for start-up costs and a $600,000 grant from the U.S. Department of Justice

(Interview, Gomez, March 17, 2008). Motorola hopes to market its Motomesh technology to

other cities, as well as to “scale up” its technology to other areas in Los Angeles. In fact,

Nickerson Gardens, our control area, is purportedly the next project in line to receive video

surveillance, which may be again provided by Motorola (Interview, Graham, March 26, 2008),

given the strength of the company’s relationship with LAPD and the South Bureau.

Types of Partnerships

A distinction should be drawn between donations by technological sponsors like Motorola

(Jordan Downs) and General Electric (MacArthur Park) and broader community and business

partnerships (Hollywood and Santee Alley). These latter partnerships may provide more

flexibility for the LAPD to choose which cameras and vendors to use. For example, HEDPOA

was responsible for finding a video surveillance vendor, in collaboration with the LAPD for the

BID. To this end, HEDPOA hired an FBI consultant to write and issue a request for proposal

(RFP) and to vet potential vendors. While an appropriate vendor was eventually settled upon,

HEDPOA said it also had to receive “buy in” from both HED property owners and the LAPD; a

task that could prove difficult and time-consuming without a strong collaborative partnership

between relevant stakeholders (Interview, MacPherson, April 14, 2008).

Legacy Issues and Maintenance

These unique partnerships and donations provide a good model for cash-strapped local

governments and resource-constrained police departments, but local government may become

locked into older or inadequate technologies that may make interoperability, maintenance and

expansion more costly and difficult. For instance, the City had problems getting some of the

cameras installed in MacArthur Park in 2003, paid for by a donation from General Electric,

replaced — an issue that was raised by the Los Angeles City Council in the early part of 2008.36

The Los Angeles Times reported “glitches with equipment that records and stores video images”

and revealed that “data storage lasts only 12 hours before it is recorded over” (Blankenstein, et

al. 2008). According to Sergeant Daniel Gomez, South Bureau Tactical Technology Unit, no

35 According to the L.A. Times, “Garcetti has pushed for cameras in his district, which runs from Hollywood east to

Echo Park and north to Glassell Park. He recently obtained a $180,000 grant to install 14 cameras in areas around

parks there” (Garrison 2004).

- 45 -

36 Councilman Ed Reyes, whose district contains MacArthur Park, asked the Los Angeles City Council for $150,000

to add six security cameras to replace some of the cameras and add new cameras near the park. However, the Los

Angeles Times story about the maintenance issues forced Reyes to include an “accountability provision” to his

motion that would report back issues regarding equipment. To the contrary, Sgt. Gomez told researchers that most

problem cameras he was aware of had been adequately handled by GE within 72 hours of reporting them.

financial plan for replacing cameras, outside the normal maintenance provided by the

manufacturer, exists. Should a complete system overhaul become necessary at some point, a new

funding agreement will be required, either with GE or another company.

In addition, difficulties in negotiating new maintenance contracts may arise between private and

public stakeholders as CCTV systems expand. For example, the original maintenance contract

with LAPD’s current vendor in Hollywood, MetroVideo, has expired. All of the cameras are

functioning properly, but if technical problems do arise, clients with maintenance contracts are

prioritized over those without such contracts. The LAPD indicated that part of the difficulty in

negotiating a new maintenance contract is deciding who will pay for it: The 14 cameras now

span several business improvement districts and two Los Angeles City Council offices.37

Additional Public Resources

Similarly, a technologically advanced CCTV system like that in Jordan Downs requires

additional public dollars to integrate into existing LAPD technology. For example, Motorola’s

Motomesh system, when fully installed, would allow LAPD officers to access camera footage in

their cars while within the Jordan Downs area, but the amount of bandwidth currently available,

while substantial, is not sufficient to permit this. Funding for upgrading the wireless network has

only recently been budgeted by the City. The currently available bandwidth is provided by

optical fiber that was laid in the area with funding from a grant for disadvantaged communities

before the video-surveillance project was conceived of, an example of seemingly unrelated

programs assisting each other. As adept as law enforcement might become in securing funding

for hardware and support, the efforts and expertise of other city departments are also required for

the smooth functioning of different aspects of a video surveillance program.

Finding funding partnerships, applying for grants, or deciding on vendors for innovative projects

like video surveillance is a somewhat specialized skillset in the LAPD. A core group of officers

in the LAPD, led by Sgt. Gomez, have learned “on the job” how to apply for federal grants and

to negotiate with private donors. These officers stressed that one must be “opportunistic” in

seeking funders for new projects. The generosity of corporate and business alliances to local

government is certainly not something local government can or should ignore, but such an

approach should be thought through strategically. As the City expands its deployment of CCTV

cameras, it should consider closely evaluating area technologies and “scaling up” accordingly.

Simultaneous Crime Reduction Strategies

Simultaneous crime-reduction programs that paralleled CCTV deployment, a factor identified in

our meta-analysis as influential in CCTV effectiveness, were present in both Hollywood and

Jordan Downs. Even though neither of the camera systems in our study locations was

individually directly related to a statistically significant reduction in crime, overall crime rates in

our test and control locations did fall between 10 and 25 percent, attesting to the importance of

these suites of strategies that can include everything from staggering closing times for nightclubs

to mobilizing a community to watch out for its children on their way to school. These ongoing

- 46 -

37 The HED paid $25,000 for the initial maintenance contract with MetroVideo for the first five cameras deployed to

Hollywood Boulevard.

programs stress the importance of cameras as one tactical element, rather than a stand-alone

crime-reduction strategy.

Surveillance Coverage Area and Camera Placement

Another factor that emerged as influential in determining the effectiveness of CCTV in our

review of previous evaluations was the extent of surveillance coverage in an area. While we do

not have the technical expertise to properly evaluate the effective placement of cameras in either

location, how the LAPD deployed these cameras, and the problems they encountered, deserves

some exposition.

Hollywood

In Hollywood, the five cameras located along Hollywood Boulevard’s Walk of Fame surveil a

roughly one-mile area that suffers from the highest overall crime rate in all of Los Angeles.38

The captain of the Hollywood Division Police Station, Clay Farrell, indicated that 25-30 percent

of all crime in the Hollywood Area occurs along Hollywood Boulevard; in 2007, for instance, he

said that 21 percent of Part I crime occurred within the “Box.”

According to HEDPOA, the Business Improvement District (BID) consulted with the LAPD to

determine where crime clusters were occurring in the area before deciding on camera placement.

Assistant Executive Director MacPherson: “You have to kind of find the ideal locations and then

work backwards. You don’t want to cluster too many cameras together but do want a continuous-

type surveillance system” (Interview, MacPherson, April 14, 2008). One issue was that some

“ideal locations” were owned by absentee owners who weren’t available or weren’t interested in

having the cameras on their building. Another problem was finding a building to harbor the

camera which would offer a full panoramic view of the area. One factor that facilitated

placement was that the Department of Transportation had already built poles to monitor traffic

flow in heavily condensed areas. Another implementation challenge was getting power to the

cameras and from what source. Finally, because the imagery is sent over a wireless signal, if the

cameras were not in a clear line of site of the Hollywood Area Police Station, where the cameras

are monitored, they had to set up “repeaters” to bounce the signal in the right direction. “I

remember standing on some area rooftops while we were trying to determine just where to put

the cameras (ibid).”

Jordan Downs

In Jordan Downs the six cameras currently cover approximately one square mile around the

housing complex. Cameras placed on 103rd Street cameras were erected in large measure to

monitor and protect children en route to and from school; other locations were chosen for

maximum coverage of high traffic areas. Another concern was covering routes of ingress and

egress to and from the complex in order to potentially capture images of rival gang members

entering or leaving the area. Sgt. Gomez said that once the LAPD and community had agreed on

- 47 -

38 According to Jane Faerber, a management analyst in the Hollywood Community Police Station’s Crime Analyst

Division (CAD), one of the reporting districts that runs along Hollywood Boulevard (RD645) has the most Part I

crimes in the City (Interview, Jane Faerber, April 8, 2008).

the locations, installation was relatively easy, but that some poles to which the cameras could be

attached were not present in each of the chosen locations. But, according to Gomez “dropping in

poles is cheap” (Interview, March 17, 2008).

Technology and Monitoring

Our review of news reports indicated that many of the purported problems with cameras installed

throughout San Francisco were attributed to those cameras’ low-quality resolution and slow

frame rate, in addition to the inability of police to “actively monitor” locations remotely. By

most measures, cameras in both Jordan Downs and Hollywood are relatively “high-tech” with

perhaps a slightly more advanced system in place in Jordan Downs (see Appendix D on p. 90 for

specifications).

Camera Technology

Both camera systems in our test locations produce high-resolution, high-quality video footage.

All cameras are remotely operable via pan, tilt and zoom features, and are supported by enough

data-transmission bandwidth to allow for the manipulation of, and uninterrupted recording from,

several cameras at once over an encrypted wireless signal. The cameras record at 25 frames per

second in Jordan Downs and 30 frames per second in Hollywood — approximately equivalent to

a video camera and significantly faster than the 2-to-4 frames per second of the San Francisco

system.

Head End Monitoring Systems

The observation systems that serve as the “head ends” where the cameras are monitored by law

enforcement differ substantially at the two sites. The observation system in the Southeast Police

Station, where Jordan Downs cameras are monitored, is software-based, whereas Hollywood’s

observation system is hardware-based. This means that images being sent from Jordan Downs to

Southeast appear as video windows next to each other on a single computer monitor; such a

system allows for more viewing capacity than a hardware-based system. For example, officers

monitoring the cameras in Hollywood must view a stacked set of 14 video monitors, a difficult

task for even the most observant and experienced LAPD officer.

Moreover, the software-based system allows for multiple networked screens, including a large,

flatscreen TV viewable from anywhere in the monitoring office. Controls for manipulating the

cameras in Jordan Downs appear in each small, video window and images can be blown up to

full screen with two clicks of the mouse. Officers in Hollywood’s rather small monitoring office,

by contrast, must manipulate one camera at a time with a single joystick. Some studies have

indicated the difficulty in monitoring multiple video screens simultaneously by a single

individual (Gill et al. 2005). These factors, compounded by the constant foot traffic in the

Hollywood area and the limitations of a single controller watching multiple video screens at

once, may be a vital operational deficiency in the Hollywood system, particularly for the

immediate intervention of crimes as they occur.

- 48 -

Future Implementation

The LAPD’s goal is to fully install the Motorola Motomesh system in Jordan Downs, which

would allow officers to view surveillance video from the cameras in their cars. The system will

provide free Wi-Fi to the residents of the complex while withholding one secure channel for

LAPD activities. On this channel, officers would be able to watch video from any or all the

cameras active in the area while in their cars. Additionally, the cars themselves would act as

hotspot nodes, intensifying the mesh signals as they traveled in the area. Realization of the full

capacity of the system requires that all nine of the planned cameras be installed. (The seventh

camera’s future location is currently being planned.) As mentioned elsewhere, such

implementation depends on the expenditure of more City resources to supplement bandwidth for

transmission. There is no guarantee, however, that a fully implemented system would have a

discernible effect on crime rates and it may impose more intrusively on residents, intensifying a

feeling of being “in prison” in their own community (Email correspondence, Tamika Taylor,

May 1, 2008).

Operations, Training and Turnover

How cameras are monitored, including positioning and staffing, differed substantially at our test

sites. While neither test location showed any significant results in crime reduction, interviews

revealed different successes and challenges in this regard for each location.

