2019 Surveillance Camera Policy and Anti-Theft Budget Vote

Sponsors

• Ryan Pierce

History

• Announced: TBD
• Language locked: TBD
• Voted on: TBD

Background

Overview

As Pumping Station: One has expanded, it has experienced an increase in tool theft. Surveillance cameras, first installed in 2015 to monitor the building entrances and exits, have limited use in preventing many thefts. A January 2019 revision to the Membership Agreement permitted expansion of cameras if approved by the Board, paving the way for recording members working inside the space (“Membership Agreement,” 2019). In response to theft, proposals on the organization’s e-mail list have ranged from expanding camera usage to instituting criminal background checks. This vote uses an evidence-based approach to reduce tool theft by taking the following actions:

• Updating the surveillance camera policy to prohibit cameras from monitoring internal workspaces. More cameras will only waste time and money, will create privacy issues, and will only serve to distract the organization from implementing real solutions.
• Budgeting $1000 to allow a team of members to prototype and experiment with anti-theft solutions.

More cameras won’t solve the problem

In the past two months, Ryan Pierce conducted an analysis of surveillance camera issues in makerspaces. His primary research consisted of five interviews of PS:One members who hold positions leadership at PS:One and other makerspaces. He also explored surveillance camera practices and policies at other makerspaces.

Pierce (2019) concluded that surveillance cameras monitoring PS:One’s entrances may have contributed to preventing the theft of large, expensive tools. However, none of his interview subjects could recall any instance where analysis of security camera footage was useful in investigating the theft of smaller tools, such as the angle grinders that vanished from the Hot Metals area. These tools could easily fit in a backpack. Further, Sky Nova, who is responsible for managing the surveillance camera system, needed to deny a request to review camera footage because the time window was too long. Frequently, thefts at PS:One are not promptly noticed, and one cannot expect volunteers to scrutinize weeks of footage from multiple cameras. Small items and large time windows render surveillance cameras useless.

Retail stores rely heavily on saturating their sales floors with cameras for loss prevention. In an article concerning retail theft, Colapinto (2008) describes a Manhattan department store with hundreds of cameras, each controlled by a joystick, operated from a command center with twenty screens monitored by three staff members. This approach could never work at PS:One. While more cameras may increase the likelihood of capturing video of a theft, it also increases the amount of video review needed, and PS:One does not have the volunteers needed to find the needle in the haystack.

Pierce (2019) also reported that while no interview subject objected to PS:One’s monitoring of the entrances, some expressed concerns about privacy that would be exacerbated by cameras installed to record interior spaces. At present, the Board has authorized an additional internal camera to record in Hot Metals, which has not yet been installed. One interview subject disagreed with the decision to record in Hot Metals.

This vote explicitly limits cameras to those focused on the building’s external entrances and exits, the building’s exterior, and network, server, and camera infrastructure. Additional language was added to permit mounting cameras on the outside of the building, should this be necessary. Recording networking and computing infrastructure is necessary to capture attempts to disable the camera system; these areas do not overlap with member work areas so should not present a privacy risk. These changes preserve privacy while eliminating collection of data that PS:One cannot reasonably analyze with volunteer labor. This enables the organization to focus on real solutions while avoiding the temptation to throw more cameras at the problem.

Background checks are deeply problematic

One solution proposed on the mailing list was to screen current and future members via a criminal background check. Ignoring the significant cost and labor to administer this, such an approach harbors intrinsic racial bias, would likely decrease PS:One’s diversity, and probably would do nothing to reduce tool theft.

Colapinto (2008) notes that actress Winona Ryder shoplifted from Saks Fifth Avenue, and that one of President George H. W. Bush’s top advisers fraudulently returned purchases at a Target store. Anyone can steal. Assuming that only convicted felons are responsible for stolen angle grinders and cordless drills, and revoking memberships based on past felony convictions in the interest of protecting tools, has no basis in fact.

Felony convictions carry heavy racial bias. Research by Bonczar and Beck (1997) projects that 28.5% of Black males and 16.0% of Hispanic males are likely to be incarcerated, compared with 4.4% of White males. Goldman, Cooper, and Kugler (2018) demonstrate through experiments that Whites, when evaluating fictional job applications for a position, are more likely to rank a Black candidate with a non-violent felony drug conviction lower than a an identically qualified White candidate with an identical felony conviction. These results become even more concerning when viewed in the context of PS:One’s demographics. A recent member survey showed that 82.4% of PS:One members identify as White while only 2.6% identify as African American, and recent cohorts are more White than older cohorts, implying that the space is becoming more White with time (Stein & Ciardi, 2018). Any process to screen members based on criminal background checks would be fraught with racial bias, given both the substantially higher rate of Black felony convictions and the tendency to forgive Whites for felonies that would be disqualifying for Blacks. This would only serve to make PS:One more White. Given this level of racial bias and PS:One’s already problematic demographics, screening members based on criminal background checks is unconscionable.

