REPORT HIGHLIGHTS

OFFICE OF CITY AUDITOR SEATTLE

October 26, 2009

Cal Anderson Park

Surveillance Camera Pilot Program Evaluation

Audit Background

From January through February 2008, the City of Seattle installed three surveillance cameras in Capitol Hill's Cal Anderson Park. In June 2008, the Seattle City Council adopted an ordinance that created the Surveillance Camera Pilot Program and established controls over the cameras' use.

Audit Objectives

The ordinance governing the pilot program requires that the City Auditor conduct a program evaluation to address:

* Departmental compliance with the policies of the ordinance;

* The effect of the cameras on crime deterrence;

* The effect of the cameras on crime detection and investigation; and

* The effect of the cameras on the public perception of safety.

This program evaluation is intended to help the Seattle City Council decide whether to grant additional authority to the Department of Parks and Recreation, the Seattle Police Department, and the Department of Information Technology to operate surveillance cameras in Cal Anderson Park, or to install surveillance cameras in other City parks (SMC 18.14.030).

Findings

Departmental Compliance Each department involved with the pilot program complied with its policies, with one exception: the Department of Parks and Recreation (DPR) had the technical capability to access the surveillance cameras' pan, tilt, and zoom (PTZ) controls, which was prohibited by ordinance. According to the Department of Information Technology (DoIT), DoIT recently disabled DPR's PTZ controls to comply fully with the ordinance.

Crime Deterrence We were unable to determine whether the surveillance cameras in Cal Anderson Park had a deterrent effect on crime due to data limitations and a time frame, restricted by ordinance requirements, which limited our scope to three months of data.

Crime Detection and Investigation SPD's first use of the cameras to investigate a crime was in August 2009 following reports of roving gangs in Cal Anderson Park.

Public Perception of Safety Our survey's methodology limitations prevented us from confidently generalizing its results beyond our pool of respondents. The surveillance cameras in Cal Anderson Park appear to have had a minimal effect on our survey respondents' perception of safety in the park, which was dependent upon their awareness of the presence of the cameras. We found that less than one-third of the 103 respondents to the survey we conducted in Cal Anderson Park claimed to know about the cameras. Of those who knew about the cameras, only about four percent claimed that the cameras affected their perception of safety in the park. We were unable to determine whether this effect on respondents' perception of safety is sufficient for the purposes of the pilot program, due to the program's lack of clear goals, specific performance measures, and defined benchmarks to determine program success or failure.

Recommendations

We made nine recommendations to improve the City's ability to evaluate the Surveillance Camera Pilot Program, to increase the effectiveness of the cameras, and to improve departmental compliance with the requirements of Ordinance 122705.

Departmental Compliance

If the Seattle City government (the City) wants to fully comply with Chapter 18.14 of the Seattle Municipal Code (SMC), which authorizes the Surveillance Camera Pilot Program, the following recommendations should be implemented:

Recommendation 1: To aid program oversight, The Department of Parks and Recreation (DPR), the Department of Information Technology (DoIT), and the Seattle Police Department (SPD), should amend their log books to include a place where the user may record the name, title, and signature of the authorizing supervisor as one of the personnel whose involvement is required for using the cameras or viewing camera footage. According to DPR they have made changes to comply with this recommendation. Also, the City should consider clarifying SMC 18.14.090 to indicate that the authorizing supervisor is among the "personnel involved," whose name and title must appear in the log book entries.

Recommendation 2: To help prevent future confusion about City department responsibilities in implementing the Surveillance Camera Pilot Program, the City should consider revising Section 2 of Ordinance 122705 by replacing the reference to the "City Clerk" with the "City Auditor."

Crime Deterrence, Detection and Investigation

If the City wants to understand the effect of the cameras on crime deterrence:

Recommendation 3: The City should: (a) develop a more valid and geographically specific incident reporting method than it currently uses, (b) install cameras in places where they may be both visible and able to record images throughout each park, and (c) select parks that are easily compared to one another in terms of characteristics that are relevant to the implementation of the program (e.g., camera positions and visibilities, crime trends, park user demographics, landscapes, etc.).

Recommendation 4: The City should improve the method by which it records and retains park exclusion notices, which prohibit people caught committing violations of City ordinances from entering a park for a year. DPR has indicated they agree with this recommendation and will work with SPD on recordkeeping and retention issues.

Recommendation 5: If the City wishes to assess the effectiveness of the surveillance cameras in assisting SPD in detecting and investigating crimes, then it should authorize another evaluation of the Surveillance Camera Pilot Program after at least one year from the time SPD employees completed their training on the cameras. This should give SPD enough time to make more of its employees aware of the cameras, and to use the cameras to detect and investigate crimes.

Public Perception of Safety

If the City wants to maximize the public's perception of safety and measure the cameras' effect on that perception:

Recommendation 6: The City should increase public awareness of the surveillance cameras in City parks by installing more visible signs and periodically engaging in publicity campaigns.

Recommendation 7: The City should consider modifying the Seattle Municipal Code to authorize SPD personnel to view the live surveillance footage more frequently, while balancing public safety concerns against privacy concerns.

Recommendation 8: The City should establish clear program goals, specific performance measures, and defined benchmarks to determine program success or failure. Definitions of success or failure should address expected or desired levels of change in the areas of crime deterrence, crime-solving, and perception of public safety.

Recommendation 9: To inform its decision about whether to continue using surveillance cameras in parks, the City should consider: (a) authorizing and funding a more extensive survey to evaluate the effect of the surveillance cameras on park users' perception of public safety; (b) seek further public comment on the presence of the cameras in Cal Anderson Park by requiring DPR to hold one or more public hearings in Cal Anderson Park; and (c) using the public comments DPR will receive as it convenes community meetings in other parks per SMC 18.14.040.B to inform the decision about whether to install cameras in yet more parks.

Office of City Auditor

Cal Anderson Park Surveillance Camera Pilot Program Evaluation

October 26, 2009

Auditor-in-Charge: Dean Runolfson, Audit Intern

Audit Manager: Mary Denzel

City of Seattle

Office of City Auditor

Our Mission:

To help the City of Seattle achieve honest, efficient management and full accountability throughout City government. We serve the public interest by providing the Mayor, the City Council, and City department heads with accurate information, unbiased analysis, and objective recommendations on how best to use public resources in support of the well-being of the citizens of Seattle.

Background:

Seattle voters established our office by a 1991 amendment to the City Charter. The office is an independent department within the legislative branch of City government. The City Auditor reports to the City Council and has a four-year term to ensure his/her independence in selecting and reporting on audit projects. The Office of City Auditor conducts financial-related audits, performance audits, management audits, and compliance audits of City of Seattle programs, agencies, grantees, and contracts. The City Auditor's goal is to ensure that the City of Seattle is run as effectively and efficiently as possible.

How We Ensure Quality:

The office's work is performed in accordance with the Government Auditing Standards issued by the Comptroller General of the United States. These standards provide guidelines for staff training, audit planning, fieldwork, quality control systems, and reporting of results. In addition, the standards require that external auditors periodically review our office's policies, procedures, and activities to ensure that we adhere to these professional standards.

An equal opportunity employer

Street Address: 700 5th Avenue, Suite 2410, Seattle, WA

Mailing address: PO Box 94729, Seattle, Washington 98124-4729

Phone Numbers: Office: (206) 233-3801 Fax: (206) 684-0900

E-mail: davidg.jones@seattle.gov

Website: seattle.gov/audit

City of Seattle

Office of City Auditor

David G. Jones, Acting City Auditor

October 26, 2009

The Honorable Greg Nickels

Seattle City Councilmembers

City of Seattle

Seattle, Washington 98104

Dear Mayor Nickels and City Councilmembers:

Attached is our report, Cal Anderson Park Surveillance Camera Pilot Program Evaluation. Our objectives for this work were to assess departmental compliance with the Surveillance Camera Pilot Program polices specified in Ordinance 122705, and to evaluate the effectiveness of the surveillance cameras within Cal Anderson Park in: (1) deterring and solving crime, and (2) improving the public's perception of safety within Cal Anderson Park.

We incorporated responses into this report from the Department of Parks and Recreation, the Department of Information Technology, the Seattle Police Department, and the Department of Finance. We have included final comments from Executive departments in Appendix J to the report. Our responses to these comments are contained in Appendix K.

We appreciate the cooperation received from the Department of Parks and Recreation, the Department of Information Technology, the Seattle Police Department, the Department of Finance, the City Attorney's Office, the City Clerk's Office, and citizens who provided information and insights during our evaluation. If you have any questions or comments about this report, please call Mary Denzel, Audit Manager at (206) 684-8158, or me at (206) 233-1095.

Sincerely,

David G. Jones

Acting City Auditor

Enclosure

This page intentionally left blank

Table of Contents

I. Introduction and Background 1

II. Scope and Methodology 2

III. Departmental Compliance 5

IV. Crime Deterrence, Detection, and Investigation 7

V. Perception of Safety 16

Appendix A Ordinance 122705 21

Appendix B Departmental Compliance with Seattle Municipal Code Requirements 28

Appendix C Quasi-Experimental Evaluation Methods 37

Appendix D Crime Trends in the Cal Anderson Park Neighborhood 40

Appendix E Survey Overview and Analysis 47

Appendix F Survey Questions 53

Appendix G Survey Question Frequency Tables 59

Appendix H Responses to Question 2c Regarding Safer Areas in Cal Anderson Park 69

Appendix I Responses to Question 11 Regarding Surveillance Cameras in Cal Anderson Park 71

Appendix J Executive Departments' Comments on This Report 75

Appendix K Auditor's Comments on the Executive's Comments 80

This page intentionally left blank

I. Introduction and Background

Seattle City Council Ordinance 122705 directed the Office of City Auditor (OCA) to conduct an audit of a pilot program for using surveillance cameras in City parks. This audit report fulfills this directive and documents the program evaluation we performed regarding the City of Seattle's Surveillance Camera Pilot Program. In this section we present the background and context of the Surveillance Camera Pilot Program, including: (1) planning and installation of the surveillance cameras; (2) scope limitations on the program evaluation; (3) features of the surveillance camera system; and (4) trends in the public surveillance industry.

Surveillance Camera Pilot Program Planning and Installation

According to the Department of Parks and Recreation (DPR) Superintendent, the proposal to install surveillance cameras in Seattle City parks came from the Center City Parks Task Force. This task force included surveillance cameras as one option to revitalize Seattle's downtown City parks and public spaces. Some of the other options included: increased lighting, landscape design changes, and creating the Park Ranger program. In his September 17, 2007 speech presenting the 2008 budget, Seattle Mayor Greg Nickels proposed funding several Task Force recommendations:

My budget proposal will hire and train uniformed Park Rangers to help people enjoy our open space while providing an extra set of eyes to prevent illegal activities. We will supplement the new patrols with a pilot program to install security cameras to discourage crime. We will invest in additional lighting, features and landscaping and we will expand music, food, art, and other activities to keep these parks fun and compelling.1

After the Seattle City Council adopted the 2008 budget, the City completed the installation of three surveillance cameras in Cal Anderson Park in February 2008 (see Figure 1).