Camera Configurations

Camera configuration while not being controlled manually emerged as a relevant operational

issue. Hollywood’s cameras are currently monitored 10-12 hours a day, 90 percent of the time at

night, according to Captain Farrell (Interview, Farrell, March 27, 2008).39 In Jordan Downs,

cameras are actively monitored during times of the day and/or night when criminal activity is

expected to be high and are always viewed during school hours. Area police in Southeast said

that the specific times of monitoring are determined by reviewing COMPSTAT data on a daily

basis, followed by discussion amongst officers as to the situation “on the ground.”

Cameras in Hollywood roam when not controlled manually, moving quickly in multiple

directions rather than focusing on a particular area. Therefore, one might surmise that reviewing

footage after a crime to extrapolate clues or identify license plates may be more difficult here

than where a camera is fixed on a single site. For instance, one recent homicide in Jordan Downs

was essentially “solved” thanks to the cameras fixating in a single direction from which police

anticipated a perpetrator might approach the area.

Leadership, Training and Turnover

Turnover of staff in police departments is anecdotally reported to negatively affect the ability of

CCTV surveillance to function as an efficient crime-fighting tool by bleeding away the

- 49 -

39 As discussed on page 53, in “Privacy Considerations and Community Involvement,” the LAPD is also considering

working with HPOA to install a single monitor that an operator at Andrews Security headquarters can “scroll

through” to observe the cameras, but such a proposition is still in preliminary stages.

institutional memory necessary to fully utilize the technology. Some studies (Carley 1992) reveal

that hierarchical and team structures excel at maintaining different types of institutional memory

and in implementing different types of projects. The functioning of small teams, or units, within

the LAPD, an organization with an overarching hierarchical structure, sets the stage for the

organization to suffer from both the disadvantages of hierarchical and team structures with

regard to personnel turnover.

The team structure is necessary, however, for a pilot project like that in Jordan Downs to

initialize. Teams learn complex systems faster and are more flexible with new material, while

hierarchies tend to be slow and distort information through simplification as it moves up the

chain of command.40 Indeed, “implement[ing] a proposed change often requires ... creating a

specialized subunit that will take on the tasks” (Wilson, p. 231). Teams are very susceptible to

fragmentation when they experience turnover, however, and the mission of the team can then be

lost within the greater hierarchy.

The approach of Hollywood Station personnel toward the surveillance system exhibits some of

these characteristics. The initial team, led by Captain Michael Downing, present at the time

CCTV surveillance was brought into the station no longer functions there. Currently, only a few

officers at the station have much knowledge of why and exactly when the cameras were erected

in the locations they were. (Instead, researchers were referred to the HEDPOA for many of the

details used in this analysis.) It is also unclear if any specialized training is currently given to

camera operators in Hollywood. Although no statistically significant results were realized in

either location, the system in Southeast Station has greater potential for utility as a capability

multiplier due to a tight-knit team with special training that runs the video unit. Should the team

there experience a loss of personnel and expertise, however, that advantage might be lost to the

station, the private partner and the community.

Privacy Considerations and Community Involvement

The tension between potential privacy infringement and efficacy was discussed in some detail in

Privacy, Efficacy and Public Opinion (p. 10), but how privacy protections are managed at the

local level is an important consideration for policymakers implementing CCTV. In addition, the

extent of community involvement and discussion of matters such as privacy, as well as managing

expectations of program effectiveness, are also important for policymakers to consider.

Station Practices

According to LAPD Sgt. Gomez, “each police station has a policy that governs the use and

policy of the camera systems,” since the first cameras were installed at MacArthur Park in 2004

(Email correspondence, Gomez, April 14, 2008). Gomez also stated that each division captain is

responsible for the proper maintenance of records, and random audits of captured video archives

- 50 -

40 The team structure accompanied by specialized training in the gathering and utilization of video evidence has

served the officers in the video unit at Southeast Station well, allowing them to learn from each other as they discuss

and compare observations (Interview, Graham, March 26, 2008). The special training these officers have received

does not preclude their taking part in other types of police work, but other officers are not able to step into their

roles, should they be absent (Interview, Gomez, March 13, 2008).

ensure that the data is being used and stored properly. Recorded video that is kept as evidence of

a crime is booked into the Property Division, where police and personnel from the Los Angeles

City Attorney can obtain access based on need. The LAPD’s Discovery Section handles all other

requests. (Note: The Los Angeles City Attorney’s office was contacted but did not respond to our

request for information).

Access to camera footage is always available upon request to the district attorney’s office.

Officers reviewing footage will mark any that can potentially be used as evidence in solving a

crime as evidentiary and put it aside. All other footage is available to the public through the same

channels they would use to obtain a police report (Interview, Hernandez, May 1, 2008).

Protecting Privacy Rights

Jordan Downs. The issue of protecting privacy rights did not appear of primary concern to the

people we interviewed; a somewhat unexpected finding given their location in a residential

setting where the potential for misuse may be high. (For example, pointing powerful cameras

into the windows of unknowing citizens in their apartments.) However, some residents have a

general feeling of being imprisoned, knowing that their movements could be captured and

viewed by remote observers when inhabitants of other parts of the city can move about

unobserved. According to Tamika Taylor of the Los Angeles Commission of Children, Youth

and Their Families: “It seems more like a liberty issue” than a privacy issue to them (Email

correspondence, May 1, 2008).

Potential privacy concerns may have been reduced initially by the early involvement of

community members. Before installation commenced in Jordan Downs, LAPD officers

participated in a year-long series of meetings to discuss with community members why the

cameras were being considered for their area, what the LAPD hoped to achieve with the

cameras, and what outcomes could reasonably be expected. The video-surveillance program also

became a “catalyst” for open and frank discussion between the police and the community as to

what was needed to improve surroundings in the area, according to the LAPD.

Congruent anecdotal evidence from the LAPD and community leaders reveals that some

community support for the cameras has eroded since the program began. Sgt. Graham noted that

the attitudes expressed toward the cameras in community meetings are more negative now than

when the cameras were first installed. Many of the faces at the meetings have changed, and he

believes that misinformation about what the cameras can and cannot do, and how they were paid

for, is the basis of some discontent (Interview, Graham, March 26, 2008). Community activists,

Pastor Mike Cummings of We Care Outreach Ministries and Tamika Taylor, Community

Program Assistant, at the Los Angeles Commission of Children, Youth and Their Families

agreed, as some residents now resent that scarce resources appear to be consumed by the CCTV

system without meaningful results (Interview, Taylor, April 4, 2008; Interview, Cummings,

April 5, 2008).

- 51 -

Hollywood. In Hollywood, too, the issue of privacy was reportedly less of a concern among

community members than was expected. HEDPOA’s Security Committee conducted outreach to

inform property owners and other community members about when the cameras would be

installed, how the cameras would be used, and in what manner the LAPD would use them.

According to HEDPOA, there has not been much opposition in the Hollywood area from the

community (ibid). “There were some initial concerns [about privacy] … but we don’t really

receive any phone calls complaining about [the cameras]” (MacPherson, Interview, April 14,

2008). From MacPherson’s observation, “the benefits seem to outweigh the privacy costs to most

community members in the effected areas” (ibid).

The apparent lack of opposition to cameras on Hollywood Boulevard along privacy grounds may

be due to the Boulevard’s “public” nature — there are few hours of the day when the Boulevard

is not heavily covered by pedestrian foot traffic — but also because of residents’ experience with

CCTV in the past. Yucca Street, a highly active residential area just north of Hollywood

Boulevard, was one of the first areas in the country to use video monitoring to target criminal

activity during the 1990s. At the time, landlords placed cameras on buildings and submitted

footage of “suspected drug activity” to area police (Garrison 2004). The Yucca Street cameras

were removed in 1999, but new cameras were installed once again, during the second

deployment of cameras to Hollywood in 2006.

Moreover, 20 to 30 signs are spread throughout the Hollywood Area warning passersby that

“Video Monitoring in Progress For Your Safety.” The signs are one aspect of the area privacy

guidance that serves not only as a public notice, but also as a potential deterrent to would-be

criminals.

According to an excerpt of a Hollywood area order, “The Los Angeles Police Department and

Hollywood Area are sensitive to individual citizen's constitutional rights and expectations of

privacy” (Los Angeles Police Department, 2004). In addition to the required signage, the “CCTV

system monitors only public areas where courts and case law have held there is no reasonable

expectation of privacy” (ibid).41

The possibility of granting access to LAPD monitors by HEDPOA’s private security detail —

Andrews Security — does raise some additional concerns regarding privacy. The City Attorney’s

office was present for all the security meetings to help advise HEDPOA and the LAPD about

privacy, and wrote an indemnification agreement severing responsibility from HEDPOA for use

of the cameras. However, both the LAPD and Andrews would like to give the responsibility for

camera monitoring along the Boulevard to Andrews, arguing that such capability may enhance

Andrews ability to better detect and reduce crime in the area. Yet, HEDPOA suggests that

liability issues, among other policy considerations, may derail the effort before it is ever placed

in front of HEDPOA’s Security Committee. It may also raise eyebrows among civil liberties

groups: CCTV of public space by an organization unaccountable to the public seems a prime

target for legal action.

- 52 -

41 The Hollywood area also has some guidance regarding the storage of video equipment. Digital video must be

maintained by the Hollywood Area for a minimum of 30 calendar days, but other guidance was not provided

regarding public requests for video footage.

CONCLUSIONS FOR POLICYMAKERS

Our statistical analysis and qualitative research have guided us to a number of lessons that

policymakers considering video surveillance of public space should consider. We describe each

briefly in our concluding remarks.

CCTV is a tool not a panacea

Policymakers considering video surveillance of public places by law enforcement should not

presume that crime reduction or prevention will occur automatically — or at all. Surveillance

may prove useful for targeting certain types of crimes in certain areas as a tool to aid law

enforcement, but it may be less effective in deterring or reducing other types of crimes in other

areas. For instance, crimes of passion, or crimes in which an individual may be inebriated or

unaware of the cameras’ existence, may not be deterred by video surveillance. Moreover, as the

literature on CCTV suggests, video surveillance may be less effective in open environments or

high traffic areas. In order to adequately cover an area like Hollywood Blvd., for instance, by

creating a “continuous” type surveillance system, LAPD would need to vastly expand the

number of cameras, and, likewise, the number of operators — an unlikely scenario. Also, areas

that are susceptible to “crime waves” like Jordan Downs may not be the most effective

deployments to deter crime, though they may be useful in solving crime. In sum, cameras used in

conjunction with larger crime-reduction strategies should be viewed as one tactical element, not

a strategy in-and-of themselves.

Public-private partnerships save costs, but raise new policy questions

The unique partnerships and donations that the City of Los Angeles has used to procure its

cameras may provide a good model for cash-strapped governments, but policymakers should

also be aware of potential risks. For instance, law enforcement may become locked into older or

inadequate technologies that may make interoperability, maintenance and expansion more costly

and difficult in the long term. In addition, difficulties regarding privacy and liability may emerge

if private security patrols are ever granted access to CCTV monitors.

Sustainability can be difficult

Sustainable CCTV surveillance requires strong leadership, and, ultimately, additional resource

allocation. Turnover of personnel that may have initiated a surveillance system in a particular

area or helped to realize some initial success may leave an implementation gap if new personnel

are not recruited and trained. New priorities for subsequent administrators, strained resources and

the absence of a well-defined training regimen may further limit later success. Policymakers

might consider making use of retired police detectives, as Chicago has, to monitor the cameras.

Need for universal guidance on use, storage and utility of video surveillance

- 53 -

Another issue for policymakers to consider is formalized privacy and use guidelines. While

LAPD suggests that they have local policies in place at each location that limits the accessibility

of stored footage, it seems reasonable to offer some explicit, citywide policies regarding these

matters. This is particularly true as the ubiquity of cameras in Los Angeles and elsewhere grows.