Analysis of crime prevention strategies

Simmons (2018) describes a problem similar to PS:One’s thefts. In a university library, 12 unattended patron-owned laptops were stolen in an academic year. Tools and laptops are both small, easily concealed, valuable property. An analysis of the problem using Routine Activity Theory and Rational Choice Theory yielded several solutions that the library implemented. The following year, no laptops were reported stolen.

Routine Activity Theory asserts that crime exists within the context of normal, legal activities, it requires “a motivated offender, an attractive target, and the lack of a capable guardian”, and eliminating any one of these three factors prevents the crime (Cohen & Felson; Cullen et al. as cited by Simmons, 2018). Eck expands on this concept with the Crime Triangle consisting of “two nested triangles … [where] the inner triangle consists of three elements of a crime – the setting or place, the offender, and the target. The outer triangle represents three elements of supervision that a motivated offender must circumvent for a crime to occur – a place manager, a handler, and a guardian” (as cited by Simmons, 2018).

Preventing theft by changing the inner triangle frequently is not feasible. PS:One cannot close its space, ban all members, or remove all tools. Since a crime cannot occur unless all three of the outer triangle controls are defeated, this area deserves examination. Having a handler supervise each member individually is not feasible. A place manager may take many forms, such as staff members and security guards, which is infeasible for PS:One, or a device, such as a surveillance camera. Indeed, Simmons (2018) mentions use of cameras in this capacity, however they proved too expensive for the library, which still eliminated laptop theft without relying upon them. As indicated above, surveillance cameras will not work at PS:One. This leaves the guardian, e.g. controls that can be placed on the property itself.

Rational Choice Theory asserts that offenders use a rational decision-making process, based around a cost/benefit assessment, when deciding to commit a crime (Clarke & Cornish; Cullen et al.; Matsueda, Kreager, & Huizinga as cited by Simmons, 2018). This forms the basis for Situational Crime Prevention, which suggests five approaches: “increase the offender's required effort, increase the offender's risk, reduce the reward gained from the criminal activity, reduce the provocation or motivation to commit the crime, and remove any offender excuses (Clarke & Eck; Cornish & Clarke as cited by Simmons, 2018). Simmons (2018) noted the utilization of these approaches in the successful library anti-theft campaign.

PS:One already employs some of these strategies. Security cameras focused on exits have increased offender risk of stealing large, expensive, tools that would quickly be noted as missing. Some area hosts have reduced the reward for theft by defacing the space’s tools with garish paint and obvious PS:One logos. Anti-theft solutions are an excellent addition that could increase an offender’s effort and risk.

Anti-theft solutions

This vote allocates $1000 to research and prototype anti-theft solutions. This amount should be adequate to explore one or more approaches in a limited capacity, but it is still small enough relative to PS:One’s budget that the organization can afford to risk it to research new technologies that may not work. It need not be a complete solution, e.g. only a few tools and one exit door might be monitored initially. Ideally, this research will lead to a workable design of one or more solutions, and costs to implement a full solution will become known. Funding for this can come from a new member vote. While this vote does not dictate or prescribe any solutions to examine, two ideas are presented here for reference as possibilites:

Electronic Article Surveillance – This technology, according to Colapinto (2008), is used in retail stores to sound an alarm when unpaid merchandise leaves the store. A similar concept, such as a longer range equivalent of an RFID tag fitted on vulnerable tools, combined with an antenna system near each exterior door, can track when the tool is in proximity of the door. It could register a silent alarm signal that would log the identity of the tool, the antenna location, and a timestamp. A volunteer could verify the tool is missing and then review the exterior surveillance camera footage from that door at that time. The volunteer need only examine footage for a small time window. Even if the tool were concealed in a backpack, it would still trigger the alarm. This greatly increases the offender’s risk. Should it become known that tool thieves have been caught, that reduces the future provocation to steal.

Tool Checkout – Tools could be stored in individually locked cages or boxes, or they could be tethered to something immovable. PS:One members now have RFID keyfobs, used for building entry. This fob could be used to check out a tool, freeing it from its cage, box, or tether. The identity of the user and the tool, as well as the date and time, would be logged. If the user fails to check in the tool, and the tool is missing from the space, the user could be held accountable. This increases offender effort and risk to steal a tool, as well as decreases provocation by having fewer small, unattended, valuable tools. This also could protect against unauthorized tool use by requiring that the user appear in a database of authorized users before permitting the user to check out the tool.