Figure 1 Surveillance Cameras Installed in Cal Anderson Park

At a Seattle City Council Parks and Seattle Center Committee meeting on April 22, 2008, the Committee discussed potential privacy issues raised by the presence of surveillance cameras in City parks. Between April and June 2008 the City Council developed protocols to govern the use of surveillance cameras in City parks. In June 2008, the City Council passed Ordinance 122705, amending Seattle Municipal Code (SMC) Chapter 18.14 and creating the Surveillance Camera Pilot Program.2 Ordinance 122705 authorized the expansion of the surveillance camera system into Hing Hay, Occidental, and Victor Steinbrueck Parks. The Mayor signed Ordinance 122705 into law in June 2008. Due to budget limitations, no additional cameras have been installed.

Features of the Surveillance Camera System

The City installed three surveillance cameras in Cal Anderson Park. They are located on two light poles that stand at the north end of the Bobby Morris playfield, along a walkway that runs the width of the park. Figure 2 on the following page shows the three camera locations with green circles. Control and monitoring stations are located in Department of Parks and Recreation (DPR), Department of Information Technology (DoIT), and Seattle Police Department (SPD) facilities, and connect to the cameras via a server in the Seattle Municipal Tower. These control and monitoring stations enable the surveillance camera system's authorized users to manipulate the cameras (i.e., pan, tilt, and zoom), view live surveillance footage, and both view and retrieve recorded surveillance footage (see Figure 3). The user may pan 360 degrees , tilt 180 degrees , and optically zoom to enlarge objects up to 35 times their original size. According to DoIT, it is working with DPR to disable the pan, tilt, and zoom capabilities (though not the monitoring capabilities) at the DPR control and monitoring station to comply with the pilot program's protocols.

Public Surveillance Industry Trends

From our literature review, we noted three main trends in the public surveillance industry: (1) surveillance cameras are being installed in an increasing number of cities in the United States (U.S.); (2) U.S. cities are integrating their surveillance camera systems into their emergency communications operations; and (3) lack of funding in both Europe and the United States is resulting in a decrease in the live monitoring of surveillance camera footage. Independent studies of surveillance cameras' effect on crime suggest that the cameras may help to deter property crime within the area immediately around the cameras, but they appear to have little or no measurable effect on other crimes. While privately owned surveillance camera footage aids law enforcement in identifying and arresting suspects in criminal investigations including recent investigations conducted by the SPD experts in academia have yet to come to a consensus on whether or not surveillance cameras are a cost effective tool for helping police to make criminal arrests that result in convictions.

II. Scope and Methodology

The criteria stated in Ordinance 122705 suggest three broad questions we used to frame our program evaluation:

1. Are the Department of Parks and Recreation, the Seattle Police Department, the Department of Information Technology, and the Office of City Auditor complying with the policies set forth in Ordinance 122705?

2. Do the surveillance cameras in Cal Anderson Park aid in deterring and solving crime?

3. Do the surveillance cameras in Cal Anderson Park affect the public perception of safety?

Below, we discuss the scope of our research questions and introduce the general methods by which we addressed them:

Figure 2 Surveillance Camera System in Cal Anderson Park

Figure 3 Surveillance Camera System Monitoring and Control Interface

Question 1: Departmental Compliance The ordinance requires that we assess departmental compliance with the program policies governing: (1) the installation and use of the cameras; (2) handling of video recordings created by the cameras; (3) live monitoring of the cameras; (4) prohibiting SPD active monitoring for an improper purpose; (5) access to recordings created by the cameras; (6) recordkeeping; and (7) auditing and evaluation of the pilot program. To assess compliance with these policies, we conducted site visits of department facilities, interviewed department representatives, examined the logbooks that the departments used to record their use of the surveillance camera system, and reviewed randomly selected surveillance recordings.

Question 2: Crime Deterrence, Detection and Investigation The ordinance requires that we consider the following information to assess the degree to which the cameras deter or help to solve crime:

1. Crime statistics and the number of calls for assistance in the monitored parks for the first 12 months after the cameras are installed;

2. Crime statistics and the number of calls for assistance in the monitored parks for the two years preceding the installation of the cameras;

3. Crime statistics and the number of calls for assistance for the areas just outside the monitored parks for the first 12 months after the cameras are installed;

4. Crime statistics and the number of calls for assistance for areas just outside the monitored parks for the two years preceding the installation of the cameras;

5. A description of crime trends over the past five years in the neighborhoods where the monitored parks are located;

6. The number of crimes detected in the monitored parks due to the presence of the cameras. Data should be provided for the first 12 months after the cameras are installed;

7. The number of crime investigations aided by the use of video recordings obtained by the cameras. A description of how the video recordings were helpful to each investigation should be included. Data should be provided for the first 12 months after the cameras are installed.3

After assessing the usefulness and geographic specificity of the data, we analyzed crime statistics and trends from both Computer Aided Dispatch (CAD) data between 2003 and 2009, and Park Exclusion Notice (PEN) data between 2006 and 2009, by creating and graphing incident numbers during a unit of time (e.g. three months or one year).

Question 3: Public Perception of Safety The ordinance requires that we conduct a survey to assess the effect of the surveillance cameras in Cal Anderson Park on the public's perception of safety in that park. We created and conducted a face-to-face survey that asked questions of park users about their knowledge of the presence of the surveillance cameras and their effect on respondents' perception of safety in the park. We then tested the survey results for statistical significance.

Scope Limitations on the Program Evaluation

While Ordinance 122705 is not clear about the start date of the pilot program, it requires that the City Auditor provide "a program evaluation within 18 months of the completion of the installation of the cameras" (emphasis added). We began our program evaluation in October 2008 with the intent to analyze data from all four parks included in the pilot program. However, by November 2008 the City Council significantly reduced the funds allocated to the Center City Parks Security Capital Improvement Program, which included the Surveillance Camera Pilot Program. By February 2009, it was clear that Cal Anderson Park would be the only park with the pilot program's surveillance cameras. To best comply with Ordinance 122705, we chose to complete our program evaluation within 18 months of the installation of the cameras in Cal Anderson Park. The Ordinance states that the program evaluation shall include crime statistics and the number of calls for assistance for the first 12 months after the cameras are installed. For Cal Anderson Park, this period was February 2008 through February 2009.

Although the cameras were installed in February 2008, we had only three months of data on which to base our evaluation of the efficacy of the Surveillance Camera Pilot Program in investigating and solving crime due to the following: (1) the cameras were not available for use until protocols were established in June 2008 with the passage of Ordinance 122705, and (2) SPD was not able to complete training for the staff that would use the cameras until November 2008. So SPD did not have use of the cameras for crime investigation between February 2008 and November 2008; and they were only used for training purposes between June and November 2008. Consequently we could only measure the impact of camera-usage on crime investigation from mid-November 2008 through February 2009.

III. Departmental Compliance

Ordinance 122705 governing the Surveillance Camera Pilot Program requires that the City Auditor, "audit the DPR, DoIT, and SPD log books; corresponding 911 calls and other police contacts; and video recordings throughout the pilot program to aid in program evaluation and to assess compliance with this chapter".4 During the period we conducted our program evaluation, there were no corresponding 911 calls or other police contacts. However, we audited the log books and video recordings because the cameras had been used for training purposes between June and November 2009. We found that by the time we completed our audit work City departments involved with the Surveillance Camera Pilot Program were in compliance with the City policies and protocols established in Ordinance 122705.

After providing a brief description of the general roles of the departments in the pilot program, we present our assessment of departmental compliance with policies and procedures regarding the camera system. We then discuss our conclusions and recommendations regarding departmental compliance.

Description of Department Roles

SPD is the end-user of the surveillance camera footage and recordings. The ordinance grants SPD authority to use the camera footage: (1) proactively (i.e., to prevent a crime from occurring or escalating) by engaging in live and active monitoring5 under the limited conditions described in SMC 18.14.060; and (2) retroactively (i.e., to investigate crimes that have already occurred) by reviewing surveillance recordings. DPR and DoIT provide technical support to SPD. DoIT owns the cameras and related equipment; it is responsible for ensuring the surveillance system is in working order. DPR owns the surveillance recordings; it is responsible for providing copies of the recordings upon request to both SPD and citizens who make a public disclosure request. The City Auditor is responsible for a program evaluation of the pilot program and its participating departments. For more information regarding department roles in the pilot program, see Table 1.

Summary of Findings and Recommendations Related to Program Compliance Assessment

We assessed departmental compliance with policies stated in the ordinance regarding: (1) the installation and use of the cameras; (2) video recordings created by the cameras; (3) live monitoring of the cameras; (4) prohibiting SPD active monitoring for an improper purpose; (5) access to recordings created by the cameras; (6) recordkeeping; and (7) auditing and evaluation of the pilot program. Appendix B reports departmental compliance with each specific policy.

We found that City departments involved with the Surveillance Camera Pilot Program complied with the City policies and protocols established in Ordinance 122705. The one exception we found was that the Department of Parks and Recreation had the technical capability to access the surveillance cameras' pan, tilt, and zoom controls, which is prohibited by Ordinance 122705. According to the Department of Information Technology (DoIT), DoIT recently disabled DPR's pan, tilt, and zoom controls to comply fully with the ordinance.

However, while SMC 18.14.090.A and B require that DPR, DoIT, and SPD record the personnel involved whenever the departments access the surveillance camera system, none of the departments clearly recorded the title of the supervisor who authorized department personnel to access the surveillance camera system.6 While Ordinance 122705 does not explicitly state the supervisor's title should be included in the log books, it is a good practice for demonstrating compliance with the ordinance provision that all personnel required enter their names in the log book.

The language of Ordinance 122705 requires that, "DPR, DoIT, SPD, and the City Clerk shall take the steps necessary to implement this ordinance."7 Ordinance 122705 assigns no responsibilities to the City Clerk, whereas the City Auditor is one of the departments responsible for implementing the ordinance.

Recommendation 1: To help ensure that a department supervisor authorizes personnel accessing the surveillance camera system, DPR, DoIT, and SPD should consider amending their log books to include a place where the user may record the name, title, and signature of the authorizing supervisor as one of the personnel whose involvement is required for using the cameras or viewing camera footage. Also, the City should consider clarifying SMC 18.14.090 to require the departments to include the name, title, and signature of the authorizing supervisor among the "personnel involved" in the log book entries.

Recommendation 2: To help prevent future confusion about City department responsibilities in implementing the Surveillance Camera Pilot Program, the City should consider revising section 2 of Ordinance 122705 by replacing the "City Clerk" with the "City Auditor."

IV. Crime Deterrence, Detection, and Investigation

Ordinance 122705 governing the Surveillance Camera Pilot Program requires that the City Auditor conduct a program evaluation that includes the following information:

1. Crime statistics and the number of calls for assistance in the monitored parks for the first 12 months after the cameras are installed;

2. Crime statistics and the number of calls for assistance in the monitored parks for the two years preceding the installation of the cameras;

3. Crime statistics and the number of calls for assistance for the areas just outside the monitored parks for the first 12 months after the cameras are installed;

4. Crime statistics and the number of calls for assistance for areas just outside the monitored parks for the two years preceding the installation of the cameras;

5. A description of crime trends over the past five years in the neighborhoods where the monitored parks are located;

6. The number of crimes detected in the monitored parks due to the presence of the cameras. Data should be provided for the first 12 months after the cameras are installed;

7. The number of crime investigations aided by the use of video recordings obtained by the cameras. A description of how the video recordings were helpful to each investigation should be included. Data should be provided for the first 12 months after the cameras are installed.8

We found that we were unable to determine whether the surveillance cameras in Cal Anderson Park had a deterrent effect on crime due to data limitations because (1) Computer Aided Dispatch (CAD) data from the Seattle Police Department may not record the location of an actual incident, and (2) available CAD data are unable to distinguish between crimes that occurred in the park from those that occurred just outside the park. Furthermore, because of requirements imposed by Ordinance 122705 we were limited to analyzing only three months of data. The cameras have only been used once, in August 2009, for crime investigation, but were not helpful in that instance.