Other cities, such as Fresno, have formalized guidelines for use by law enforcement.42

Deterrence and enforcement are strongly intertwined

Policymakers should also be aware of the linkages between CCTV’s possible deterrent effect and

the ability of police to enforce existing laws. Law enforcement that does not have the resources

to respond to crime and enforce laws, such as certain types of property crimes and minor

offenses where cameras are located, may lead to the belief by criminals that some crimes are

inconsequential. Following the “broken windows” theory of policing, criminals may eventually

believe that they can commit more serious crimes in the absence of consequences, nullifying any

deterrent effect the cameras may have.

Need for further research into detection, apprehension and prosecution

Finally, all of our research and findings point in one direction: Local program characteristics are

the key to the utility of video surveillance. However, police in Los Angeles currently do not

record when they use CCTV to catch criminals, elicit confessions, or provide footage that is

influential to prosecutors in a trial or plea agreement. Research that digs deeper, from initial

deployment and following the adjustments of law enforcement along the way to see how CCTV

data is used to enhance the capability of law enforcement, might provide clearer evidence as to

its cumulative effects.

42 The Code of Practice in the U.K. can be found here:

http://www.ico.gov.uk/upload/documents/library/data_protection/detailed_specialist_guides/cctv_code_of_practice.

pdf

- 54 -

REFERENCES

ABC News/Washington Post Poll (2007, July 29). “Surveillance Cameras Win Broad Support.” Embargoed copy,

available at http://abcnews.com/pollvault.html

ABI Research (2008, March 21). “Video Surveillance: A Market Poised for $46 Billion of Explosive Growth,” Press

Release, available at http://www.abiresearch.com/abiprdisplay.jsp?pressid=1081.

Advancement Project, The (2006, August). “City of Los Angeles Gang Activity Reduction Strategy: Phase 1,”

available at http://www.advanceproj.com/doc/gang_phase1.pdf

American Community Survey (2003). “Demographic Profile Highlights: California 2003,” available at

www.factfinder.uscensus.gov/, last accessed April 04, 14 2008.

Armitage, R.,G. Smyth, and K. Pease (1999). Burnley CCTV Evaluation. In N. Tilley (ed.), Surveillance of Public

Space: CCTV, Street Lighting and Crime Prevention, Vol. 10. Monsey (New York): Criminal Justice Press.

Baig, Edward C., Marcia Stepanek and Neil Gross (1999, April 5). “Privacy: The Internet wants your personal info.

What’s in it for you?” Businessweek.

BBC News (2005, July 23). “Good Response to CCTV Bombs Plea,” available at

http://news.bbc.co.uk/1/hi/uk/4710239.stm.

Beck, Charlie (2008, February 5). Personal Interview. Deputy Chief Commanding Officer. 77th Community Police

Station, LAPD.

Blankstein, Andrew and Ari B. Bloomekatz (2008, January 16). “Police cameras' view fading; After initial success,

a lack of maintenance on MacArthur Park units may be linked to a rise in serious crimes.” Los Angeles

Times.

Brantingham, Paul and Patricia Brantingham (2003). “Anticipating the Displacement of Crime Using the Principles

of Environmental Criminology,” In Theory for Practice in Situational Crime Prevention. Crime Prevention

Studies, vol. 16. Edited by Martha J. Smith and Derek B. Cornish. Monsey, NY: Criminal Justice Press.

Britt, Chester L. and Michael R. Gottfredson (2003). “Editors Introduction.” In Control Theories of Crime and

Delinquency, (Advances in Criminological Theory, Vol. 12. New Brunswick, NJ: Transaction Publishers.

Brown, B. (1995). CCTV in Town Centres: Three Case Studies. Crime Detection and Prevention Series, Paper 68.

London: Home Office.

Bulwa, Demian (2007, October 5). “Newsom wants to add 25 cameras,” San Francisco Chronicle, available at

http://www.sfgate.com/cgi-bin/article.cgi?file=/c/a/2007/10/05/MNR7SK0VL.DTL.

Bulwa, Demian (2008, February 7). “New criminal justice chief wants cops monitoring cameras” San Francisco

Chronicle.

Bulwa, D. and Stannard, M. (2007, August 17). “Security Camera: Is it worth the cost?” San Francisco Chronicle.

Burrows, J.N. (1979). “The impact of closed circuit television on crime in the London Underground.” In P.

Mayhew, R.V.G. Clarke, J.N. Burrows, J.M. Hough, and S.W.C. Winchester, Crime in Public View (pp.

21-29). (Home Office research study no. 49.) London: HMSO.

California Department of Justice (2006). Crime in California 2006, Advanced Release. Bureau of Criminal

Information and Analysis.

California Department of Justice (2002). How Much Does Crime Cost. In “Crime in California 2002.” Bureau of

Criminal Information and Analysis.

California Opinion Index (March, 2005). “Major Issues Facing Californians,” The Field Poll.

Carley, Kathleen (1992, February). “Organizational Learning and Personnel Turnover,” Organization Science, Vol.

3, No. 1, pp. 20-46.

CBS News (2002, April 21). “Poll: Americans OK with video scrutiny,” available at CBSNews.com.

Center for Democracy and Technology (2008, January 18). “CDT Comments on CCTV: Developing Best

Practices.” Docket No. DHS-2007-0076 Submitted to the Department of Homeland Security.

Clarke, Ronald (1983). “Situational Crime Prevention: Its Theoretical Basis and Practical Scope.” Crime and

Justice, Vol. 4.

Clarke, Ronald and Ross Homel (1997). “A Revised Classification of Situational Crime Prevention Techniques.” In

Crime Prevention at a Crossroads, edited by Stephen P. Lab. Cincinnati, OH. Anderson.

COMPSTAT Weekly Meeting (2008, April 16). South Bureau, LA Mart.

COMPSTAT Document (2008). Historical Data from LAPD Statistical Data 1934-2006. 2007 Data from Year End

Comparison Report.

Cummings, Mike. (2008, April 5). Personal Interview. Pastor, We Care Outreach Ministries.

- 55 -

Davenport, Justin (2007, September 07). “Tens of thousands of CCTV cameras, yet 80 percent of crime

unsolved.” Evening Standard.

De La Fuente, Ignacio (2005). “The Right Way to Confront Liquor Stores,” Official Website of Ignacio De La

Fuente, President of the Oakland City Council, available at

http://www.idelafuente.com/News/ReduceCrimeLiquorStoresDec2005.htm, last accessed May 2, 2008.

Ditton, J., E. Short, S. Phillips, C. Norris, and G. Armstrong (1999). The Effect of Closed Circuit Television on

Recorded Crime Rates and Public Concern about Crime in Glasgow (Final report). Edinburgh: The

Scottish Office.

Doan, Lynn. (2006, March 31). “LAPD to Zoom in on Jordan Downs Crime.” Los Angeles Times.

El Cerrito, City of (2006-2007). El Cerrito City Ordinance: Amending El Cerrito City Ordinance, Chapter 6.90.

Economist, The. “New Model Police” U.S. Edition. June 9, 2007.

Economic Research Associates (2004). “Viability of the AMPAS Museum Project in Hollywood.” Prepared for the

Hollywood Chamber of Commerce.

Edmunds, M., M. Hough, and N. Urquia (1996). Tackling Local Drug Markets. Crime Detection and Prevention

Series, Paper 80. London: Home Office.

Fabrikant, Richard (1979, January). “The Distribution of Criminal Offenses in an Urban Environment: A Spatial

Analysis of Criminal Spillovers and of Juvenile Offenders,” American Journal of Economics and

Sociology, Vol. 38, No. 1, pp. 31-47.

Faerber, Jane (2008, April 08). Personal Interview. Management Analyst, Crime Analysis Detail, LAPD.

Farrell, Clay (2008, March 27). Personal Interview. Captain, Hollywood Area Community Police Station, LAPD.

Farrington, D.P., Bennett, T.H., and Welsh, B.C. (2002). “Rigorous Evaluations of the Effects of CCTV on Crime.”

Unpublished manuscript, Cambridge: Institute of Criminology, University of Cambridge.

Flight, S., Y. v. Heerwaarden, and P. v. Soomeren (2003). “Does CCTV Displace Crime? An Evaluation of the

Evidence and a Case Study from Amsterdam.” In M. Gill (ed.), CCTV. Leicester: Perpetuity Press.

Fussey, Pete, (2007). “An Interrupted Transmission? Processes of CCTV Implementation and the Impact of Human

Agency, Surveillance & Society,” Special Issue on ‘Surveillance and Criminal Justice’ Part 1, 4(3): 229-

256, available at www.surveillance-and-society.org.

Fyfe, N.R. and Bannister, J (1996). “City Watching: Closed Circuit Television Surveillance in Public Spaces.” Area,

Vol. 28, No 1.

Garrison, Jessica (2004, October 28). “Cameras to Keep Watch in Hollywood,” The Los Angeles Times.

Geffner, David (2005, February 21). “New policing strategies significantly reducing area's crime problem,” Los

Angeles Business Journal.

Gelman, Lauren (March 9, 2004). “California Court of Appeal Upholds Video Surveillance Obligation for Cyber

Cafes,” The Center for Internet and Society, Stanford Law School.

http://cyberlaw.stanford.edu/packets001996.shtml.

Gill, Martin, and Angela Spriggs (2005). Assessing the Impact of CCTV. Home Office Research Study, Number 292,

London: Home Office Research, Development and Statistics Directorate.

Gill, Martin, Angela Spriggs, Jenna Allen, Martin Hemming, Patricia Jessiman, Deena Kara, Jonathan Kilworth,

Ross Little, Daniel Swain (2005). “Control room operation: Findings from control room observations,”

Home Office Research, Development and Statistics Directorate; Online report.

Gilling, Daniel (1997). Crime Prevention: Theory, Policy and Politics. London, England: UCL Press Limited.

Godown, Jeff (2008, February 28). Personal Interview. Lieutenant, COMPSTAT, LAPD.

Gomez, Dan (2008, March 13). Personal Interview. Sergeant, Parker Center Station, Tactical Training Unit, South

Bureau, LAPD.

Government Accountability Office (2003). “Video Surveillance: Information on Law Enforcement’s Use of Closed-

Circuit Television to Monitor Selected Federal Property in Washington, D.C.” (GAO-03-748), Washington,

D.C.: Government Accountability Office.

Graham, Don (2008, March 26). Personal Interview. Sergeant, Southeast Community Station, LAPD.

Grandmaison, R. and Tremblay, P. (1997). Évaluation des effets de la télé-surveillance sur la criminalité commise

dans 13 stations du Métro de Montréal. Criminologie. Vol. 30, 93-110.

Greenfield, Kent (Summer 1991). “Cameras in Teddy Bears: Electronic Visual Surveillance and the Fourth

Amendment." University of Chicago Law Review. Vol. 58, 1045-1077.

Griffith, M. (n.d.). Town Centre CCTV: An Examination of Crime Reduction in Gillingham, Kent. University of

Reading, accessible at www.crimereduction.gov.uk/ cctv33.pdf.

Guy, James (2008, January 18). “Fresno camera assists in 2 arrests,” Fresno Bee.

- 56 -

Harada, Y., S. Yonezato, M. Suzuki, T. Shimada, S. Era, and T. Saito (2004). Examining Crime Prevention Effects

of CCTV in Japan. Paper presented at the American Society of Criminology Annual Meeting, Nov. 17-20,

Nashville, Tennessee.

Heilbrun, Alfred B., Jr., Lynn C. Heilbrun, and Kim L. Heilbrun (1978). “Impulsive and Premeditated Homicide: An

Analysis and of Subsequent Parole Risk of the Murderer,” The Journal of Criminal Law and Criminology,

Vol. 69, No. 1 (Spring), pp. 108-114.

Helford, D. (2006, April 1). L.A. Gets Federal Grant to Fight Gang Violence, Los Angeles Times.