Arguments against these systems include cost and implementation complexity. Certainly, cost exceeds that of an angle grinder. However, small thefts add up over time, and such a system could easily pay for itself. Doing nothing, or installing more cameras that do not catch any theft, can make matters worse; past successful thefts provide provocation for future thefts by encourage copycats and decreasing the offender’s perceived risk. Also, based on a conversation with a member, solutions like these have been considered before, and should funds become available, some members would be interested in prototyping solutions.

Together, eliminating internal camera surveillance, which is ineffective and intrudes upon member privacy, and funding research for anti-theft solutions will provide the best likelihood of reducing tool theft at PS:One.

Language

The “Surveillance Cameras” section of the Membership Agreement shall be changed from:

1. PS:One reserves the right to place surveillance cameras within the facility where necessary and appropriate to monitor certain public areas by a vote of the directors.
2. Surveillance cameras shall be clearly denoted.
3. All camera locations shall be publicly documented.
4. Only the Board of Directors, DRC, or designated agents may have access to security footage on an as-needed case-by-case basis. Any instance of accessing camera footage shall be recorded in the following member meeting’s minutes and shall include the reason and individual accessing the footage. In DRC matters, where providing a public reason would conflict with confidentiality, surveillance footage access should be recorded in the process notes of the DRC for that raised issue.
5. All camera footage will be kept for a maximum of 6 months, unless part of an ongoing or archive of DRC or Board proceedings.
6. Members may not install recording devices or cameras in the facility without a vote of the membership or vote of the directors.

to:

1. PS:One reserves the right to place surveillance cameras within and on the facility where necessary and appropriate to monitor certain public areas by a vote of the directors.
2. Surveillance cameras shall only record exterior entrances and exits, the exterior of the building and adjacent land, and infrastructure such as servers and network switches used by the surveillance camera system. Surveillance camera placement shall, to the greatest extent practical, avoid recording of areas where members commonly work.
3. Surveillance cameras shall be clearly denoted.
4. All camera locations shall be publicly documented.
5. Only the Board of Directors, DRC, or designated agents may have access to security footage on an as-needed case-by-case basis. Any instance of accessing camera footage shall be recorded in the following member meeting’s minutes and shall include the reason and individual accessing the footage. In DRC matters, where providing a public reason would conflict with confidentiality, surveillance footage access should be recorded in the process notes of the DRC for that raised issue.
6. All camera footage will be kept for a maximum of 6 months, unless part of an ongoing or archive of DRC or Board proceedings.
7. Members may not install recording devices or cameras in the facility without a vote of the membership or vote of the directors.

The Board is authorized to spend $1,000 to research and prototype anti-theft systems to protect PS:One’s tools. The Board may, at its discretion, delegate this spending authority to a working group of one or more members. This authorization expires at midnight December 31, 2020.

Results

Quorum: TBD Present: TBD

For: TBD Against: TBD Abstain: TBD

References

Bonczar, T.P. and Beck, A.J. (1997), “Bureau of justice statistics special report, lifetime likelihood of going to state or federal prison, NCJ-160092”. Retrieved from https://bjs.gov/content/pub/pdf/Llgsfp.pdf

Colapinto, J. (2008). Stop, Thief! New Yorker, 84(26), 74–83. Retrieved from https://search-ebscohost-com.ezproxy1.lib.asu.edu/login.aspx?direct=true&db=aph&AN=34042746&site=ehost-live

Goldman, B., Cooper, D., & Kugler, T. (2019). Crime and punishment. International Journal of Conflict Management (Emerald), 30(1), 2–23. https://doi-org.ezproxy1.lib.asu.edu/10.1108/IJCMA-04-2018-0055

Membership agreement. (2019, January 24). Retrieved November 20, 2019, from Pumping Station: One Wiki: https://wiki.pumpingstationone.org/Membership_agreement

Pierce, R. (2019) Makerspaces and Surveillance – 2019 meets 1984 (Unpublished paper). Arizona State University, Tempe, AZ.

Simmons, H. (2018). A Framework for the Analysis and Management of Library Security Issues Applied to Patron-property Theft. Journal of Academic Librarianship, 44(2), 279–286. https://doi-org.ezproxy1.lib.asu.edu/10.1016/j.acalib.2017.12.021

Stein, A., & Ciardi, A. (2018, December 10). Results from PS1’s first-ever member survey. Retrieved from https://pumpingstationone.org/2018/12/results-from-ps1s-first-ever-member-survey/