In this chapter we discuss: (1) relevant crime statistics that SPD collects; (2) crime trends in the area around Cal Anderson Park; and (3) crime detection and investigation with the surveillance cameras in Cal Anderson Park.

Crime Statistics

We were unable to confidently provide crime statistics and the number of calls for assistance (i.e., 911 calls and officer back-up calls) in or just outside Cal Anderson Park for any period of time due to data limitations. We explain this conclusion by discussing: (1) the nature and usefulness of crime data that SPD collects; and (2) quasi-experimental methods to determine from crime statistics the effect the surveillance cameras may have on crime.

Data Limitations Prevent Valid Comparison of Crime Statistics

We consulted with SPD representatives before requesting the crime statistics that Ordinance 122705 requires, and learned that:

* Computer Aided Dispatch (CAD) data9 may not record the location of an actual incident. For example: if someone was assaulted in Cal Anderson Park, the victim could go to a hospital and report the crime upon arrival; thus, the record would tie the crime to the hospital, and not the park. This reduces the validity of locations in the CAD data.

* CAD data are unable to distinguish between crimes that occurred in the park and those that occurred just outside the park. Each officer reports incident locations in a different way. The variation in reporting further reduces the validity of the locations in CAD data. Though reporting the closest physical address is the preferred and most common method, this is not possible for officers who report an incident that occurred in a park. If an officer records that an incident occurred in "Cal Anderson Pk" (as written in CAD data), then the geocoding software that SPD crime analysts use to map incident locations would place the incident at the park's official address (1100 E Pine St), located at the intersection of 11th Ave. and E Pine St. When mapped, any crimes that occurred in Cal Anderson Park would be mixed with those crimes that occurred at that intersection. Thus, we are unable to present the locations of criminal incidents within Cal Anderson Park.

* Not every criminal incident is included in police crime data. For example, after we interviewed a board member of the Cal Anderson Park Alliance10 on November 18, 2008, the citizen provided us with a photograph of what was described as, "the worst case of vandalism in the history of the park" (see Figure 4). However, there is no record of this vandalism in the CAD data on or after the date of the incident (March 30, 2008).

Figure 4 Vandalism on the Cal Anderson Park Fountain, 3/30/08

SPD Data Collection Does Not Support Use of Quasi-Experimental Methods

We conducted a literature review to identify methods other program evaluations have used to assess the effectiveness of public surveillance cameras in deterring crime. Three significant methods used in camera evaluations are: 1) Measuring the distances of crime incidents from the cameras, 2) Comparing those areas in which the public can see the cameras with areas in which the public cannot see the cameras; and 3) Comparing the crime trends between similar areas with surveillance cameras, which requires that differences between experimental and comparison groups be minimized or controlled for evaluation purposes. We could not apply these quasi-experimental methods for this evaluation because all crimes in Cal Anderson Park are identified as occurring at a single location, so it is impossible to measure the distance of a crime from the cameras, or to determine if a crime happened in an area where the public can see the cameras. Furthermore, it is very difficult to compare one City park to another with sufficient rigor for quasi-experimental methods. See Appendix C for additional information on obstacles to using quasi-experimental methods to derive conclusions from the City's crime statistics.

Recommendation 3: If Seattle wishes to evaluate the effectiveness of the surveillance cameras in deterring crime in City parks, then:

(a) SPD should develop a more valid and geographically specific incident reporting method than it currently uses. This improved reporting method should enable police officers to conveniently, accurately, and uniformly report the location of an incident in a park or other area where there is no street address.

(b) The City should install cameras in places where they may be both visible and able to record activities throughout each park. This should help to fulfill the evaluative need to identify crimes that occurred within view of the cameras.

(c) The City should select parks that are easily compared to one another in terms of significant characteristics that are relevant to the implementation of the program (e.g., camera positions and visibilities, crime trends, park user demographics, landscapes, etc.).

Crime Trends

As required by Ordinance 122705, we provide crime trends in the area around Cal Anderson Park. These trends cannot establish the effect of the surveillance cameras on crime in Cal Anderson Park because there are too many variables that affect crime rates. Nevertheless, as required by Ordinance 122705, we present and discuss: (1) crime trends from CAD data in three census tracts that surround Cal Anderson Park between February 2003 and February 2009; and (2) park exclusions in Cal Anderson Park from the beginning of 2006 through the end of February 2009. Table 2 describes crime trends for major crime categories from 2003 through 2009.

CAD Data Trends: February 2003 through February 2009

Ordinance 122705 requires that we describe crime trends, "in the neighborhoods where the monitored parks are located."11 The City Attorney's Office reported that there is no legal definition of "neighborhood" in Seattle. We were going to define the park's neighborhood by the police beats12 that surround it, but SPD representatives informed us that the beat boundaries had been changed in 2008. We then decided to define the neighborhood around Cal Anderson Park to consist of the three census tracts that surround it (see Figures 5 and 6). 13

Figure 5 Cal Anderson Park Neighborhood by Census Tract

Figure 6 Cal Anderson Park Neighborhood by Census Tract (enlarged)

The census tracts' boundaries have remained consistent since the 2000 census. We analyzed statistics and trend data for the Cal Anderson Park neighborhood, aggregating census tracts 074, 075 (contains Cal Anderson Park), and 084. Ordinance 122705 also does not specify what crime trends we need to describe. After viewing CAD data (see Table 3 for an example), we learned that we could define crime trends by: incident type (TYPE), Miscellaneous Incident Report (MIR) code, priority (PRI), and disposition (DISP). For the purposes of this evaluation, we found that: (1) MIR codes were too detailed and specific; (2) PRI was not detailed enough; and (3) DISP referred to the result of an incident, rather than describing the incident itself. We chose to describe crime trends as defined by TYPE, because that definition provided sufficient detail without unnecessary specificity.

The CAD data we analyzed do not show any consistent trend. Appendix D displays quarterly crime counts for twenty-one crime types from 2003 through the first quarter of 2009.

Table 3 CAD Data Sample

Park Exclusion Notices

Police officers and park rangers issue park exclusion notices (PENs) for offenses that occur in the parks, including violations of any: (1) provision of the SMC or RCW; (2) park rule (as designated by the Superintendent of Parks; or (3) provision of SMC Chapter 18.12 (City of Seattle Parks Code). Individuals who receive PENs are banned from entering one or more parks for up to one year, depending upon the park location and the severity of the offense.

Neither SPD nor DPR compile records of PENs in an electronic database. To learn the total number of PENs SPD and DPR issued in Cal Anderson Park during some period, the paper records of the notices had to be handcounted. We received numbers of PENs from SPD, based upon hand-counted records from the East Precinct, counted and provided by the Community Police Team (see Table 4). SPD did not have data on PENs issued before 2006. These data suggest a decrease in the number of PENs issued in Cal Anderson Park during the year 2008. Though the timing of the installation of the surveillance cameras in January-February 2008 supports the hypothesis that the surveillance cameras may have contributed to this decrease, we are unable to determine whether this is the case. Three possible alternative hypotheses are: (1) PENs are correlated with the crimes in the CAD data that appear to be downward trending during 2008; (2) the initiation of the Park Rangers Program in April 2008 may have had a deterrent effect on behavior that would warrant PENs; and (3) the decrease is simply a random fluctuation.

Recommendation 4: If Seattle intends to include park exclusion notices in its evaluation of surveillance camera effectiveness in deterring crime in City parks, then it should use a more sophisticated method of tallying the notices over time. The City should keep records of the notices for a longer period (i.e., at least five years). It should also consider changing the format of the notices to include specific geographical locations within the park where the offender committed the violation.14 DPR has indicated they agree with this recommendation and will work with SPD on record retention issues.

Crime Detection and Investigation

Ordinance 122705 requires that we report: (1) the number of crimes detected in Cal Anderson Park due to the presence of the surveillance cameras; and (2) the number of crime investigations aided by the use of video recordings produced by the surveillance cameras. SPD reported to us on June 11, 2009, that both numbers were zero: SPD had not detected any crimes in Cal Anderson Park due to the presence of the cameras. SPD made a request to view surveillance recordings of Cal Anderson Park for a criminal investigation in August 2009, but the camera footage was not useful in that investigation. However, in August 2009 private security camera footage helped SPD solve three crimes in other areas of the City. There is no question that when a crime is captured on video, the film can help in solving the crime and sometimes with prosecution as well.

An SPD official stated that SPD has been very cautious in its use of the cameras to comply with the requirements of Ordinance 122705. SPD reported that the program is "very young," and that SPD anticipates that the cameras will be used more as its employees become more aware of the cameras. SPD trained fourteen 911 Communications Center employees for the Surveillance Camera Pilot Program (see Figure 7). These employees completed their training by November 14, 2008.

Recommendation 5: If the City wishes to assess the effectiveness of the surveillance cameras in assisting SPD in detecting and investigating crimes, then it should authorize another evaluation of the Surveillance Camera Pilot Program after at least one year from the time SPD employees completed their training on the cameras. This should give SPD enough time to make more of its employees aware of the cameras, and to use the cameras to detect and investigate crimes.

Figure 7 SPD Surveillance Camera Program Training Memorandum

V. Perception of Safety

Ordinance 122705 governing the Surveillance Camera Pilot Program requires that the City Auditor conduct an evaluation of the program that includes, "A survey to determine whether the installation of surveillance cameras affects how the public perceives the safety of the parks in which the cameras are located."15

We conducted a face-to-face survey of persons in and just outside Cal Anderson Park in March 2009 and April 2009 to help determine whether our respondents' perception of safety changed due to the presence of the surveillance cameras. We administered the survey to 103 respondents. We present a summary of our survey analysis here. Our survey's methodology limitations prevented us from confidently generalizing its results beyond our pool of respondents. Nevertheless, the surveillance cameras in Cal Anderson Park appear to have had a minimal effect on our survey respondents' perception of safety in the park, which was dependent upon their awareness of the presence of the cameras. We found that less than one-third of the 103 respondents to the survey we conducted in Cal Anderson Park claimed to know about the cameras. Of those who knew about the cameras, only about four percent claimed that the cameras affected their perception of safety in the park. We were unable to determine whether this effect is sufficient for the purposes of the pilot program, due to the program's lack of clear goals, specific performance measures, and defined benchmarks to determine program success or failure.

See: (1) Appendix E for additional discussion of the survey protocol and the limitations to our survey methodology; (2) Appendix F for the survey introductory text and questions; (3) Appendix G for the frequencies of our survey responses; (4) Appendix H for responses to the open-ended question, "What are the areas of [Cal Anderson Park that you feel are safer compared to other areas of the park], and why are they safer?"; and (5) Appendix I for responses to the open-ended question, "Do you have any comments you would like to share with me about the presence of surveillance cameras in Cal Anderson Park?"