Herhold, Scott (2007, August 2). “Cameras make downtown safer, but at what cost?” San Jose Mercury News.

Hernandez, Manuel (2008, March 13). Personal Interview. Officer, Parker Center Station, Tactical Training Unit,

LAPD.

Hier, Sean, Kevin Walby and Josh Greenberg (2006). Supplementing the Panoptic paradigm: surveillance, moral

governance and CCTV, In Theorizing Surveillance: The Panopticon and Beyond. Edited by David Lyon.

London, England: William Publishing.

Hollywood Entertainment District Property Owners Association (2005, October 24). “Public Safety Cameras

Contributing to Crime Reduction in BID; Expansion Underway,” Bid Biz. Vol. IX, No. 6.

Hood, J. (2003). “Closed Circuit Television Systems: A Failure in Risk Communication?” Journal of Risk Research

6(3):233-251.

Knight, Heather (2008, March 21). “Crime cameras not capturing many crimes,” San Francisco Chronicle.

Laurence, Jeremy (2005, April, 24). “Injuries from violent crime fall to five year low,” The Independent.

LAPD Online (2008, January 29). “Watts beset by retaliatory shootings,” (first published in the Los Angeles Times)

available online at http://www.lapd.com/article.aspx?a=6022.

LA Weekly (2007, December 7). “L.A. Gangs: Nine Miles and Spreading,” available at

http://www.laweekly.com/news/news/la-gangs-nine-miles-and-spreading/17861/?page=4

Leovy, J. (2006, July 13). “A Complex Portrait of Rampart’s Redemption,” The Los Angeles Times.

Los Angeles Police Department (Oct. 1, 2004). Hollywood Area Order No. 2. Excerpt sent via email April 28, 2008

from Tracy Yasunaga, Management Assistant Hollywood Area Community Police Station.

MacPherson, Sarah (April 14, 2008). Personal Interview. Associate Director, Hollywood Entertainment District

Property Owners Alliance.

Mayhew, Pat (1981). “Crime in Public View: Surveillance and Crime Prevention.: In Environmental Criminology.

Edited by Paul J. Brantingham and Patricia L. Brantingham. Beverly Hills, CA: Sage Publications.

Mazerolle, L., D. Hurley, and M. Chamlin (2002). “Social Behavior in Public Space: An Analysis of Behavioral

Adaptations to CCTV,” Security Journal 15(3): 59-75.

McGrath, R.D. (2005, December 19). “End of the Rainbow: South Central Los Angeles ushers in anew era of racial

tension—this time between blacks and Hispanics,” The American Conservative.

Moore, Martha T. (2005, July 17). Cities opening more video surveillance eyes. USA Today.

Morrison, Kerry (April 14, 2008). Personal Interview. Executive Director. Hollywood Entertainment District

Property Owners Alliance.

Mozingo, J. (2008, May 1). L.A. police aggressively target hard-core gangs. The Los Angeles Times.

Murphy, Thomas (1999). “The Admissibility of CCTV Evidence in Criminal Proceedings.” International Review of

Law, Computers and Technology. Vol. 13, No 3.

Musheno, M.C., Levine, J.P., and Palumbo, D.J. (1978). “Television surveillance and crime prevention: Evaluating

an attempt to create defensible space in public housing.” Social Science Quarterly, 58, 647-56.

Newman, Oscar (1972). Defensible Space: Crime Prevention through Urban Design, New York, NY:

MacMillan.

Nieto, Marcus (1997). “Video Surveillance: Is it an Effective Crime Prevention Tool,” California Research Bureau.

Nieto, Marcus, Kimberly Johnson-Dodds, and Charlene Wear Simmons (2002, March). “Public and Private

Applications of Video Surveillance and Biometric Technologies,” California Research Bureau, p. 17-18.

Nisperos, Neil. (February 1, 2008). “Unit gets tough on taggers: Montclair police share data with other gang units,”

Inland Valley Daily Bulletin, available at http://www.dailybulletin.com/ci_8143131.

Norris, Clive and Gary Armstrong (1997). The unforgiving Eye: CCTV surveillance in public space, Centre for

Criminology and Criminal Justice at Hull University.

Office of U.S. Senator Joseph Lieberman (2007, August 01). “Lieberman seeks more effective use of camera

surveillance,” Press Release.

Ozer, Nicole (2006, January 21). “In Safety’s Name, Technology Breaches Privacy,” The Daily Journal.

Paton-Simpson, Elizabeth (2000). “Privacy and the Reasonable Paranoid: The Protection of Privacy in Public

Places,” University of Toronto Law Journal, pp. 305-346, University of Toronto Press.

- 57 -

BID Biz (2005, October) “Public Safety Cameras Contributing to Crime Reductions in BID; expansion underway,”

BID Biz, a publication of the Hollywood Entertainment District Property Owners Association.

Ratcliffe, J. (2006). Video Surveillance of Public Places, Problem-Oriented Guides for Police Response Guides

Series. No 4. U.S. Department of Justice.

RNCOS (2007, November 11). “Technological Advancements Pushing Global Electronic Surveillance Industry

High,” Press Release, available at http://www.newswiretoday.com/news/25601/.

Rubin, Joel (2008, March 19). Jump in Homicides Not Related to Racial Animosity, LAPD Says, Los Angeles

Times.

Sarno, C., M. Hough, and M. Bulos (1999). Developing a Picture of CCTV in Southwark Town Centres: Final

Report,. London: Criminal Policy Research Unit, South Bank University.

Sarno, C. (1995). Impact of CCTV on crime. In M. Bulos (ed.), Towards a Safer Sutton? Impact of Closed Circuit

Television on Sutton Town Centre (pp. 4-32), London: London Borough of Sutton.

San Jose City Council (2006, June 27), Meeting Minutes.

http://www.sanjoseca.gov/clerk/2006_CnclMins/062706Min.pdf.

Savage, Charlie (2007, August 12). “U.S. Doles Out Millions for Street Cameras,” Boston Globe.

SCS Fact Sheet, Ralph and Goldy Lewis Center for Regional Policy Studies (September, 2006). Transportation,

Crime and The Environment Top SoCal Problems, SCS Fact Sheet, Vol. 2 No 2.

Schlosberg, M. and Ozer, N. (2007). Under the Watchful Eye: The Proliferation of Video Surveillance Systems in

California. San Francisco: American Civil Liberties Union.

Schwabe, W. Davis, L., and Jackson, B. (2003). Challenges and Choices for Crime-Fighting Technology, Santa

Monica: RAND.

Shepherd, Jonathan and Sivarajasingam, S (2005). “Injury research explains conflicting violence trends,” Injury

Prevention, Vol. 11.

Short, E., and J. Ditton (1996). Does Closed Circuit Television Prevent Crime? An Evaluation of the Use of CCTV

Surveillance Cameras in Airdrie Town Centre. Edinburgh: The Scottish Office Central Research Unit.

Sivarajasingam, V., J. P. Shepherd, and K. Matthews (2003). “Effect of Urban Closed Circuit Television on Assault

Injury and Violence Detection.” Injury Prevention, 9(4):312-316.

Skinns, D. (1998). Doncaster CCTV Surveillance System: Second Annual Report of the Independent Evaluation.

Doncaster: Faculty of Business and Professional Studies, Doncaster College.

Sousa, William H. and George L. Kelling (2006). “Reaching the Tipping Point in Los Angeles: MacArthur Park.”

Unpublished Draft, sent by author with permission, February 14, 2008.

Squires, P. (1998). An Evaluation of the Ilford Town Centre CCTV System. Brighton (United Kingdom): Health and

Social Policy Research Centre.

Squires, P. (2003). An Independent Evaluation of the Installation of CCTV Cameras for Crime Prevention in the

Whitehawk Estate, Brighton. Brighton (United Kingdom): Health and Social Policy Research Centre.

Stohs, Brent (2004, October 15). “Privacy, Free Speech and the Garden Grove Cyber Cafe Experiment,” Duke Law

& Technology Review, No. 0012.

Taylor, Tamika (April 4, 2008). Personal Interview. Community Program Assistant, Los Angeles Commission of

Children, Youth and Their Families.

Tilley, N. (1993). Understanding Car Parks, Crime and CCTV: Evaluation Lessons from Safer Cities, Home Office

Crime Prevention Unit series, paper 42, London: Home Office.

United State Department of Justice (2005). Customized tables, based on the FBI’s UCR crime report. State and

National Crime Trend Estimates, All States, All Variables. US Bureau of Justice Statistics.

United States Census Bureau (2007). Los Angeles (City) Quick Facts, available at

http://quickfacts.census.gov/qfd/states/06/0644000.html.

Viahos, James (2008, January). “New Surveillance Cameras: High Tech Law Enforcement Spying Programs,”

Popular Mechanics.

Wagers, Michael (2007). Broken Windows Policing: The LAPD Experience. Dissertation for Graduate Program in

Criminal Justice at Rutgers University.

Wagner, Douglas. (2008, March 11). Personal Interview. Management Analyst II, COMPSTAT, LAPD.

Washington, Jude (2008, May 2). Personal Interview. Officer, Southeast Community Police Station, LAPD.

Webb, B. and Laycock, G. (1992). Reducing Crime on the London Underground: An Evaluation of Three Pilot

Projects, Home Office Crime Prevention Unit series, paper 30. London: Home Office.

Welsh, B. C., and D. P. Farrington (2002). Crime Prevention Effects of Closed Circuit Television: A Systematic

Review. Home Office Research Study, Number 252, London: Home Office Research, Development and

Statistics Directorate.

- 58 -

Welsh, Brandon C., and D.P. Farrington (2004). “Surveillance for Crime Prevention in Public Space: Results and

Policy Choices in Britain and America.” Criminology & Public Policy.

- 59 -

Williamson, D. and McLafferty, S. (2000). “The effects of CCTV on crime in public housing: An application of GIS

and spatial statistics,” Paper presented at the American Society of Criminology meeting, November 15-19,

2000, San Francisco, California.

Winge, S., and J. Knutsson (2003). “An Evaluation of the CCTV Scheme at Oslo Central Railway Station,” Crime

Prevention and Community Safety: An International Journal 5(3):49-59.

Winton, R and Vives, R. (2008, January 30). “Shootings follow killing of gang leader,” The Los Angeles Times,

Retrieved February 15, 2008, from

http://www.latimes.com/news/local/lamewatts30jan30,1,4287409.story?track=rss&ctrack=4&cset=true.

Wood, Daniel (2007, September 25). “Los Angeles Turns Cameras on Gang Graffiti,” Christian Science Monitor.

Personal Interviews, Email Correspondence, Meetings

Los Angeles Police Department (LAPD)

Beck, Charlie. Deputy Chief Commanding Officer. 77th Community Police Station, LAPD. Personal Interview. Feb.

5, 2008.

Faerber, Jane. Management Analyst. Crime Analyst Detail. Hollywood Police Station. Apr. 8, 2008.

Farrell, Clay. Captain. Hollywood Police Station, LAPD. Mar. 27, 2008.

Godown, Jeff. Lieutenant. COMPSTAT, LAPD. Personal Interview. Feb. 28, 2008.

Gomez, Dan. Sergeant. Parker Center Station, Tactical Training Unit, LAPD. Personal Interview. Mar. 13, 2008.

Graham, Don. Sergeant. Southeast Community Station, LAPD. Personal Interview. Mar. 26, 2008.

Hernandez, Manuel. Officer. Parker Center Station, Tactical Training Unit, LAPD. Personal Interview. Mar. 13,

2008.

Ligouri, Jason. Officer. Parker Center Station, Tactical Training Unit, LAPD. May 2, 2008.