Analysis and Discussion of Survey Results

To determine, "whether the installation of surveillance cameras affects how the public perceives the safety of"16 Cal Anderson Park, we focused our analysis of the survey results on: (1) respondents' knowledge of the surveillance cameras in Cal Anderson Park; and (2) the difference in the respondents' perception of safety due to the presence of the cameras in the park.

Knowledge of the Cameras is Key to the Perception of Safety

We chose to focus on respondents' knowledge of the surveillance cameras in Cal Anderson Park in our attempt to determine the effect that the cameras have on their perception of park safety, because knowledge of the cameras is a vital part of the logic behind surveillance camera efficacy. Doctor Jerry Ratcliffe, PhD, wrote in a paper sponsored by the U.S. Department of Justice Office of Community Oriented Policing Services:

Numerous studies have tried to determine if the presence of cameras in public places reduces fear of crime in people who use the area... The findings are mixed but generally show there is some reduced level of fear of crime among people in CCTV areas, but only among people who were aware they were in an area under surveillance. Most studies exploring the perception of surveillance areas found that less than half the interviewees were aware they were in a CCTV area.17

Evidence suggests that even though implementers install a system, have a publicity campaign, and place signage, there is no guarantee the population will be aware of the cameras. In Glasgow, Scotland, 15 months after 32 cameras were installed in the city center, only 41 percent of those interviewed were aware of the cameras. These findings are similar to other research that found only one-third of respondents were aware they were within the vision of a public-street CCTV system.18

Our survey results appear to be similar to those published in recent literature pertaining to surveillance cameras. Fewer than one-third (26 percent) of our survey's respondents claimed to know about the presence of the cameras in Cal Anderson Park. Of those who did know about the cameras, the media was the most common means (50 percent) by which the respondents learned of the cameras' presence. The signage in the park appeared to have very little effect on informing our respondents about the cameras. The signs were removed from the park sometime after August 2008, and reposted between February19 and March 2009.20 This may help to explain why only 5 of the 103 respondents reported seeing a sign about the cameras. DPR representatives suspect vandals removed the signs from the park.

Recommendation 6: If the City wishes to increase public awareness of the surveillance cameras in City parks, then it should consider making the signs that are to inform the public of the cameras in the park more obvious to park users. The City should also consider periodically engaging in publicity campaigns that use different means of communicating to the public that there are surveillance cameras in the park (e.g., mass media and public meetings). These and similar methods of advertising should increase public awareness of the surveillance cameras in City parks, compared to the current level of awareness.

Perception of Safety Due to the Presence of Cameras

We designed our survey to directly measure the change in the respondents' perception of safety in Cal Anderson Park due to the presence of the surveillance cameras. After the survey provider informed the respondents about the cameras in Cal Anderson Park, between 22 and 25 percent of respondents reported an increase in their perception of safety due to the presence of the cameras. Though we also asked about nighttime safety, we focused our analysis on daytime safety in the park, because fewer than 17 percent of the respondents reported usually visiting the park after 6:00 PM.

Though some respondents reported an increase in their perception of daytime safety in the park due to the presence of the cameras, more than 70 percent who answered the question reported that the daytime safety in Cal Anderson Park is "about the same" due to the cameras. This result is similar to that of the following question from a survey performed by a Seattle University criminal justice student in October 2006 on Seattle Central Community College students: "Would the presence of camera surveillance make you feel safer in the park?" Of the 88 respondents who answered that question, 63 (nearly 72 percent) responded, "No."21

According to the authors of a CITRIS Report22 on a recent study of the City of San Francisco's Community Safety Camera (CSC) Program, "Video surveillance may increase the response of police officers toward areas where suspicious behavior is occurring. These officers could arrest offenders or deter potential offenders with their presence."23 Cal Anderson Park is located only one block from SPD's East Precinct. Park users might expect that if the park is under police surveillance, then police could respond quickly. However, Ordinance 122705 governing the use of the cameras limits SPD's ability to access the surveillance camera system:

A. Only SPD personnel are authorized to utilize the cameras for live monitoring, including for active monitoring:

1. When SPD has a reasonable suspicion to believe that any criminal activity, whether felony, gross misdemeanor, or misdemeanor, may be in progress within the area visible from the camera;

2. As part of an ongoing investigation into criminal activity that SPD has a reasonable suspicion to believe has occurred within the area visible from the camera; or

3. During a state of emergency declared by the Mayor as provided in applicable law.24

Live or active monitoring is not included in San Francisco's CSC Program. The authors of the CITRIS Report note that a significant issue with San Francisco's forensic approach to using their CSC system involves the lack of feedback cycle to those who:

SPD employees only monitor the surveillance cameras in real-time in exceptional circumstances. We asked our respondents "How safe would you feel Cal Anderson Park would be due to the presence of the cameras if no one was watching the camera footage in real time"? 97 percent of respondents reported they would not feel safer because of the cameras, knowing that no one was watching the camera footage. Roughly 15 percent of respondents reported feeling a decrease in their perception of safety in the park if no one is watching the surveillance cameras. About 82 percent of respondents reported "about the same" feeling of safety if no one is watching the cameras. Finally, about 3 percent of respondents reported feeling an increase in safety if no one is watching the camera footage. (See Appendix G, Table G26.)

Recommendation 7: If the City wishes to improve the public's perception of safety attributable to surveillance cameras in parks, then it should consider modifying the SMC to authorize SPD personnel to view the live surveillance footage more frequently. More frequent viewing should help enable SPD to efficiently deploy officers to the parks when criminal or suspicious activity is observed through the surveillance cameras. After personnel begin viewing the life surveillance footage more frequently, the City should consider informing the public that someone is viewing the camera footage in real-time and that SPD has the ability to more quickly respond to criminal activity within the surveillance cameras' view in the parks. These changes should both increase the public's perception of safety in the parks where surveillance cameras are located, and decrease the fear that is sometimes associated with the knowledge that no one is watching the cameras. However, increasing SPD's authority to access the surveillance camera system requires balancing public safety concerns against privacy concerns.

Survey Analysis Conclusion

We concluded that the surveillance cameras in Cal Anderson Park appear to have a minimal effect on the public's perception of safety in the park. Of the 98 respondents who answered questions about their awareness of the cameras and the cameras' effect on their perception of safety in Cal Anderson Park, only 4 percent claimed both knowledge of the cameras and that the cameras had an effect on their safety. In the absence of clear program goals, specific performance measures, and defined benchmarks to determine program success or failure, this percentage may be small enough to question the overall utility of the program, as the City currently executes it, to increase the public perception of safety in the park. On the other hand, anecdotal evidence from the respondents' comments about park safety and the effect of the cameras suggest other crime-deterrent and safety-enhancing methods contribute to an increased perception of safety. These methods may include: increased lighting, decreased visual obstructions, and increased appropriate activity in the parks. (See Appendices H and I for survey respondent comments).

Recommendation 8: If the City wishes to improve its ability to evaluate the Surveillance Camera Pilot Program, then it should consider establishing clear program goals, specific performance measures, and defined benchmarks to determine program success or failure. Definitions of success or failure should address expected or desired levels of change in the areas of crime deterrence, crime-solving, and perception of public safety. To do this:

(a) Council should consider amending the SMC to require the Executive to collect and regularly report relevant program data to measure progress on achieving program goals and performance relative to appropriate benchmarks; or

(b) The Executive, as a good management practice, should consider collecting and regularly reporting such data without being required to do so by an amendment to the SMC.

Recommendation 9: Ordinance 122705 requires that the City Council grant additional ordinance authority to continue operating surveillance cameras in Cal Anderson Park beyond 90 days after the City Auditor submits the program evaluation to the City Council.26 If the City believes public opinion about the surveillance cameras is an important factor in determining whether to grant this additional authority then it should consider:

(a) Authorizing and funding a more extensive survey (e.g., a random-digit dialing telephone survey) to evaluate the effect of the surveillance cameras on the public's perception of safety;

(b) Requiring DPR to hold one or more public hearings in Cal Anderson Park to obtain public comments about what would increase their perception of safety and about the presence of surveillance cameras in that park; and

(c) Using the public comments DPR will receive as it convenes community meetings in other parks per SMC 18.14.040 B to inform the decision about whether to install cameras in other parks.

(d)

Appendix A Ordinance 122705

City of Seattle Legislative Information Service

Information updated as of June 16, 2008 12:41 PM

Council Bill Number: 116225

Ordinance Number: 122705

AN ORDINANCE relating to surveillance cameras in Seattle parks: authorizing a pilot program to install surveillance cameras in selected parks; adopting City policies regarding the installation and use of such cameras; and establishing a new Chapter 18.14 in the Seattle Municipal Code.

Date introduced/referred: May 27, 2008

Date passed: June 9, 2008

Status: Passed

Vote: 8-0 (Excused: Conlin)

Committee: Parks and Seattle Center

Sponsor: RASMUSSEN

Index Terms: BUDGET, DEPARTMENT-OF-PARKS-AND-RECREATION, PERSONAL-SAFETY, CAPITOL-HILL, INTERNATIONAL-DISTRICT, PIONEER-SQUARE, PIKE-MARKET, PIKEPLACE-MARKET, SAFETY, CRIME-PREVENTION

Note: Center City Security Project, Funding for cameras in certain City parks

Text

AN ORDINANCE relating to surveillance cameras in Seattle parks: authorizing a pilot program to install surveillance cameras in selected parks; adopting City policies regarding the installation and use of such cameras; and establishing a new Chapter 18.14 in the Seattle Municipal Code.

BE IT ORDAINED BY THE CITY OF SEATTLE AS FOLLOWS:

Section 1. A new Chapter 18.14 of the Seattle Municipal Code is established as follows:

Chapter 18.14 Surveillance Cameras in Seattle Parks Pilot Program.

18.14.010 Statement of Purpose.

The purpose of this chapter is to authorize a pilot program regarding the installation and use of surveillance cameras in Seattle parks and to establish City policies regarding the installation and use of such cameras. The policies described in this chapter apply to cameras as defined in this chapter. Unless specified otherwise in a particular policy, the policies contained in this chapter apply to all persons employed by the City of Seattle, including agents retained on a temporary, contract, or voluntary basis.

18.14.020 Definitions.

A. "Cameras" means surveillance cameras: 1) installed in City parks as part of the pilot project authorized by this chapter; and 2) cameras installed in Cal Anderson Park beginning in January 2008 with funds from the Facility and Structure Maintenance BCL (K320A) of the Parks and Recreation Fund (10200). "Cameras" do not include surveillance cameras installed on City property other than in parks, cameras installed in police cars, cameras installed along pubic rights of way intended to record traffic violations, and cameras that are intended primarily to monitor the interiors or entrances of City buildings, including buildings located within City parks.

B. "Live monitoring" means a person viewing images live in real time as they are being captured and recorded by a camera.

C. "Active monitoring" means a person manipulating the point and zoom features of a camera in live monitoring mode in order to focus the camera on a particular person.

18.14.030 Pilot Program Authorization.

The Seattle Department of Parks and Recreation (DPR) and the City's Department of Information Technology (DoIT) are authorized to install and operate surveillance cameras in the following Seattle parks as part of a pilot program: Cal Anderson Park, Hing Hay Park, Occidental Park, and Victor Steinbrueck Park. Additional ordinance authority is required to install cameras in other City parks or to operate cameras in the parks identified above beyond ninety days after the program evaluation described in SMC Section 18.14.100 is submitted to the City Council.

18.14.040 City Policies Regarding the Installation and Use of Cameras.

A. Cameras are authorized to record video only and are not authorized to record audio.

B. In order to inform the community of the intended installation and to seek public comment, DPR shall convene, prior to installation, a community meeting for each park in which the installation of cameras is proposed under the pilot project.

C. Cameras should be installed to primarily record events that take place on public property. It is not a violation of this policy if cameras incidentally record events that occur on private property.

D. DPR shall prominently post at least one sign in each park where a camera is located informing the public about the presence of the camera(s).

E. The cameras and related equipment shall be the property of DoIT.

F. A monitor for the cameras and the controls that enable active monitoring shall be located in SPD's 911 Center.

G. Up to one additional monitor may be located at a DPR facility and at a DoIT facility.

18.14.050 City Policies Regarding Video Recordings Created by the Cameras.

A. The cameras may record images continuously twenty-four (24) hours per day, seven days per week.

B. Recordings shall be the property of DPR.

C. The administrative purposes served by recordings created by these cameras will typically be completed within fourteen (14) days. Cameras and equipment shall therefore normally be set to automatically record over previously recorded video recordings after a period of fourteen (14) days.

D. Authorized DPR and DoIT employees may under the following circumstances view video recordings, override or extend the automatic fourteen-day period, or retain a copy of a video recording: 1) in order to comply with a court order, the Washington Public Records Act, discovery requirements in a legal proceeding, or other applicable law; 2) as part of a criminal, civil, or administrative investigation; 3) to evaluate the video recording for possible use in a criminal, civil, or administrative legal proceeding in which the City is, or is reasonably expected to become, a party; and, 4) for system training, testing, maintenance, or repair.

E. A DPR or DoIT supervisor must authorize the viewing of video recordings by DPR or DoIT personnel.

18.14.060 Policies Regarding Live Monitoring of the Cameras.

A. Only SPD personnel are authorized to utilize the cameras for live monitoring, including for active monitoring:

1. When SPD has a reasonable suspicion to believe that any criminal activity, whether felony, gross misdemeanor, or misdemeanor, may be in progress within the area visible from the camera;

2. As part of an ongoing investigation into criminal activity that SPD has a reasonable suspicion to believe has occurred within the area visible from the camera; or

3. During a state of emergency declared by the Mayor as provided in applicable law.

B. SPD, DoIT, or City Auditor personnel are authorized to use the cameras for live monitoring, including for active monitoring:

1. For system training, testing, maintenance, or repair; or,

2. As part of the audit and program evaluation described in Section 18.14.100.

C. A SPD supervisor must authorize SPD personnel to utilize a camera for live monitoring, including for active monitoring.

18.14.070 Policy Prohibiting SPD Active Monitoring for An Improper Purpose.

SPD personnel shall not actively monitor a camera in order to focus the camera on a particular person based solely on the person's race, color, age, sex, marital status, sexual orientation, gender identity, political ideology, creed, religion, ancestry, national origin, or the presences of any physical disability. This section is not intended to limit: 1) incidental live monitoring of any person or object in view of the cameras when the camera is focused on a permitted subject; or 2) focusing active live monitoring on a possible suspect, witness, victim, or other person based upon the suspect, witness, victim, or other person's reported description or characteristics.

18.14.080 Policies Regarding Access to Recordings Created by the Cameras.

A. Except as permitted by SMC Section 18.14.050, only SPD personnel may view the video recordings created by the cameras:

1. When SPD has a reasonable suspicion to believe that any criminal activity, whether felony, gross misdemeanor, or misdemeanor, may be in progress within the area visible from the camera;

2. As part of an ongoing investigation into criminal activity that SPD has a reasonable suspicion to believe has occurred within the area visible from the camera; or

3. During a state of emergency declared by the Mayor as provided in applicable law.

C. SPD, DoIT, or City Auditor personnel may view the video recordings created by the cameras:

1. For system training, testing, maintenance, or repair; or,

2. As part of the audit and program evaluation described in Section 18.14.100.

C. A SPD supervisor must authorize the viewing of video recordings by SPD personnel.

18.14.090 Recordkeeping.

A. DPR and DoIT shall maintain a log recording the date, time, and duration of the video recording(s) reviewed; the personnel involved; and the reason(s) for viewing the video whenever department personnel view video recordings.

B. SPD shall maintain a log recording the date, time, and duration of the video viewed; the personnel involved; and the reason(s) for viewing the video whenever department personnel view video recordings or conduct live monitoring of a camera, including active monitoring.

C. The log books provided for in this section shall be retained by the City for ninety days after the report described in SMC Section 18.14.100 is submitted to the City Council.

18.14.100 Auditing and Evaluation of Surveillance Camera Pilot Program.

A. The City Auditor shall audit the DPR, DoIT, and SPD log books; corresponding 911 calls and other police contacts; and video recordings throughout the pilot program to aid in program evaluation and to assess compliance with this chapter.

B. The City Auditor shall also perform and provide to the City Council a program evaluation within eighteen (18) months of the completion of the installation of cameras pursuant to the pilot program.

C. The program evaluation provided for in this section shall include the following information:

1. Crime statistics and the number of calls for assistance in the monitored parks for the first 12 months after the cameras are installed;

2. Crime statistics and the number of calls for assistance in the monitored parks for the two years preceding the installation of the cameras;

3. Crime statistics and the number of calls for assistance for the areas just outside the monitored parks for the first 12 months after the cameras are installed;

4. Crime statistics and the number of calls for assistance for areas just outside the monitored parks for the two years preceding the installation of the cameras;

5. A description of crime trends over the past five years in the where the monitored parks are located;

6. The number of crimes detected in the monitored parks due to the presence of the cameras. Data should be provided for the first 12 months after the cameras are installed;

7. The number of crime investigations aided by the use of video recordings obtained by the cameras. A description of how the video recordings were helpful to each investigation should be included. Data should be provided for the first 12 months after the cameras are installed;

8. A survey to determine whether the installation of surveillance cameras affects how the public perceives the safety of the parks in which the cameras are located; and,

9. Additional information and analysis that the City Auditor deems useful.

D. The City Council will review the program evaluation to be completed by the City Auditor and consider any possible deterrence effects of the cameras on crime before authorizing continued operation of the cameras beyond the completion of the pilot program.

18.14.110 No Effect on Admissibility.

Neither compliance with nor a failure to comply with the policies contained in this chapter shall affect the admissibility of video recordings as evidence in criminal, civil, or administrative proceedings.

18.14.120 Civil Liability.

A. Subject to the limitations of this section, a person shall have a right of action against the City based on this chapter for damages up to Five Hundred Dollars ($500.00) proximately caused by a willful and deliberate violation of the provisions of SMC 18.14.070.

B. No cause of action may be based upon the activity of departmental personnel in complying with a court order, or an action authorized by this chapter.

C. The City reserves all defenses at law consistent with this chapter, including but not limited to consent, privilege, participation, and waiver, and as to departmental personnel or a City official, any defense arising in the employer/employee or principal/agent relationship.

D. No cause of action may be based upon this chapter against the Mayor, the City Council, any City department head, any departmental personnel, or any other City officer or employee, individually, for any action or omission made in good faith in the scope and course of his or her duties. In the event such a lawsuit is brought against a City officer or employee, individually, for such an action or omission, and the officer or employee cooperates fully in defense of the lawsuit, the City Attorney may represent the individual and defend the litigation. If the claim is deemed a proper one or judgment is rendered against the City officer or employee individually, the judgment shall be paid by the City in accordance with its procedures for the settlement of claims and payment of judgments.

18.14.130 Employee Discipline.

Any City personnel who violates policies contained in this chapter, or any implementing rule or regulation, may be subject to the disciplinary proceedings and punishment authorized by the City Charter, Article XVI.

For City personnel who are represented under the terms of a collective bargaining agreement, this section prevails except where it conflicts with the collective bargaining agreement, any memoranda of agreement or understanding signed pursuant to the collective bargaining agreement, or any recognized and established practice relative to the members of the bargaining unit.

Section 2. DPR, DoIT, SPD, and the City Clerk shall take the steps necessary to implement this ordinance.

Section 3. This ordinance shall take effect and be in force thirty (30) days from and after its approval by the Mayor, but if not approved and returned by the Mayor within ten (10) days after presentation, it shall take effect as provided by Municipal Code Section 1.04.020.

Passed by the City Council the ____ day of ________________________,

2008, and signed by me in open session in authentication of its

passage this _____ day of ___________________, 2008.

_________________________________

President __________of the City Council

Approved by me this ____ day of _____________________, 2008.

_________________________________

Gregory J. Nickels, Mayor

Filed by me this ____ day of __________________________, 2008.

____________________________________

City Clerk

May 30, 2008

Version #5

t

Fiscal Note

Appendix B Departmental Compliance with Seattle Municipal Code Requirements

We assessed departmental compliance with policies stated in ordinance 122705 regarding: (1) the installation and use of the cameras; (2) video recordings created by the cameras; (3) live monitoring of the cameras; (4) prohibiting SPD active monitoring for an improper purpose; (5) access to recordings created by the cameras; (6) recordkeeping; and (7) auditing and evaluation of the pilot program. Tables B1 through B7 report departmental compliance with a specific policy by using the color scheme: fullcompliance (green), partial compliance or area of concern (yellow), noncompliance (red).

Appendix C Quasi-Experimental Evaluation Methods

Ordinance 122705 requires that the City Auditor include crime statistics and the number of calls for assistance (i.e., 911 calls and officer backup calls) in and just outside Cal Anderson Park in our program evaluation.27 However, we were unable to confidently provide these statistics and information about the number of calls for assistance due to data limitations. We explain this conclusion by discussing: (1) the nature and usefulness of crime data that SPD collects; and (2) quasi-experimental methods to determine from crime statistics the effect the surveillance cameras may have on crime. We address the first issue in section IV of this report. This appendix addresses the second issue: obstacles to using quasi-experimental methods to derive conclusions from the City's crime statistics.

Study by the Center for Information Technology Research in the Interest of Society (CITRUS)

We conducted a literature review to discover what methods other program evaluations have used to assess the effectiveness of public surveillance cameras in deterring crime. We learned that the Center for Information Technology Research in the Interest of Society (CITRIS), at the University of California, Berkley, published a comprehensive evaluation28 in December 2008 on the effectiveness of a public surveillance camera program in San Francisco. The program evaluation consisted of a multi-disciplinary approach that included a quasi-experimental statistical evaluation of crime reports. The surveillance cameras were located almost exclusively on street corners. The address coordinate system at these locations allowed the authors to compare criminal incidents at varying distances from the cameras. The authors used three evaluation strategies to statistically assess the effectiveness of the surveillance cameras:

* Evaluation Strategy 1 compared crime rates at different distances from the cameras. The authors presumed that that the surveillance cameras would deter crime within, but not beyond, their view range, as supported by existing theoretical literature.29 After performing site visits, they concluded that the cameras were not visible at distances greater than 100 feet. By comparing crimes in areas defined by five 100-foot concentric circles around the cameras, they were able to determine whether crime declined within area one (i.e., 0-100 feet of the cameras) relative to areas two through five (i.e., 100-200-ft, 200-300-ft, 300-400-ft, and 400500-ft). When the authors observed such a decline, they considered it evidence that something impacted crime in area one, "beyond all of the other determinants of criminal activity in the larger area surrounding the camera location."30 However, this strategy was unable to address the possible displacement effect of the cameras, and did not consider changes in areaor city-wide crime over time.