Washington, Jude. Officer. Southeast Community Police Station, LAPD. May 2, 2008.

Wagner, Doug. Management Analyst II. COMPSTAT, LAPD. Mar. 11, 2008.

Community Leaders Interviewed

Cummings, Mike. Pastor. We Care Outreach Ministries. Apr. 5, 2008.

MacPherson, Sarah. Associate Director. Hollywood Entertainment District Property Owners. Apr. 14, 2008.

Morrison, Kerry. Executive Director. Hollywood Entertainment District Property Owners. Apr. 14, 2008.

Taylor, Tamika. Community Program Assistant. Los Angeles Commission of Children, Youth and Their Families.

Personal Interview. Apr. 4, 2008.

De La Torre, Ruben. Lieutenant. Media Relations, LAPD. Email Correspondence. Jan. 23, 2008 and Jan. 25, 2008.

Gomez, Dan. Sergeant. Parker Center Station, Tactical Training Unit, LAPD. Email Correspondence. Mar. 11, 2008

and Apr. 14, 2008.

Gray Jr., Louis H. Commanding Officer. Parker Center Station, LAPD. Email Correspondence. Jan. 18, 2008; Jan.

25, 2008, Feb. 25, 2008, Apr. 18, 2008 and May 1, 2008.

Taylor, Tamika. Community Program Assistant. Los Angeles Commission of Children, Youth and Their Families.

Email Correspondence. May 1, 2008.

Meeting Attended

COMPSTAT Weekly Meeting, South Bureau, April 16, 2008.

APPENDICES

- 60 -

APPENDIX A. Evidence of CCTV in Crime Prevention

- 61 -

Author Location

Description of

CCTV System Results Methodology

Armitage,

Smyth, and

Pease, 1999

Burnley,

United

Kingdom

Cameras in the

town center

Desirable effect: 28% reduction in crime vs. a 1%

reduction control area 1, and a 9% increase in

control 2. Violent crimes decreased by 35% vs. a

20% decrease in control 1 and 0% change in

control 2. Vehicle crimes decreased by 48% vs. 8%

declines in both control areas. Burglary declined by

41% vs. increases of 9% and 34% in the two

respective control areas. There was some

displacement of burglary as well as diffusion of

benefits for total crime, violent crime, and vehicle

crime.

Good design: comparison

of crime rates 12 months

before and after CCTV

intervention in

experimental and control

areas.

Brown, 1995

Newcastle

Upon Tyne,

United

Kingdom

16 cameras in

the city center

Undesirable effect: crime did not fall as much in

the experimental area as in the control area (21.6%

decline vs. 29.7% decline in the control area). But

there were reductions in certain types of crime:

57% reduction in burglary, compared to 38.7%

decline in control; 50% reduction in theft from

vehicle compared to 38.9% decline in the control,

47% reduction in vehicle theft compared to 40.5%

decline in control, and 34% reduction in criminal

damage.

Good design: comparison

of crime 26 months

before and 15 months

after in experimental and

control areas.

- 62 -

Brown, 1995 Birmingham,

United

Kingdom

9 cameras in the

city center

Desirable effect: 4.3% decline in crime compared

to a 131.6% increase in the control. Reduction in

robbery, burglary, and theft from person.

Good design: comparison

of crime rates 12 months

before and after CCTV

intervention in

experimental and control

areas.

Burrows, 1979 London,

United

Kingdom

Cameras in 4

subway stations

Desirable effect: robbery declined 22.9% compared

to 23.1% and 116.3% increases in the two

respective control areas. Theft decreased 72.8%

compared to declines of 26.5% and 39.4% in the

control areas.

Good design: comparison

of crime rates 12 months

before and after CCTV

intervention in

experimental and two

control areas.

Farrington,

Bennet, and

Welsh, 2002

Cambridge,

United

Kingdom

30 cameras in

the city center

Undesirable effect: crime did not decline as much

as in the control areas: 13.8% decline vs. 26.9%

decline in control area. Violent crime fell 6% vs. a

33.8% decline in the control. Vehicle crimes

decreased by 53.1% vs. a 54% decline in the

control area. The percentage of people that

reporting being victimized increased by 2.1

percentage points vs. an increase of 2.2 percentage

points in the control.

Good design: comparison

of crime rates 11 months

before and after CCTV

intervention in

experimental and control

areas.

Flight,

Heerwaarden,

and Soomeren,

2003

Amsterdam,

The

Netherlands

29 cameras in

city center

Desirable effect: significant reduction in crime,

compared to no change and slight increases in the

control areas.

Good design: comparison

of crime rates 12 months

before and after CCTV

intervention in

experimental, adjacent,

and control areas.

- 63 -

Gill and

Spriggs, 2005

"City

Outskirts,"

United

Kingdom

47 cameras

around a park,

hospital, and

industrial area

Desirable effect: crime was reduced by 28% in the

target area, compared to a 1% decrease in the

control area. Results are statistically significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

Hawkeye,

London,

United

Kingdom

556 cameras in

parking lots

along the city

outskirts

Desirable effect: crime declined by 73% compared

to a 10% decline in the control area. Results are

statistically significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"City

Hospital,"

United

Kingdom

9 of a total of

67 cameras

around a

hospital are

evaluated

Desirable effect: crime decreased by 33% in the

target area vs. a 6% reduction in the control;

however due to the large difference in size these

results are not statistically significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"South City,"

United

Kingdom

51 cameras in a

mixed-income

city center in

southern

England

Null effect: crime decreased by 10%, but the

control area experienced a greater reduction in

crime (-12%); these results are not statistically

significant

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"Shire

Town,"

United

Kingdom

12 cameras in a

Midlands town

center

Null effect: 4% reduction in crime compared to a

3% increase in crime in comparison site; however,

results are not statistically significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

- 64 -

Gill and

Spriggs, 2005

"Market

Town,"

United

Kingdom

9 cameras in an

affluent market

town

Null Effect: crime increased by 18% in the target

area, vs. a 3% increase in the division as a whole;

however, due to the size differences of the control

and target areas, this result is not statistically

significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"Borough

Town,"

United

Kingdom

40 cameras

around a small

town center

Null effect: crime increased slightly by .3% in the

target area, while crime increased by 13% in the

control site. The results were not statistically

significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"Northern

Estate,"

United

Kingdom

11 cameras in a

residential

neighborhood

with "deprived

housing" in

North England

Null effect: 10% reduction in crime in the target

area compared to a 21% increase in crime in the

control area; results were not statistically

significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"Eastcap

Estate,"

United

Kingdom

12 cameras in a

residential area

with "deprived

housing" in

southeast

England

Null effect: Crime increased by 2% in the target

area, vs. a 5% increase in the control area; results

were not statistically significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"Dual

Estate,"

United

Kingdom

14 cameras in a

low income

residential area

with some

commercial

businesses

Undesirable effect: Crime increased by 4% in the

target area vs. a 19% decrease in the control area.

Results are statistically significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

- 65 -

Gill and

Spriggs, 2005

"Southcap

Estate,"

United

Kingdom

148 cameras in

a residential

area of South

London with

"deprived

housing"

Null effect: crime increased by 14%, compared to a

13% crime reduction in the control area. Results

are not statistically significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"Borough"

United

Kingdom

8 cameras

installed in

lamp posts in a

mixed-income

residential area

in southeast

England

Undesirable effect: crime increased by 73% while

the control area experienced a 12% increase in

crime. Results are statistically significant.

Good design: comparison

of crime rates crime rates

1 year before and after

CCTV intervention in

experimental and control

areas.

Gill and

Spriggs, 2005

"Deploy

Estate,"

United

Kingdom

11 cameras

along 3

residential

streets, and one

shopping area

Null effect: 21% increase in crime in the target area

compared to a 3% increase in the control area.

Results are not statistically significant.

Good design: comparison

of crime rates crime rates

1 to 2 years before and

after CCTV intervention

in experimental and

control areas.

Grandmaison

and Tremblay,

1997

Montreal,

Canada

13 subway

stations with 10

cameras each

Null effect: there was not significant change in

crime rates: total crime dropped 20% compared to

18.3% decline in the control; robbery declined by

27% compared to 30.8% in the control; assault

declined 27.5% compared to an increase of 5.6% in

the control; theft and fraud declined by 15.5% vs. a

16% decline in the control.

Good design: comparison

of crime rates 18 months

before and after in

experimental and control

areas.

- 66 -

Griffith, n.d. Gillingham,

United

Kingdom

Seven cameras

in the town

center

Desirable effect: crime fell 44% compared to 22%

decline in the control area; there was a reduction in

robberies and vehicle crime.

Good design: comparison

of crime rates 1 year

before and 4 years after

CCTV intervention in

experimental and control

areas.

Harada et al.,

2004 Tokyo,

Japan

Cameras in

Kabukicho area

of Tokyo

Desirable effect: 22% decrease in crime compared

to 11% decrease in control area and 9% decrease in

adjacent buffer areas. Reductions in vehicle crime

and larceny; small decrease in violent crime.

Good design: comparison

of crime rates 12 months

before and after CCTV

intervention in

experimental, adjacent,

and control areas.

Mazerolle et al.,

2002 Cincinnati,

Ohio

Cameras

installed in a

strip mall in a

Northside

neighborhood

Null effect: calls for service increased by 1.8% vs.

a 0% increase in the control

Good design: compared

the number of calls for

service 23 months before

and 6 months after the

implementation of the

CCTV system.

Mazerolle et al.,

2002 Cincinnati,

Ohio

Cameras

installed in a

mixed-use

commercial and

residential

neighborhood

Null effect: calls for service increased by 9.8% vs.

a 0% increase in the control.

Good design: comparison

of calls for service 23

months before and 4

months after the CCTV

intervention.

Mazerolle et al.,

2002 Cincinnati,

Ohio

Cameras

installed in

Findley Market

Place

Null effect: calls for service increased by 16.9% vs.

a 17.1% increase in the control area.

Good design: compared

calls for service 24.5

months before and 3.5

months after CCTV

intervention in an

experimental and control

area.

- 67 -

Musheno,

Levine, and

Palumbo, 1978

Bronxdale

Houses, Ney

York City,

USA

Cameras in a

public housing

complex

Uncertain effect: crime fell 9.4% (32 to 29) in the

experimental area, and fell 19.2% (26 to 21) in the

control area.

Good design: comparison

of crime rate 3 months

before and after CCTV

intervention in

experimental and control

areas.

Poyner, 1991 Guildford,

United

Kingdom

Cameras in the

parking lot at

the University

of Surrey

Undesirable effect: crime decreased less in the

experimental areas than in the control areas; theft

from vehicles declined by 73.3% compared to a

93.8% decline in the control area.

Good design: compared

theft from vehicles 24

months before and 10

months after the CCTV

intervention in

experimental and control

areas.

Sarno et al.,

1999 Camberwell,

London 17 cameras in

town center

Desirable effect: 12% reduction in crime; street,

vehicle, and violent crimes rates decreased at a

faster pace than prior to CCTV, while control areas

experienced an increase in crime.

Good design: comparison

of crime rates 24 months

before and after CCTV

intervention in

experimental, adjacent,

and control areas.

Sarno et al.,

1999

London,

United

Kingdom

12 cameras on

the East Street

market

Null effect: 10% decline in crime, with reductions

in vehicle crime, property damage, and robbery;

however, the crime in the buffer and control areas

decreased at a quicker rate than in the experimental

area. Street crime increased.

Good design: comparison

of crime rates 24 months

before and after CCTV

intervention in

experimental, adjacent,

and control areas.

Sarno et al.,

1999

London,

United

Kingdom

34 cameras

around the

Elephant and

Castle shopping

center

Desirable effect: crime fell by 17%; the number of

street robberies declined substantially.

Good design: comparison

of crime rates 24 months

before and after CCTV

intervention in

experimental, adjacent,

and control areas.

- 68 -

Sarno, 1995 London,

United

Kingdom

Cameras in a

parkling lot in

the Borough of

Sutton

Desirable effect: crime fell 57.3% vs. declines of

36.5% and 40.2% in the two control areas.

Good design: comparison

of crime 12 months

before and after the

CCTV intervention in

experimental and 2

control areas.

Sarno, 1995 London,

United

Kingdom

11 cameras in

the town center

in the Borough

of Sutton

Undesirable effect: crime did not decline as much

in the experimental area as in the control areas:

12.8% decline vs. 18% decline in control 1 and a

30% decline in control 2.

Good design: comparison

of crime rates 12 months

before and after CCTV

intervention in

experimental and 2

control areas.

Short and

Ditton, 1996

Airdrie,

Scotland,

United

Kingdom

12 cameras in

town center

Desirable effect: 35% reduction in crime vs. 12%

reduction in control area. Crimes of dishonesty and

vandalism decreased sharply. Some other types of

crime increased, likely due to increased detection.

Good design: comparison

of crime rates 24 months

before and after CCTV

intervention in

experimental and control

areas.

Sivarajasingam,

Shepherd, and

Matthews, 2003

United

Kingdom Five towns with

CCTV

Null effect: increased police detection of and

intervention in violent crimes and reduced hospital

visits due to violent crime compared to control

groups.

Good design: comparison

of crime rates and

assualt-related

hospitalizations 24

months before and after

CCTV intervention in

experimental and control

areas.

- 69 -

Skinns, 1998 Doncaster,

United

Kingdom

63 cameras in

the city centre

Desirable effect: crime dropped 21.3% vs. an

increase of 11.9% in the control. Vehicle crime fell

after the cameras were introduced; however, there

was no affect on other property crimes (burglary,

theft, shoplifting, and criminal damage). There was

some displacement of crime to outlying areas.

Good design: comparison

of crime rates 24 months

before and after CCTV

intervention in

experimental and control

areas.

Squires, 1998 Ilford, Essex,

United

Kingdom

Surveillance of

town center Desirable effect: 17% reduction in crime vs. a 9%

increase in the control.

Good design: comparison

of crime rates 6 months

before and after CCTV

intervention in

experimental and control

areas.

Squires, 2003

East

Brighton,

United

Kingdom

10 cameras in a

low-income

housing

complex

Null effect: crime continued to increase at a rate

comparable to the control.

Good design: comparison

of crime rates before and

after CCTV intervention

in experimental and

control areas; however,

crime data were not

broken down into

categories of crime.

Tilley, 1993 Hartlepool,

United

Kingdom Parking Lot

Desirable effect: vehicle theft declined by 59% vs.

a 16.3% decline in the control; Theft from vehicles

declined by 9.4% vs. a 3.1% increase in the control

area

Good design: comparison

of crime data 15 months

before and 30 months

after the CCTV

intervention.

- 70 -

Tilley, 1993 Bradford,

United

Kingdom Parking Lot

Desirable effect: vehicle theft declined by 43.5%

vs. increases of 5.9% and 31.8% in the two control

areas. Theft from vehicle declined by 68.8% vs.

increases of 4.5% and 6.1% in the two control

areas.

Good design: comparison

of crime data 12 months

before and after the

CCTV intervention in the

experimental area and

two control areas.

Tilley, 1993 Coventry,

United

Kingdom Parking Lot

Desirable effect: vehicle theft declined by 50.5%

vs. a 53.6% decline in the control area. Theft from

vehicles declined by 64.4% vs. a 10.7% in the

control area.

Good design: comparison

of crime rates 8 months

before and after in the

experimental area to

crime rates 16 months

before and after in the

control area.

Webb and

Laycock, 1992

London,

United

Kingdom

Cameras in

Underground

stations

Desirable effect: the number of robberies declined

by 62.3% compared to a 50% decrease in the

control area. The specific affect of CCTV versus

the other aspects of the strategy is unclear.

Good design: comparison

of crime rates 12 months

before and after CCTV

intervention in

experimental and control

areas. The specific affect

of CCTV versus other

aspects of the strategy is

unclear.

Webb and

Laycock, 1992

London,

United

Kingdom

Cameras in the

Oxford Circus

subway station

Undesirable effect: crime increased more in the

experimental area than in the control area:

robberies increased 47.1% compared to 21.4% in

the control; theft increased by 11.0% compared to a

1.9% decline in the control; assault increased by

29.4% compared to 36.4% in the control

Good design: compared

crime rates from 28

months before and 32

months after CCTV

intervention in

experimental and control

areas.

Williamson and

McLafferty,

2000

Brooklyn,

New York

Cameras in the

Albany and

Roosevelt

public housing

complexes

Null effect: no change in crime in the housing

project and the .1 mile buffer, compared to a 5.3%

and 4.0% decline in the control areas. There was a

reduction in major felonies.

Good design: comparison

of crime rates 18 months

before and after in

experimental, adjacent,

and control areas.

Winge and

Knutsson, 2003 Oslo,

Norway

6 cameras in the

Central Train

Station

Uncertain effect: overall recorded crime increased,

particularly violent and narcotics offenses;

however, the authors suggest that crime did not

actually increase, but rather that crime detection

improved. There was some decrease in robbery and

bicycle theft.

Good design: comparison

of crime rates 12 months

before and after CCTV

intervention in

experimental, adjacent,

and control areas.

- 71 -

- 72 -

APPENDIX B: SORTING THE DATA

We identified “suppressible” crimes, by sorting the data through several filters, which are

detailed below.

Crimes

1. Sorted by Hollywood, Jordan Downs and Nickerson Gardens police reporting-districts.

2. Sorted by crime type and crime code. Types of crimes likely to be suppressed by video

surveillance given the presence of the cameras and their locations were chosen. Crime types

excluded are Stolen Boat, Recovered Boat, Kidnap, Some Miscellaneous and Other coded

crimes, Rape, Recovered Vehicle, Sex Crimes, Trespass. These are not crimes that would have

occurred within the purview of the cameras we are studying, and therefore, would not be

suppressible by the cameras’ presence. Rape and Child Abuse crime-data was excluded due to

privacy restrictions. The crimes included are listed in the table below.

- 73 -

Crime Type Crime Codes

Aggravated Assault (AGG) 230 Assault with a Deadly Weapon

231 ADW against a Police Officer

235 Child Abuse (aggravated assault)

236 Spousal Abuse (aggravated assault)

250 Shots Fired at Moving Vehicle

251 Shots Fired Inhabited Dwelling

Burglary/Theft from a Vehicle (BTFV) 330 Burglary from Vehicle

331 Theft from Vehicle

410 Burglary from Vehicle (attempted)

420 Theft from Vehicle (petty)

421 Theft from Vehicle (attempted)

Burglary (BURG) 310 Burglary

320 Burglary (attempted)

Grand Theft Auto (GTA) 510 Vehicle, Stolen

520 Vehicle , Stolen (attempted)

Grand Theft Person (GTP) 350 Theft from Person

351 Pursesnatch

352 Pickpocket – Pickpurse

353 Drunkroll

450 Theft from Person (attempted)

451 Pursesnatch (attempted)

452 Pickpocket (attempted)

453Drunkroll (attempted)

Homicide (HOM), Miscellaneous (MISS) 110 Homicide

113 Manslaughter, Negligence

Robbery (ROBB) 210 Robbery

220 Robbery (attempted)

Vandalism (VAND) 740 Vandalism ($400+)

745 Vandalism (< $400)

Miscellaneous – Other Theft (MISS) 341 Theft, Grand (> $400)

440 Theft, Petty (< $400)

441 Theft (attempted)

480 Bicycle Stolen

485 Bicycle Stolen (attempted)

Other (OTH) 622 Battery on Fireman

623 Batter on Police Officer

624 Battery – Misdemeanor

625 Other Misdemeanor Assault

626 Spousal Abuse

753 Shots Fired

755 Bomb Threat

756 Bomb or Poss./Mfr. Destruct. Device

761 Brandishing Weapon

762 Lewd Conduct

805 Pimping

806 Pandering

882 Inciting Riot

886 Disturbing the Peace

995 Suspicious Activity Reports

3. Sorted by premise code. Included codes of locations that likely would be affected by the

presence of the video cameras in their specific locations, either by being in view of the cameras,

being open to the public and possibly in view of a camera or through location in the immediately

surrounding area to which crime might be displaced. Codes for locations that do not exist in any

of the test or control areas, that are not open to the public and that are by definition indoors were

excluded.

- 74 -

Category Description

Outside 101 Street/Parkway

102 Sidewalk

103 Alley

104 Driveway

105 Ped Overcrossing

106 Tunnel

107 Vacant Lot

108 Parking Lot

109 Park Playground

110 Freeway (inc. ramp)

116 Other outside

118 Construction Site

119 Porch (residential)

121 Yard

131 Redline Platform

132 Redline Mezzanine

135 MTA Prop/Prk. Lot

139 Stairwell

140 Balcony

142 Drive Thru

143 Escalator

145 Mailbox

142 Drive Thru

143 Escalator

145 Mailbox

146 Patio

147 Pool/Public

148 Public Restroom

- 75 -

123 Parking

Underground

124 Bus Stop

125 Pay Phone

127 Trashcan/Dumpster

128 Bus Stop/Layover

Outside

149 Riverbed

151 Tow Yard

152 Underpass/Bridge

Transportation 114 Taxi

122 Vehicle, Pass./Truck

Business 201 Jewelry Store

202 Liquor Store

203 Other Business

204 Mfr. Co.

205 Gun/Sporting Goods

206 TV/Radio/Appliance

207 Bar/Cocktail

208 Auto Sales

210 Restaurant/Fast Food

211 Pawn Shop

213 Warehouse

214 Bus Depot

215 Train Depot

216 Swap Meet

217 Auto Repair Shop

218 Beauty/Barber Shop

219 Cleaners

220 Nail Salon

221 Public Storage

222 Laundromat

223 Video Rental Store

224 Surplus Store

225 Music Store

228 Bowling Alley

229 Check Cashing

233 Tattoo Parlor

234 Optical Office

Manufacturing 301 Gas Station

Stores 401 Mini-mart

402 Market

403 Drug Store

404 Department Store

405 Clothing Store

406 Other Store

407 Hardware

408 Auto Supply

409 Beauty Supply

Schools 704 Elementary *

720 Jr. High *

721 High School *

722 College/Univ. *

723 Private/Preschool *

724 Trade School *

729 Special School *

Miscellaneous 705 Slip/Dock/Marina

706 Adult Bookstore

707 Garage/Carport

710 Other Premise

717 Health Spa/Gym *

726 Police Facility *

727 Shopping Mall *

Religious Facility 708 Church/Temple *

730 Synagogue *

731 Mosque *

Entertainment 711 Arcade

716 Theater/Movie

733 Bar/Sports Bar

735 Night Club

736 Skateboard Park

737 Skating Rink

City/County Government 725 Gov’t Facility *

732 Post Office *

* Crimes with these codes that occurred at an address were assumed to be indoors and excluded;

those that were recorded at intersections were assumed to be outside and included.

4. Geocoded into ArcMap GIS software to identify crimes in the target, buffer, and control areas.

Arrests

1. Sorted by Hollywood, Jordan Downs and Nickerson Gardens police reporting district.

2. Pulled data for arrests for Part II crimes (“victimless” crimes that usually go unreported unless

an arrest is made, such as drunkenness or prostitution) that could likely be detected by video

surveillance. These arrest categories are weapons, narcotics, drunkenness, gambling and other

miscellaneous crimes. Many categories had very low numbers given the size of the reporting

districts and time period of the cases represented. Aggregated Part II arrest data was utilized to

gauge generally the ability of video surveillance to detect or deter Part II crime.