* Evaluation Strategy 2 compared differences in crime rates between different locations, before and after the cameras were installed. This second strategy compensates for the weaknesses in the first strategy by comparing crime rates in areas one through five both preand postinstallation of the cameras across nineteen comparison sites in a difference-in-difference (DD) method.31 A DD decline in crime in area one suggests an actual local deterrence effect. If one or more of the DDs for areas two through five were greater than zero, that suggested that crime was being displaced from area one into the surrounding areas. The authors applied additional statistical tools to address the difficulty in matching and comparing non-identical comparison and treatment areas.32

* Evaluation Strategy 3 strengthened the first two strategies by comparing patterns of crime occurring in public versus private places. The authors repeated the first two methods after stratifying the cameras into public and private places to assess whether the cameras affect these two areas differently. This provided, "a basic falsification check of the results from the first two strategies."33

Quasi-Experimental Methodologies Could Not Be Used In Our Program Evaluation

While these quasi-experimental methodologies may be ideally designed to help in our own assessment of the surveillance cameras' effectiveness in deterring crime in Cal Anderson Park, we could not use them.

The first evaluation strategy was not possible because of the geographical limitations of the crime data. Our data needed to be specific enough to measure differences in crime rates between the areas in view of the cameras, and out of view of the cameras. Even if we assumed that all crime records with locations labeled "Cal Anderson Pk," refer to incidences that actually occurred in the park, we could not determine whether a crime occurred within views of the cameras. There are many areas within Cal Anderson Park, where the view of the cameras are obscured by some element of the landscape most notably, the semicircle of trees that stand north of the cameras (see Figure 3). These trees block the view of the cameras in more than half the area of Cal Anderson Park. This prevents us from separating crimes with locations labeled "Cal Anderson Pk" into groups that occurred within and out of view of the cameras.

The second evaluation strategy was not possible because we lacked one or more comparison parks. The CITRIS evaluators chose comparison locations where the cameras either were already installed or were going to be installed. Cal Anderson Park is the only Seattle city park that has surveillance cameras within it. Though the pilot program intended to install surveillance cameras in Hing Hay, Occidental, and Victor Steinbrueck Parks, budget constraints have prevented their installation. These parks are also poor matches with Cal Anderson Park. During our meeting with representatives from the SPD on February 4, 2009, we learned that Cal Anderson Park may not be compared to any other park, due to its many unique properties (e.g., size, location, recent renovation, local demographics, etc.). The discussion caused us not to seek a comparison park.

The third evaluation strategy was not possible because the first two evaluation strategies were not possible. Though we considered using other methodologies, none could adequately determine the cameras' effectiveness.

Appendix D Crime Trends in the Cal Anderson Park Neighborhood

Ordinance 122705 requires that we describe crime trends, "in the neighborhoods where the monitored parks are located" (SMC 18.14.100.C.5). The City Attorney's Office reported that there is no legal definition of "neighborhood" in Seattle. We decided to define the neighborhood around Cal Anderson Park to consist of the three census tracts that surround it. We analyzed statistics and trend data for the Cal Anderson Park neighborhood, aggregating census tracts 074, 075 (contains Cal Anderson Park), and 084. Ordinance 122705 also does not specify what crime trends we need to describe. After viewing Computer Aided Dispatch (CAD) data, we chose to present crime trends as defined by individual crime incident TYPE.

We did not consider all of the incident types, and purposely excluded some that could not occur in the Cal Anderson Park neighborhood (e.g., "harbor water emergency"). In addition, we only selected for a single spelling of the CAD data types in the CAD data to avoid misinterpreting the data. This appendix contains charts showing incident type trends for the Cal Anderson Park neighborhood from the beginning of February 2003 through the end of February 2009. We provide written descriptions of these trends in section IV, Table 2, in this report. SPD informed us that, for the period August 1, 2008 through February 28, 2009, about seven days' worth of data were missing due to technical issues that occurred in November concerning the transfer of data in the SPD Communications Center. The missing data include both dispatched calls and onview (officer-initiated) events. It is unlikely that this omission affects the presentation of the crime trends significantly.

Figure D1 Abduction Trends in Cal Anderson Neighborhood

Figure D2 Arrest Trends in Cal Anderson Park Neighborhood

Figure D3 Arson Trends in Cal Anderson Park Neighborhood

Figure D4 Assault Trends in Cal Anderson Park Neighborhood

Figure D5 Property Damage Trends in Cal Anderson Park Neighborhood

Figure D6 Disturbance Trends in Cal Anderson Park Neighborhood

Figure D7 Harassment Trends in Cal Anderson Park Neighborhood

Figure D8 Miscellaneous Trends in Cal Anderson Park Neighborhood

Figure D9 Narcotics Trends in Cal Anderson Park Neighborhood Figure D10 Nuisance Trends in Cal Anderson Park Neighborhood Figure D11 Prowler Trends in Cal Anderson Park Neighborhood Figure D12 Purse Snatching Trends in Cal Anderson Park Neighborhood

Figure D13 Rape Trends in Cal Anderson Park Neighborhood Figure D14 Robbery Trends in Cal Anderson Park Neighborhood

Figure D15 Sex Offense Trends in Cal Anderson Park Neighborhood

Figure D16 Gunshots Trends in Cal Anderson Park Neighborhood

Figure D17 Suicides and Attempts Trends in Cal Anderson Park Neighborhood

Figure D18 Suspicious Trends in Cal Anderson Park Neighborhood

Figure D19 Theft Trends in Cal Anderson Park Neighborhood Figure D20 Prostitution Trends in Cal Anderson Park Neighborhood

Figure D21 Person with Weapon Trends in Cal Anderson Park Neighborhood

Appendix E Survey Overview and Analysis Ordinance 122705 governing the Surveillance Camera Pilot Program requires that the City Auditor conduct an evaluation of the program that includes, "A survey to determine whether the installation of surveillance cameras affects how the public perceives the safety of the parks in which the cameras are located."34

We conducted a face-to-face survey of persons in and just outside Cal Anderson Park in March and April 2009 to help determine whether the public perception of safety changed due to the presence of the surveillance cameras. We administered the survey to 103 respondents. This appendix includes a discussion of the survey protocol and the limitations to our survey method.

Survey Protocol

We administered our survey at a variety of times and locations to maximize the likelihood that our survey results would represent the population of Cal Anderson Park users (see Table E1).

The dates on which we administered the survey were dependent upon our schedules35. However, we were able to choose two weekdays and one Saturday to obtain proportional responses from both weekdayand weekend-visitors to Cal Anderson Park. We chose four, three-hour blocks of time each day to survey people in and just outside the park. These blocks of time helped us obtain responses from different park users in the morning, early afternoon, late afternoon, and evening. Finally, we chose three locations36 to administer the surveys: (1) at the southeast corner of the park, at the intersection of 11th Ave. and E Pine St., next to a Metro bus stop; (2) between the restrooms and the Shelter House, on the walkway where both the surveillance cameras, and signs informing the presence of surveillance cameras, are located; and (3) by the fountain, in the northern part of the park (see Figure E1).

We adhered to the following rules to minimize selection bias when administering the survey and in response to practical needs:

* The survey provider37 will administer the survey as defined in Table E1.

* While in those locations, the survey provider will attempt to engage every other individual that walks by or near the provider, but not when the provider:

o Is engaged in administering a survey to a passerby;

o Sees that the individual is talking on a cell phone or running;

o Recalls having already engaged that individual; or

o Is surrounded by a crowd or group.

* In the event of a crowd or group, the provider will attempt to find another position in the assigned area to enable the provider to more easily engage every other passerby.

* If a passerby agrees to listen to the survey provider, the provider will read aloud the introductory text (included in Appendix F) and proceed or not proceed with the survey, in accordance with the respondent's wishes.

* The survey provider will record all responses and non-responses.

* If the respondent does not understand the question, the provider will provide a brief explanation.

The survey provider will write the answer the respondent provides to the question when the respondent gives it, whether or not the provider has finished reading all possible answers to the question.

Figure E1 Cal Anderson Park

Survey Method Limitations

In this section, we organize our discussion of the survey method limitations in terms of the four kinds of error that occur in surveys and that affect accuracy: (1) coverage error; (2) sampling error; (3) measurement error; and (4) nonresponse error. We briefly define each, and then discuss how our survey is subject to that kind of error.

Coverage Error

Coverage error is, "a discrepancy between the target population and the survey population."38 According to Ordinance 122705, "the public"39 is our target population. For the purposes of this survey, we specifically define the public to be those individuals who visit Cal Anderson Park. This includes a large number of individuals who visit any part of Cal Anderson Park at any date and time.

Our survey population consists of only those individuals present in the park on the three days when we conducted the survey. Our survey method excludes those users who were not present at the times we administered the survey (e.g., summer users). Our survey method also excludes those individuals who were not present at the locations where we administered the survey. We saw other individuals throughout the park, whom we did not contact because of the constraints of the survey methodology. We chose to administer the survey at the three locations in Table E1 because each experienced much foot-traffic.

We did not administer the surveys in other park locations because we lacked the necessary time. Among the areas we did not administer the survey, wherein other park users with potentially different perceptions of public safety visited, were: ball fields, tennis courts, picnic benches, and the playground. Consequentially, we cannot confidently generalize our findings beyond the survey population.

Sampling Error

Sampling error occurs when, "only a sample of the population of interest is surveyed."40

We surveyed only a sample of those individuals present at the times and locations listed in Table E1. To remove selection bias, we attempted to survey every-other-person who passed by the survey provider. We also did not attempt to survey other individuals for reasons of feasibility. This resulted in an even smaller sample of the population of interest. It is not possible to determine the percentage of users we surveyed, because no data exists on how many people use Cal Anderson Park at any given time41. The sample size we collected may not be large enough to help answer the question regarding changes in the perception of safety in Cal Anderson Park by those who visit the park.

Measurement Error

Measurement error occurs when, "a respondent's answer to a given question is inaccurate, imprecise, or cannot be compared in any useful way to other respondents' answers."42

Most of the survey questions required the respondents to measure their perceptions of safety on the survey date, or to remember something from more than a year before the survey date. If respondents felt an incentive to provide an answer quickly to an abstract question about how they perceive safety, then their assessment of personal safety may not be as accurate or precise as it could have been if they had more time to ponder their response. If the respondents did not remember the past correctly, then responses about earlier events and conditions might not be totally accurate or precise.

We vetted the questions before administering the survey to ensure that we wrote them in a clear, specific, and informative manner. However, the wording of some of the survey questions may still have been interpreted differently by different respondents. For example, the question: "Compared to how it was before February 2008, how safe do you feel Cal Anderson Park is now?" may have created variation in the respondents' interpretation of the period described as "before February 2008." Some respondents asked whether that was the time the City renovated the park. The City completed the renovation of Cal Anderson Park in September 2005.43 Those that thought the question referred to conditions long before February 2008 provided answers that may not be easily compared to the answers other respondents gave, if the other respondents thought the question referred to conditions immediately before February 2008.

Variation in the face-to-face administration of the survey may have produced different results with different respondents. Though the survey provider read each question to the respondents, it remained difficult to consistently control for variations in vocal inflection and intonation. By attempting to interact with the respondents (i.e., to make the respondents feel comfortable, to encourage them to participate and provide answers, or to adjust the speed and manner of the survey administration in response to the preferences of the respondents, as the survey provider perceived them) the survey provider may have produced variations in the answers of the respondents.

Nonresponse Error

Nonresponse error occurs when, "a significant number of people in the survey sample do not respond to the questionnaire and are different from those who do [respond] in a way that is important to the study."44

We attempted to survey 377 people, but successfully surveyed only 103 people. The overall nonresponse rate to our survey was 73 percent. The nonresponse rates differed greatly at each location:

* Nonresponse rate at the southeast corner = 81%

* Nonresponse rate between the restrooms and Shelter House = 69%

* Nonresponse rate by the fountain = 52%.

The survey methodology is likely responsible for the high nonresponse rate. To avoid selection bias, the survey provider randomly contacted only those individuals who walked through the areas defined in Table E1. However, that they were walking suggests that they preferred to continue to their destination, rather than stop and take a survey that would last several minutes.

This may help to explain the differences in nonresponse rates between the three locations. The survey provider stood at the southeast corner of the park, on the sidewalk along E. Pike St (see location 3 on Figure E1). Those individuals walking along the sidewalk were most likely going somewhere, and had no intention of stopping. Those walking near the bathrooms and Shelter House were more likely to be visiting the park, and not just passing through it. Those walking near the fountain were even more likely to be visiting the park, and would likely be more willing to stop and take a survey.

The high nonresponse rate may not contribute to nonresponse error, if those who declined do not differ in their perceptions of safety from those who responded. We think it reasonable to assume that the difference between respondents and nonrespondents may be negligible in terms of their perception of the safety of Cal Anderson Park. However, it is not possible for us to verify this assumption.

Appendix F Survey Questions

"Hello, may name is Dean Runolfson. I am a graduate student at the University of Washington and am working as an intern for the City of Seattle. I am conducting a survey for the Office of City Auditor about how people feel about Cal Anderson Park (this park). All of your responses will be anonymous. You may choose not to answer a question if you wish. If any part of the survey is unclear, please ask me to clarify. Please give your best guess to the answers about which you are not certain. This survey usually takes about 6 minutes to complete. Will you please take the survey?" (If accepts: "Thank you!") (If declines: "Okay. Have a nice day!")

Respondent ID number: __________ Location: ____________________ Date: ______________

Start time: _______________ Weather: _____________________________________________

Notes about respondent: __________________________________________________________

Q1: Have you ever visited Cal Anderson Park?

Universe: all persons

Values:

0. No (skip to Q1e)

1. Yes

Q1a: During the last year, how frequently have you visited Cal Anderson Park?

Universe: Q1 = 1

Values:

1. At least once per day

2. At least once per week

3. At least once per month

4. At least once every three months

5. At least once every six months

6. At least once per year

7. Less than once per year

Q1b: What time of day do you usually visit Cal Anderson Park?

Universe: Q1 = 1

Values:

1. 6:00 9:00 AM

2. 9:00 AM 12:00 PM

3. 12:00 3:00 PM

4. 3:00 6:00 PM

5. 6:00 9:00 PM

6. 9:00 PM 12:00 AM

7. 12:00 6:00 AM

Q1c: Did you visit Cal Anderson Park before February 2008?

Universe: Q1 = 1

Values:

0. No

1. Yes

Q1d: Have you visited Cal Anderson Park in or after February 2008?

Universe: Q1 = 1

Values:

0. No

1. Yes

Q1e: Do you live, work, or attend school within a 10-block radius of Cal Anderson Park?

Universe: all persons

Values:

0. No

1. Yes

Q1f: Do you regularly participate in organized activities in Cal Anderson Park, such as with sports teams, chess clubs, etc?

Universe: Q1a = 1, 2, 3, or 4 (not 5, 6, or 7)

Values:

0. No

1. Yes

Q2: Overall, how safe do you feel Cal Anderson Park is during the day?

Universe: all persons

Values:

1. Very unsafe

2. Unsafe

3. Somewhat unsafe

4. About average

5. Somewhat safe

6. Safe

7. Very safe

Q2a: Overall, how safe do you feel Cal Anderson Park is at night?

Universe: all persons

Values:

1. Very unsafe

2. Unsafe

3. Somewhat unsafe

4. About average

5. Somewhat safe

6. Safe

7. Very safe

Q2b: Are there areas of Cal Anderson Park that you feel are safer compared to other areas of the park?

Universe: all persons

Values:

0. No (go to Q2d)

1. Yes

Q2c: What are the areas of the park, and why are they safer?

Universe: Q2b = 1

Values:

Response = Open-ended question

Q2d: Compared to how it was before February 2008, how safe do you feel Cal Anderson Park is now?

Universe: Q1c = 1

Values:

1. Much more dangerous

2. More dangerous

3. Somewhat more dangerous

4. About the same

5. Somewhat safer

6. Safer

7. Much safer

Q2e: Did you ever see any illegal activity, such as drug dealing, drug using, assaults, sexual activity, etc., in Cal Anderson Park before February 2008?

Universe: Q1c = 1

Values:

0. No

1. Yes

Q2f: Did you ever see any illegal activity, such as drug dealing, drug using, assaults, sexual activity, etc., in Cal Anderson Park in or after February 2008?

Universe: Q1d = 1

Values:

0. No

1. Yes

Q3: The City installed surveillance cameras in Cal Anderson Park in February 2008. Before I told you this, were you aware of the presence of the surveillance cameras in Cal Anderson Park?

Universe: all persons

Values:

0. No (skip to Q4)

1. Yes

Q3a: In what month and year did you become aware of the presence of the surveillance cameras?

Universe: Q3 = 1

Values:

1. Jan 2008

2. Feb 2008

3. Mar 2008

4. Apr 2008

5. May 2008

6. Jun 2008

7. Jul 2008

8. Aug 2008

9. Sep 2008

10. Oct 2008

11. Nov 2008

12. Dec 2008

13. Jan 2009

14. Feb 2009

15. Mar 2009

Q3b: How did you learn about the presence of the surveillance cameras in Cal Anderson Park?

Universe: Q3 = 1

Values:

1. See the cameras

2. See a sign

3. Hear it by word of mouth

4. Learn from the media

5. Some other way

Q3c: Have you ever seen a sign in Cal Anderson Park that notified the public that there are surveillance cameras in the park?

Universe: Q3 = 1

Values:

0. No

1. Yes

Q4: Compared to how safe you felt Cal Anderson Park was before you became aware of the surveillance cameras, overall, how safe do you feel Cal Anderson Park is during the day?

Universe: all persons

Values:

1. Much less safe

2. Less safe

3. Somewhat less safe

4. About the same

5. Somewhat safer

6. Safer

7. Much safer

Q4a: Compared to how safe you felt Cal Anderson Park was before you became aware of the surveillance cameras, overall, how safe do you feel Cal Anderson Park is at night?

Universe: all persons

Values:

1. Much less safe

2. Less safe

3. Somewhat less safe

4. About the same

5. Somewhat safer

6. Safer

7. Much safer

Q4b: How safe would you feel Cal Anderson Park would be due to the presence of the cameras if no one was watching the camera footage in real time?

Universe: all persons

Values:

1. Much less safe

2. Less safe

3. Somewhat less safe

4. About the same

5. Somewhat safer

6. Safer

7. Much safer

Q5: What is your age?

Universe: all persons

Values:

Range: 15-120

Q6: What is your gender?

Universe: all persons

Values:

1. Male

2. Female

3. Transgender

Q7: This is not a perfect list, but what racial group best describes you? (show list)

Universe: all persons

Values:

1. White

2. Black

3. American Indian/Alaskan Native

4. Native Hawaiian/Other Pacific Islander

5. Asian

6. Hispanic

7. Other

Q8: What is the highest level of school you have completed or the highest degree you have received? (show list)

Universe: all persons

Values:

1. Less than 9th grade

2. 9th grade 12th grade (no high school diploma)

3. High school grad (with diploma)

4. GED

5. Vocational certificate

6. Some college, no degree

7. Associate degree in college

8. Bachelor's degree

9. Master's degree

10. Professional school degree

11. Doctorate degree

Q9: What is your living status?

Universe: all persons

Values:

1. Alone

2. Married/cohabiting

3. With roommate(s)

Q10: If you have any children, are you living with them?

Universe: all persons

Values:

0. No

1. Yes

Q11: Do you have any comments you would like to share with me about the presence of surveillance cameras in Cal Anderson Park?

Universe: all persons

Values:

Response = Open-ended question

Thank you very much for your time and your participation.

End time: __________

Appendix G Survey Question Frequency Tables

This appendix includes the frequencies of the responses to each of the survey questions except for the open-ended questions, which are included in Appendices H and I. We administered the survey to 103 respondents.

NOTE: values are subject to rounding error.

Appendix H Responses to Question 2c Regarding Safer Areas in Cal Anderson Park

This appendix includes survey responses to the open-ended question, "What are the areas of [Cal Anderson Park that you feel are safer compared to other areas of the park], and why are they safer?"

Appendix I Responses to Question 11 Regarding Surveillance Cameras in Cal Anderson Park

This appendix includes survey responses to the open-ended question, "Do you have any comments you would like to share with me about the presence of surveillance cameras in Cal Anderson Park?"

Appendix J Executive Departments' Comments on This Report

M E M O R A N D U M

DATE: October 5, 2009

TO: Office of the City Auditor

FROM: Greg Doss, Department of Finance

SUBJECT: Executive Response to the Surveillance Camera Audit

The Executive appreciates the opportunity to comment on the Audit draft. We have been working with the Seattle Police Department (SPD), the Department of Information Technology (DoIT) and the Department of Parks and Recreation (DPR) to ensure that the Department Compliance recommendations are fully implemented. As the Audit notes, most of these changes have been made.

A centerpiece of the evaluation of the Cal Anderson Park cameras is a survey conducted on three days in March and April of 2009, of passers-by in and just outside the Park. The purpose of the survey was to ascertain the public's knowledge of the surveillance cameras and their resultant perceptions of safety in the Park.

Earlier in the year when the Office of the City Auditor (OCA) proposed to do a community survey, SPD staff recommended a telephone survey of both neighborhood residents and Park users. The Police Department has considerable experience in conducting community surveys on perceptions of public safety and has found this methodology to be both valid and reliable. The OCA staff chose not to follow this approach.

Because of the methodology used for the survey in this Audit, we believe that its results should be interpreted with caution. The survey respondents are at best a very limited representation of "the public" since neighborhood residents and active athletic populations are likely excluded. Second, there was a very high non-response rate to the survey. Nearly four times as many people chose not to participate and there is no way to know how these persons' views compare with those who chose to respond. For this reason, it may not be possible to generalize the survey respondents' views to those of other passers-by, much less to "the public." Third, it is likely that the time of year and inclement weather at the time of the survey produced an unrepresentative subset of Park users. Finally, there is a potential mismatch between those persons most likely impacted by Park safety and those surveyed. For example, about a third of the respondents reported visiting the Park less than once a month, nearly 40 percent neither live, work or attend school within the area and 90 percent do not regularly participate in Park activities. Arguably, these persons are not nearly as likely to be impacted by Park safety as are those who visit the Park more frequently, those who live, work or go to school in the area, or those who regularly participate in activities there.