3. Geocoded into ArcMap GIS, and crimes that occurred within the target, buffer, and control

areas were identified.

- 76 -

APPENDIX C: STATISTICAL ANALYSIS

Relative Effect Size Test

We employed the relative effect size statistical test that allows for a comparison of before-and-

after crime rates in two different locations. The calculation of the relative effect size is detailed

below:

Average monthly crime

rate before

implementation

Average monthly crime

rate after

implementation

Target Area a b

Control Area c d

Relative Effect Size = [a/(a+b)] / [b/(a+b)]

[c/(c+d)] / [d/(c+d)]

A relative size effect greater than one indicates that crime decreased in the test area at a faster

pace than, or did not increase as quickly as the control area. Alternatively, a relative effect size

less than one signifies that crime declined at a faster rate or did not increase as substantially in

the control area than in the test area.

To determine if the observed differences in crime rates are statistically significant and not due to

random variance, we calculated the 95% confidence intervals. If the entire confidence interval is

greater than one then the crime in the target area is said to have decreased significantly more

than in the control area. Alternatively, if the entire confidence interval is less than one, then the

crime in the control area is said to have decreased significantly more than in the target area. If

the confidence interval includes values both less than and greater than one, then the differences

in crime are not statistically significant. The confidence interval calculations are described

below:

95% Confidence Interval = relative effect size +/- (relative effect size * 2 * standard error)

Standard Error = var(a)/a2 + var(b)/b2 + var(c)/c2+var(d)/d2

Relative Effect Size Results

- 77 -

We compared before-and-after crime rates both between the target and control areas, as well as

between the buffer and control areas to test for displacement of crime areas or the diffusion of

benefits from the surveilled areas. None of the changes in crime in either our Hollywood or

Jordan Downs case study areas met the criterion for statistical significant, which is likely a result

of the large variance in crime month to month. The results of the relative effect size test and the

95% confidence intervals for the target and buffer areas are listed in the tables below.

Hollywood Boulevard Target Area Descriptive Crime Statistics

Pre-CCTV Period (25 Months) Post-CCTV Period (14 Months)

Crime Type

Total #

Crimes

% of

Total

Crime

Monthly

Rate Variance Total #

Crimes

% of

Total

Crime

Monthly

Rate Variance

Change

Monthly

Rate

All Crime 1,951 100.00% 78.04 176.1 980 100.00% 70.00 62.3 -10.30%

Violent Crime 865 44.34% 34.60 67.4 475 48.47% 33.93 32.4 -1.94%

Homicide 1 0.05% 0.04 0.0 1 0.10% 0.07 0.1 78.57%

Aggravated

Assault 221 11.33% 8.84 13.6 106 10.82% 7.57 19.0 -14.35%

Robbery 148 7.59% 5.92 6.7 79 8.06% 5.64 8.6 -4.68%

Battery 495 25.37% 19.80 32.8 289 29.49% 20.64 15.5 4.26%

Property

Crime 1,054 54.02% 42.16 60.4 485 49.49% 34.64 13.9 -17.83%

Vandalism 139 7.12% 5.56 4.1 94 9.59% 6.71 6.1 20.76%

Auto Theft 140 7.18% 5.60 4.0 61 6.22% 4.36 1.6 -22.19%

BTFV 254 13.02% 10.16 14.2 84 8.57% 6.00 8.0 -40.94%

Theft 521 26.70% 20.84 19.4 246 25.10% 17.57 10.3 -15.68%

Other 32 1.64% 1.28 1.3 20 2.04% 1.43 1.6 11.61%

Hollywood Boulevard Target Area Relative Effect Size (RES) by Crime Type

Crime Type

RES Upper

Confidence

Interval

Lower

Confidence

Interval

All Crime 0.99 1.54 0.44

Violent Crime 0.98 1.69 0.27

Homicide 0.22 3.26 -2.81

Aggravated

Assault 0.91 2.39 -0.57

Robbery 0.96 2.52 -0.61

Battery 1.05 1.90 0.20

Property Crime 1.02 1.64 0.39

Vandalism 0.82 1.86 -0.23

Auto Theft 0.97 2.16 -0.21

BTFV 1.23 2.99 -0.52

Theft 1.07 1.85 0.29

Other 0.99 4.05 -2.06

- 78 -

Hollywood Boulevard Target Area Crime RES and 95% Confidence Intervals

Hollywood Boulevard Target Area Descriptive Arrest Statistics

Column1

# of

Arrests

% of

Total

Arrests

Monthly

Rate Variance # of

Arrests % of

Total Monthly

Rate Variance

change

monthly

rate

Total Arrests 2,439 100.00% 97.6 117.0 1,470 100.00% 105.0 609.4 7.63%

Drunkenness 612 25.09% 24.5 56.1 372 25.31% 26.6 60.0 8.54%

Liquor Laws 10 0.41% 0.4 0.3 4 0.27% 0.3 0.2 -28.57%

Narcotics 896 36.74% 35.8 143.5 407 27.69% 29.1 71.0 -18.89%

Prostitution 6 0.25% 0.2 0.2 0 0.00% 0.0 0.0 -

100.00%

Weapon 17 0.70% 0.7 1.0 18 1.22% 1.3 1.3 89.08%

Other 898 36.82% 35.9 296.7 669 45.51% 47.8 138.2 33.03%

Hollywood Boulevard Target Area RES by Arrest Type

Arrest Type RES Upper

Confidence

Interval

Lower

Confidence

Interval

Total Arrests 1.04 1.75 0.34

Drunkenness 1.05 2.24 -0.13

Liquor Laws 0.98 6.32 -4.37

Narcotics 1.28 2.62 -0.07

Prostitution NA NA NA

Weapon 0.86 4.16 -2.45

Other 0.94 2.13 -0.24

- 79 -

Hollywood Boulevard Target RES and 95% Confidence Intervals

Hollywood Boulevard Buffer Area Descriptive Crime Statistics

Pre-CCTV Period (25 Months) Post-CCTV Period (14 Months)

Crime Type Total #

Crimes

% of

Total

Crime

Monthly

Rate Variance Total #

Crimes % of Total

Crime Monthly

Rate Variance

Change

Monthly

rate

All Crime 1,868 100.00% 74.7 268.4 983 100.00% 70.2 146.6 -6.03%

Violent Crime 709 37.96% 28.4 61.6 395 40.18% 28.2 37.6 -0.51%

Homicide 5 0.27% 0.2 0.2 1 0.10% 0.1 0.1 -64.29%

Aggravated

Assault 219 11.72% 8.8 11.9 90 9.16% 6.4 10.7 -26.61%

Robbery 151 8.08% 6.0 5.7 84 8.55% 6.0 4.3 -0.66%

Battery 334 17.88% 13.4 15.3 220 22.38% 15.7 11.8 17.62%

Property

Crime 1,141 61.08% 45.6 164.8 576 58.60% 41.1 96.4 -9.85%

Vandalism 197 10.55% 7.9 17.1 111 11.29% 7.9 10.7 0.62%

Auto Theft 209 11.19% 8.4 14.3 103 10.48% 7.4 9.3 -12.00%

BTFV 375 20.07% 15.0 24.2 183 18.62% 13.1 38.1 -12.86%

Theft 360 19.27% 14.4 22.4 179 18.21% 12.8 22.6 -11.21%

Other 18 0.96% 0.7 0.7 12 1.22% 0.9 1.1 19.05%

- 80 -

Hollywood Boulevard Buffer Area Relative Effect Size by Crime Type

Crime Type RES Upper

Confidence

Interval

Lower

Confidence

Interval

All Crime 0.95 1.58 0.31

Violent Crime 0.97 1.77 0.17

Homicide 1.11 13.13 -10.91

Aggravated

Assault 1.07 2.67 -0.54

Robbery 0.92 2.20 -0.37

Battery 0.93 1.72 0.15

Property Crime 0.93 1.73 0.13

Vandalism 0.98 2.48 -0.52

Auto Theft 0.86 2.12 -0.39

BTFV 0.84 1.99 -0.31

Theft 1.01 2.13 -0.10

Other 0.93 4.41 -2.55

Hollywood Boulevard Buffer Area RES and 95% Confidence Intervals

- 81 -

Hollywood Boulevard Buffer Area Descriptive Arrest Statistics

Pre-CCTV Period (25 Months) Post-CCTV Period (14 Months)

Arrest Type # of

Arrests % of

Total Monthly

Rate Variance # of

Arrests % of

Total Monthly

Rate Variance

% Change

Monthly

Rate

Total Arrests 1,987 100.00% 79.5 232.8 1,180 100.00% 84.3 221.8 6.05%

Drunkenness 483 24.31% 19.3 54.9 315 26.69% 22.5 22.6 16.46%

Liquor Laws 21 1.06% 0.8 1.2 10 0.85% 0.7 1.1 -14.97%

Narcotics 822 41.37% 32.9 51.4 432 36.61% 30.9 68.3 -6.15%

Prostitution 17 0.86% 0.7 0.8 2 0.17% 0.1 0.1 -78.99%

Weapon 18 0.91% 0.7 0.7 22 1.86% 1.6 2.1 118.25%

Other 626 31.50% 25.0 50.5 399 33.81% 28.5 56.9 13.82%

Hollywood Boulevard Buffer Area RES by Arrest Type

Arrest Type RES Upper

Confidence

Interval

Lower

Confidence

Interval

Total Arrests 1.06 1.77 0.34

Drunkenness 0.98 2.11 -0.15

Liquor Laws 0.82 5.06 -3.42

Narcotics 1.10 2.09 0.11

Prostitution 2.98 20.78 -14.83

Weapon 0.74 3.33 -1.85

Other 1.10 2.21 -0.01

Hollywood Boulevard Buffer Area Arrest RES and 95% Confidence Intervals

- 82 -

Hollywood Box Matched Pair Descriptive Crime Statistics

Pre-CCTV Period (25 Months) Post-CCTV Period (14 Months)

Crime Type # of

Crimes

% of

Total

Crime

Monthly

Rate Variance Total #

Crimes

% of

Total

Crime

Monthly

Rate Variance

Change

Monthly

Rate

All Crime 6,200 100.00% 248.0 1,540.6 3,087 100.00% 220.5 473.7 -11.09%

Violent Crime 2,434 39.26% 97.4 254.8 1,314 42.57% 93.9 169.2 -3.60%

Homicide 9 0.15% 0.4 0.4 2 0.06% 0.1 0.1 -60.32%

Aggravated

Assault 690 11.13% 27.6 62.6 302 9.78% 21.6 32.3 -21.84%

Robbery 531 8.56% 21.2 30.3 271 8.78% 19.4 54.4 -8.86%

Battery 1,204 19.42% 48.2 57.7 739 23.94% 52.8 55.1 9.60%

Property Crime 3,695 59.60% 147.8 784.8 1,729 56.01% 123.5 203.3 -16.44%

Vandalism 625 10.08% 25.0 54.9 345 11.18% 24.6 33.0 -1.43%

Auto Theft 645 10.40% 25.8 40.4 274 8.88% 19.6 36.0 -24.14%

BTFV 1,178 19.00% 47.1 192.1 481 15.58% 34.4 66.6 -27.09%

Theft 1,247 20.11% 49.9 91.5 629 20.38% 44.9 39.5 -9.93%

Other 71 1.15% 2.8 3.7 44 1.43% 3.1 3.1 10.66%

Hollywood Box Matched Pair Descriptive Arrest Statistics

Pre-CCTV Period (25 Months) Post-CCTV Period (14 Months)