While we are concerned with the survey's methodology, our most significant issue is the use of a public perception survey as a central means to evaluate the effectiveness of the cameras. The City Council requested information on a total of nine measures, including many crime statistics and overall crime trends. In determining public safety results, it is more important to measure criminal activity in the park rather than the public's perception of the cameras. We acknowledge the shortcoming of SPD's data to pin-point crimes reported in the Park, but we do not believe that the Audit's focus on the survey results is an adequate proxy for measuring public safety results. The Audit's discussions about crime in the area surrounding the Park and the reduction in park exclusions are given less weight than the survey as measures of public safety.

The OCA makes a link between awareness of public safety cameras and criminal behavior. While we do not deny that deterrence can only happen when criminals know that they are being filmed, we question if this particular survey was an effective instrument to reach the criminal element. We do not believe that daytime park users are in any way representative of those who are most likely to endanger public safety.

Again, we appreciate the opportunity to comment and are thankful that the CAO engaged Executive staff during the audit review and development process.

Final Comments from the Department of Information Technology (DoIT)

From: Schmidt, Mark

To: Denzel. Mary

CC: Arnold. Dean, Schrier. Bill, Doss. Greg

Date: 10/2/2009 4:07 PM

Subject: FW: Surveillance Cameras Audit Final Review Draft for comment

Mary,

After additional review of the final draft, DoIT would like to include a few comments. My apologies for any confusion.

Background:

To test, maintain, and repair components of the surveillance camera system, DoIT uses a configuration tool that does not provide surveillance video features (full access to video or PTZ camera control), although it does provide very limited access to live video (to enable configuration of cameras).

Surveillance video features are provided by separate video review workstations owned by SPD, DPR, and DoIT.

DoIT has fully complied with the ordinance and has updated its log book with a new record each time the video system has been accessed, whether that access was via the configuration tool or via the video review workstation owned by DoIT.

Within the report, OCA makes the following recommendation:

"Recommendation 1 To aid program oversight, The Department of Parks and Recreation (DPR), the Department of Information Technology (DoIT), and the Seattle Police Department (SPD), should amend their log books to include a place where the user may record the name, title, and signature of the authorizing supervisor as one of the personnel whose involvement is required for using the cameras or viewing camera footage. According to DPR they have made changes to comply with this recommendation. Also, the City should consider clarifying SMC 18.14.090 to indicate that the authorizing supervisor is among the "personnel involved," whose name and title must appear in the log book entries."

DoIT comments:

DoIT believes that the requirements in SMC 18.14.090 Recordkeeping are impractical when applied to the use of configuration tools for system testing, maintenance and repair. Implementing Recommendation 1 (requiring management signatures every time the system is accessed) would also be impractical in this scenario, and could inhibit DoIT's ability to provide timely service and support.

DoIT fully understands the recordkeeping requirement when applied to the use of DoIT's video review workstation for training or viewing surveillance video.

Mark Schmidt

Communications Technologies Engineer

Department of Information Technology

City of Seattle

206-233-7878

mark.schmidt@seattle.gov

Appendix K Auditor's Comments on the Executive's Comments

Cal Anderson Park Surveillance Camera Pilot Program Evaluation

We appreciate the Executive branch departments' comments (DoIT, DPR, Finance, SPD) on early drafts of this report. The following are our responses to the Executive's October 5, 2009 comments (see Appendix J) on our final draft report.

Executive Comment on Survey Methodology

The Executive stated:

Earlier in the year when the Office of the City Auditor (OCA) proposed to do a community survey, SPD staff recommended a telephone survey of both neighborhood residents and Park users. The Police Department has considerable experience in conducting community surveys on perceptions of public safety and has found this methodology to be both valid and reliable. The OCA staff chose not to follow this approach.

Because of the methodology used for the survey in this Audit, we believe that its results should be interpreted with caution.

Office of City Auditor Response:

The City Auditor determined that there was insufficient budget to fund a valid and reliable phone survey, opting instead for an in-person survey capturing the views of willing participants who were passing through the park on three days in the spring of 2009. We agree that the survey's methodology limitations prevent generalizing from the results. In response to this comment we have been careful to characterize the results only in terms of the survey's respondents, see page 16 of the report:

Our survey's methodology limitations prevented us from confidently generalizing its results beyond our pool of respondents. Nevertheless, the surveillance cameras in Cal Anderson Park appear to have had a minimal effect on our survey respondents' perception of safety in the park, which was dependent upon their awareness of the presence of the cameras.

Executive comment on park users excluded from survey respondents

The Executive stated: "The survey respondents are at best a very limited representation of 'the public' since neighborhood residents and active athletic populations are likely excluded."

Office of City Auditor Response: 64 percent of respondents reported that they live, work, or attend school within ten blocks of Cal Anderson Park, and 57 percent reported visiting the park at least weekly. The Executive accurately noted that only ten percent of respondents reported they "regularly participate in organized activities in Cal Anderson Park, such as with sports teams, chess clubs, etc." Because the City does not gather data on people who use its parks, there is no way to determine whether the respondents to our survey are representative of typical park users.

Centrality of Survey Results to the Report

The Executive states: "While we are concerned with the survey's methodology, our most significant issue is the use of a public perception survey as a central means to evaluate the effectiveness of the cameras."

Office of City Auditor Response

The purpose of this work was not to evaluate the effectiveness of the cameras. Ordinance 27705 required the Office of City Auditor to conduct a program evaluation so that the City Council could use this information to determine whether to expand the camera program. The report addresses four measures, as required by the ordinance: 1) department compliance with the pilot program protocols (see pages 5-7 and Appendix B in the report), 2) a survey of the cameras' effect on the public's perception of safety (pages 16-20 and Appendices E through I), 3) analysis of crime statistics before and after installation of the cameras (pages 10-14 and Appendices C and D), and 4) determining the number of times the cameras were used to detect or investigate crimes (pages 14-15). No one measure is described in the report as being more central than the others.

Camera Effect on Perception of Public Safety

Executive statement:

The OCA makes a link between awareness of public safety cameras and criminal behavior. While we do not deny that deterrence can only happen when criminals know that they are being filmed, we question if this particular survey was an effective instrument to reach the criminal element.

Office of City Auditor Response:

The survey was not designed or intended to reach criminals. We quote research reports that indicate cameras cannot have any effect on perceptions of safety or crime deterrence if no one knows they are there. We recommend additional publicity about the cameras, which could affect both ordinary park users and potential criminals.

1 Mayor Greg Nickels, 2008 Budget Speech, Making History Today, September 17, 2007. http://www.seattle.gov/mayor/speeches/budgetSpeech2008.htm (accessed: June 3, 2009).

2 See Appendix A for the text of Ordinance 122705

3 See Appendix A, Ordinance 122705

4 See Appendix A, Ordinance 122705, SMC 18.14.100.A

5 According to Ordinance 122705, SMC 18.14.020, live monitoring means, "a person viewing images live in real time as they are being captured and recorded by a camera." Active monitoring means, "a person manipulating the point and zoom features of a camera in live monitoring mode in order to focus the camera on a particular person."

6 See Appendix A, SMC 18.14.050.E, 18.14.060.C, and 18.14.080.C

7 See Appendix A, Section 2, emphasis added

8 See Appendix A, SMC 18.14.100.C

9 Computer Aided Dispatch (CAD) data include all dispatched 911 calls and police officer on-views (i.e., officer-initiated events).

10 A nonprofit community organization that is, "dedicated to generating creative, innovative activities, programs and events to activate Cal Anderson Park." Cal Anderson Park Alliance. Home Page. http://www.calandersonpark.org/ (accessed: May 12, 2009).

11 See Appendix A, SMC 18.14.100.C.5

12 Police beats are divisions of police precincts, which divide the City into jurisdictional areas.

13 Adapted from: http://www.seattle.gov/dpd/cms/groups/pan/@pan/documents/web_informational /dpds_006734.pdf

14 DPR reported on August 5, 2009, that it had begun work on an improved recordkeeping system, and that it will work with SPD on record retention issues.

15 See Appendix A, SMC 18.14.100.C.8

16 See Appendix A, SMC 18.14.100.C.8

17 Ratcliffe, Jerry. Video Surveillance of Public Places. Problem-Oriented Guides for Police, No. 4. U.S. Department of Justice Office of Community Oriented Policing Services: 2006. pg. 12

18 Ratcliffe, pgs 8-9.

19 Dean Runolfson, OCA Graduate Intern, from a site visit of Cal Anderson Park, January 30, 2009.

20 Dean Runolfson, Survey Provider, from a site visit of Cal Anderson Park, March 25, 2009.

21 Randy Wiger, DPR Community Parks Program Coordinator, in an email attachment to Dean Runolfson, on January 12, 2009.

22 See Appendix B for more information.

23 J. King, K. Mulligan, & S. Raphael, "CITRIS Report: The San Francisco Community Safety Camera Program An Evaluation of the Effectiveness of San Francisco's Community Safety Cameras." Center for Information Technology Research in the Interest of Society (CITRIS). University of California, Berkley. (2008) http://www.citris-uc.org/files/CITRIS SF CSC Study Final Dec 2008.pdf (accessed: May 11, 2009), pg. 161

24 See Appendix A, SMC 18.14.060.A

25 CITRIS, pgs. 86-87

26 See Appendix A, SMC 18.14.030

27 See Appendix A, SMC 18.14.100.C

28 J. King, K. Mulligan, & S. Raphael, "CITRIS Report: The San Francisco Community Safety Camera Program An Evaluation of the Effectiveness of San Francisco's Community Safety Cameras." Center for Information Technology Research in the Interest of Society (CITRIS). University of California, Berkley. (2008) http://www.citris-uc.org/files/CITRIS SF CSC Study Final Dec 2008.pdf (accessed: May 11, 2009).

29 Ibid., pg. 58

30 Ibid., pg. 52

Footnotes 31, 32, 33

34 See Appendix A, SMC 18.14.100.C.8

35 We attempted to conduct the survey on Saturday, March 29, 2009, but the excessive rain resulted in very few park visitors and made it impossible to record the responses on paper that quickly became too wet to write upon or read from afterwards. We therefore chose to exclude the results from the one respondent (ID# 260) we surveyed that day.

36 We took a stratified random sample of Cal Anderson Park users. The survey locations and times are the defined strata (groups) within the population.

37 A graduate school intern (W. Dean Runolfson) from the University of Washington's Daniel J. Evans School of Public Affairs served as the survey provider.

38 Salant, Priscilla & Don Dillman, How to Conduct Your Own Survey. (excerpts) New York, Wiley, 1994, pg. 16.

39 See Appendix A, SMC 18.14.100.C.8

40 Salant, 1994, pg. 17.

41 In an interview with a DPR representative on March 17, 2009, we learned that DPR does not have data on park usage.

42 Salant, 1994, pg. 17

43 Ramirez, Marc, Celebrating a park's rebirth. Seattle Times, September 22, 2005, http://community.seattletimes.nwsource.com/archive/?date=20050923&slug=cal park23 (accessed: May 13, 2009).

44 Salant, 1994, pg. 18 (emphasis added)

---------------

------------------------------------------------------------

---------------

------------------------------------------------------------

12

39

46

79

City of Seattle

Department of Finance

Dwight Dively, Director, Department of Finance

Gregory J. Nickels, Mayor