Arrest Type # of

Arrests % of

Total Monthly

Rate Variance # of

Arrests % of

Total Monthly

Rate Variance

% Change

Monthly

Rate

Total Arrests 6,537 100.00% 261.5 1,051 4,111 100.00% 293.6 2,649.8 12.30%

Drunkenness 1,447 22.14% 57.9 300 926 22.52% 66.1 217.5 14.28%

Liquor Laws 64 0.98% 2.6 8 25 0.61% 1.8 5.0 -30.25%

Narcotics 2,308 35.31% 92.3 236 1,337 32.52% 95.5 500.3 3.44%

Prostitution 294 4.50% 11.8 41 103 2.51% 7.4 23.0 -37.44%

Weapon 75 1.15% 3.0 3 68 1.65% 4.9 10.6 61.90%

Other 2,349 35.93% 94.0 659 1,652 40.18% 118.0 392.0 25.59%

- 83 -

Jordan Downs Target Area Descriptive Crime Statistics

Pre-CCTV Period (45 Months) Post-CCTV Period (16 Months)

Crime Type # of

Crimes

% of

Total

Crime

Monthly

Rate Variance Total #

Crimes

% of

Total

Crime

Monthly

Rate Variance

Change

Monthly

Rate

All Crime 535 1.00 11.89 23.06 170 1.00 10.63 27.72 -10.63%

Violent Crime 289 0.54 6.42 14.11 82 0.48 5.13 9.32 -20.20%

Homicide 4 0.01 0.09 0.08 2 0.01 0.13 0.25 40.63%

Aggravated

Assault 82 0.15 1.82 3.19 24 0.14 1.50 1.60 -17.68%

Robbery 137 0.26 3.04 5.82 24 0.14 1.50 2.00 -50.73%

Battery 66 0.12 1.47 2.39 32 0.19 2.00 2.67 36.36%

Property Crime 237 0.44 5.27 9.75 86 0.51 5.38 20.25 2.06%

Vandalism 68 0.13 1.51 3.35 20 0.12 1.25 2.33 -17.28%

Auto Theft 68 0.13 1.51 1.85 23 0.14 1.44 3.73 -4.87%

BTFV 74 0.14 1.64 2.23 27 0.16 1.69 2.36 2.62%

Theft 27 0.05 0.60 0.61 16 0.09 1.00 1.20 66.67%

Other 9 0.02 0.20 0.21 2 0.01 0.13 0.12 -37.50%

Jordan Downs Target Area RES by Crime Type

Crime Type Relative Size

Effect

Upper

Confidence

Interval

Lower

Confidence

Interval

All Violent Crime 1.02 2.99 -0.96

Homicide 0.4 5.36 -4.56

Aggravated Assault 0.61 2.72 -1.5

Robbery 1.61 6.38 -3.17

Battery 0.85 3.51 -1.81

All Property Crime 0.65 2.21 -0.91

Vandalism 0.94 4.53 -2.65

Car Theft 0.9 4.23 -2.43

BTFV 0.36 1.81 -1.08

Theft 0.34 1.89 -1.22

Total Crime 0.84 2.14 -0.47

- 84 -

Jordan Downs Target Area Crime RES and 95% Confidence Intervals

Jordan Downs Target Area Descriptive Arrest Statistics

Pre-CCTV Period (45 Months) Post-CCTV Period (16 Months)

Arrest Type # of

Arrests % of

Total Monthly

Rate Variance # of

Arrests % of

Total Monthly

Rate Variance

Change

Monthly

Rate

Total Arrests 441 100.00% 9.80 77.28 292 100.00% 18.25 4,493.74 86.22%

Drunkenness 6 1.36% 0.13 0.17 16 5.48% 1.00 15.11 650.00%

Liquor Laws 2 0.45% 0.04 0.04 5 1.71% 0.31 1.76 603.13%

Narcotics 147 33.33% 3.27 12.09 98 33.56% 6.13 522.51 87.50%

Weapon 25 5.67% 0.56 0.72 3 1.03% 0.19 0.62 -66.25%

Other 261 59.18% 5.80 49.09 170 58.22% 10.63 1,551.13 83.19%

Jordan Downs Target Area RES by Arrest Type

Arrest Type RES Upper

Confidence

Interval

Lower

Confidence

Interval

Total Arrests 0.61 5.33 -4.11

Drunkenness 0.25 2.97 -2.47

Narcotics 0.72 6.37 -4.94

Weapon 1.88 8.08 -4.32

Other 0.52 4.67 -3.63

- 85 -

Jordan Downs Target Area Arrest RES and 95% Confidence Intervals

Jordan Downs Buffer Area Descriptive Crime Statistics

Pre-CCTV Period (45 Months) Post-CCTV Period (16 Months)

Crime Type

# of

Crimes

% of

Total

Crime

Monthly

Rate Variance # of

Crimes % of Total

Crime Monthly

Rate Variance

Change

Monthly

Crime

Rate

All Crime 285 100.00% 6.33 9.09 91 100.00% 5.69 11.03 -10.20%

Violent Crime 134 47.02% 2.98 3.79 42 46.15% 2.63 3.05 -11.85%

Homicide 3 1.05% 0.07 0.06 0 0.00% 0.00 0.00 -

100.00%

Aggravated

Assault 43 15.09% 0.96 1.13 11 12.09% 0.69 1.56 -28.05%

Robbery 60 21.05% 1.33 1.45 20 21.98% 1.25 1.27 -6.25%

Battery 28 9.82% 0.62 0.69 11 12.09% 0.69 0.50 10.49%

Property

Crime 147 51.58% 3.27 5.97 47 51.65% 2.94 3.93 -10.08%

Vandalism 39 13.68% 0.87 1.53 12 13.19% 0.75 1.00 -13.46%

Auto Theft 60 21.05% 1.33 1.50 16 17.58% 1.00 1.07 -25.00%

BTFV 42 14.74% 0.93 0.61 10 10.99% 0.63 0.78 -33.04%

Theft 6 2.11% 0.13 0.12 9 9.89% 0.56 1.20 321.88%

Other 4 1.40% 0.09 0.13 2 2.20% 0.13 0.25 40.63%

- 86 -

Jordan Downs Buffer Area RES by Crime Type

Buffer RES RES Upper Confidence

Interval Lower Confidence

Interval

All Violent Crime 0.92 2.87 -1.03

Aggravated Assault 0.70 4.08 -2.68

Robbery 0.85 3.42 -1.73

Battery 1.05 4.98 -2.88

All Property Crime 0.73 2.48 -1.01

Vandalism 0.90 4.71 -2.92

Car Theft 1.14 4.91 -2.62

BTFV 0.56 3.05 -1.93

Theft 0.13 1.09 -0.82

Total Crime 0.83 2.29 -0.62

Jordan Downs Buffer Area Crime RES and 95% Confidence Intervals

Jordan Downs Buffer Area Descriptive Arrest Statistics

Pre-CCTV Period (45 Months) Post-CCTV Period (16 Months)

Arrest Type # of

Arrests % of

Total Monthly

Rate Variance # of

Arrests % of

Total Monthly

Rate Variance

% Change

Monthly

Rate

Total Arrests 186 100.00% 4.13 10.6 87 100.00% 5.44 11.46 31.55%

Drunkenness 1 0.54% 0.02 0.0 2 2.30% 0.13 0.25 462.50%

Narcotics 65 34.95% 1.44 2.0 26 29.89% 1.63 1.85 12.50%

Weapon 12 6.45% 0.27 0.3 0 0.00% 0.00 0.00 -100.00%

Other 108 58.06% 2.40 6.1 59 67.82% 3.69 8.50 53.65%

- 87 -

Jordan Downs Buffer Area RES by Arrest Type

Arrest Type RES Upper Confidence

Interval

Lower

Confidence

Interval

Total Arrests 0.87 3.03 -1.29

Drunkenness 0.33 5.75 -5.09

Narcotics 1.20 4.68 -2.29

Other 0.62 2.58 -1.34

Jordan Downs Buffer Area Arrest RES and 95% Confidence Intervals

Nickerson Gardens Control Area Descriptive Crime Statistics

Pre-CCTV Period (45 Months) Post-CCTV Period (16 Months)

Crime Type # of

Crimes

% of

Total

Crime

Monthly

Rate Variance # of

Crimes

% of

Total

Crime

Monthly

Rate Variance

Change

Monthly

Rate

All Crime 1,032 100.00% 22.93 54.70 275 100.00% 17.19 27.36 -25.05%

Violent Crime 517 50.10% 11.49 17.53 149 54.18% 9.31 10.10 -18.94%

Homicide 10 0.97% 0.22 0.22 2 0.73% 0.13 0.12 -43.75%

Aggravated

Assault 146 14.15% 3.24 4.10 26 9.45% 1.63 2.52 -49.91%

Robbery 213 20.64% 4.73 8.29 60 21.82% 3.75 4.47 -20.77%

Battery 148 14.34% 3.29 3.66 61 22.18% 3.81 4.83 15.92%

Property

Crime 507 49.13% 11.27 21.61 119 43.27% 7.44 12.80 -33.99%

Vandalism 163 15.79% 3.62 4.56 45 16.36% 2.81 2.83 -22.35%

Auto Theft 141 13.66% 3.13 4.07 43 15.64% 2.69 2.76 -14.23%

BTFV 143 13.86% 3.18 5.47 19 6.91% 1.19 2.43 -62.63%

Theft 60 5.81% 1.33 1.55 12 4.36% 0.75 0.87 -43.75%

Other 8 0.78% 0.18 0.15 7 2.55% 0.44 0.26 146.09%

- 88 -

- 89 -

Nickerson Gardens Descriptive Arrest Statistics

Pre-CCTV Period (45 Months) Post-CCTV Period (16 Months)

Arrest Type # of

Arrests % of

Total Monthly

Rate Variance # of

Arrests % of Total Monthly

Rate Variance

Change

Monthly

Rate

Total

Arrests 1,830 100.00% 40.67 700.55 743 100.00% 46.44 187.20 14.19%

Drunkenness 60 3.28% 1.33 7.14 40 5.38% 2.50 6.80 87.50%

Liquor Laws 0 0.00% 0.00 0.00 0 0.00% 0.00 0.00 #DIV/0!

Narcotics 773 42.24% 17.18 101.47 370 49.80% 23.13 54.65 34.62%

Weapon 31 1.69% 0.69 1.13 7 0.94% 0.44 0.26 -36.49%

Other 966 52.79% 21.47 311.03 326 43.88% 20.38 63.32 -5.09%

APPENDIX D

Video Camera Specifications

Jordan Downs: Sony Ipela Hollywood: Pelco Spectra III

Color or day/night Color or day/night

Autofocus, lowlight optics Autofocus, lowlight optics

Frames per sec.: Max 25 Shutter Speed 1/2 – 1/30,000

Shutter speed Max 1 – 1/10,000 360-degree rotation

360-degree rotation 180-degree object tracking (autoflip)

180-degree object tracking Pan: 360 deg. at 150 deg./sec – .1 deg/sec.

Pan: 340 deg. at 300 deg./sec. Tilt: +2 deg. to -92 deg. at 200 deg./sec.

Tilt: 115 deg. at 300 deg./sec. Zoom (color: 16x optical, 8x digital)

Zoom: 26x optical 312x digital Zoom (day/night: 18x optical, 10x digital)

Resolution 450 lines Resolution 470 lines

Multiple image-compression formats Compatible w/fiber optic transmitters

Audio support Multilanguage onscreen menus

Manual/auto white balance Manual/auto white balance

Backlight compensation Onscreen compass

Frame integration (clear, smooth images) Color or day/night

Measuring The Effects Of Video Surveillance On Crime In Los Angeles (PDF) 210 Cameron

Navigation menu

Versions of this User Manual:

Views

Navigation