This is a short policy brief on hi-tech police surveillance in the United States.

Executive Summary

Surveillance by local police is an increasing problem across the United States of America.

Surveillance is broad, even at the local level, and for that reason, this brief is focusing specifically on surveillance based on technologies that have come available and widespread within the 21st century—though largely within the last decade. Even within these constraints, police surveillance is a tricky issue by its nature: it involves new technologies that vastly outpace older laws, civic traditions, and scholarly literature. The current figures of certain police departments’ surveillance are almost always underestimations, based on disclosed data that is hard-won through lawsuits and likely reflects only a fraction of the national issue. At present, however, there is little evidence to suggest that local surveillance is significantly affecting crime or terrorism, but the lack of oversight and transparency accompanying it set the stage for a chilling of free society. As such it is necessary to have new policies at the national level that make the recording of information more democratic if it must happen, and better regulated if that is at all possible.

Although this policy brief is aimed at the federal level of governance, there is little federal regulation on the surveillance capabilities of police departments. The best responses to the problem of police surveillance have been at city or county levels, but as this problem is particularly political and context-sensitive, my federal policy recommendations have stemmed from the best practices at local levels with high consideration of political feasibility. The best federal options promote or require transparency of local police departments.

Introduction

Surveillance tools used by local police across the United States have vastly increased the depth and sophistication of surveillance. In recent years, local police departments have been acquiring and employing new technology that has resulted in vast amounts of data being intercepted, if not wholly collected, on the citizenry regardless of guilt. As the technology and its deployment is relatively new and still developing, the consequences are not fully understood. Immediate information indicates, however, that surveillance is harmful and poses direct threats to a free society.

This policy brief is specifically focusing on new, “big-data” oriented surveillance across the country, as opposed to (for example) older forms of surveillance such as the wiretapping of the late 20th century. The reason for this focus is simple: never before have authority figures had the ability to process such massive amounts of information on their communities. More information about the devices used by local law enforcement can be found in the Appendices of this brief.

The concluding recommendations focus on political feasibility and take the best examples of surveillance management within the country as a guide. The best options for the United States are to promote or require transparency and to better regulate its dispersal of security funding towards cities.

Why is surveillance a problem?

Some problems, such as poverty or murder, are recognized as inherently just that—problems. Surveillance by authorities is not necessarily considered a problem by all people, particularly by younger generations growing up with data-consuming consumer devices. Therefore, unlike with other issues and policy briefs, surveillance must first be qualified as a problem.

Firstly, privacy is a fundamental right. This is recognized around the world in many countries, but especially so in the United States, with the fourth amendment to the constitution. Opinions of people may vary (especially over time), but that the fourth amendment exists at all is some indication that at least the United States holds privacy to be a right.

Secondly, surveillance has a chilling effect on democratic societies and institutions. Surveillance is the enemy of journalists and activists, and minority groups. The Snowden leaks have revealed a long pattern of state surveillance aimed towards dissidents and activists. So far, local facial recognition databases over-represent those with darker skin tones[1], and there is no reason to believe this will not continue.

Thirdly, there is currently very weak evidence that such massive, dragnet surveillance aids counter-terrorism operations or even helps stop crime. At a national level there may be more contention, but at the local level, one would be hard pressed to make a convincing, results-based argument in favor of surveillance. Much surveillance technology is acquired at the local level for the justification of countering terrorism, but there is very little evidence this is the actual purpose; in fact, there is more evidence to indicate the tech is used in routine police work.[2] Even if the official justification were for countering crime, evidence seems to indicate police surveillance does not prevent crime[3]—perhaps because criminals (and terrorists) have assumed they risk being surveilled on their mobile devices.

Fourthly, and finally, this implies that there is not a strong benefit in terms of cost. Though surveillance may reduce the manpower needed for some operations, it is ultimately expensive if the benefits are not very strong.

Police Surveillance in America

Measuring and qualifying police surveillance across the country is naturally a difficult, and perhaps presently impossible, task. Full disclosure of the problems contributing to knowledge of the problem are necessary for a truly robust policy analysis. Firstly, much information is left to be desired due to lack of disclosure. Most police departments do not immediately and transparently reveal their technologies and use of them until otherwise prompted by outside forces. Indeed, most of the existing information on police surveillance stems from outsiders, particularly the American Civil Liberties Union, local activists, and journalists, prompting disclosures from police. Therefore, unless otherwise noted, one should assume that stated figures are likely smaller than the actual numbers. Secondly, rapid advances in technology produce advances in surveillance (refer to the proceeding subsection for more detail), which makes the dates of facts and figures in fact very relevant. Nonetheless, those that are available paint a picture of a problem affecting many, if not the majority, of Americans.

New technology has allowed the processing of information at levels once unimaginable. An often-cited report from Georgetown Law’s Center on Privacy and Technology revealed that about half of all Americans are in a police facial recognition database.[4] One can only assume the number will increase as face recognition software becomes increasingly common. From available data, 24 states officially permit cell site simulators, though the actual number is almost certainly far higher.[5] Facial recognition cameras and software as well as cell site simulators are the more pervasive big data tools of today, though CCTV cameras and license plate readers, as well as other smaller software, are usually components of a city’s surveillance network.

Conditions producing police surveillance

First and foremost, the development of technology is absolutely key in understanding police surveillance and its growth. There are two facets of technological growth in this respect. Firstly, as police departments acquire increasingly advanced technology, they gain increasingly sophisticated surveillance measures. Secondly, as ordinary people increasingly use technology, increasingly vast amounts of data are recorded on or via consumer technologies. In other words, a key part of today’s police surveillance lies at the intersection between big data and policing.[6]

Additionally, federal grants are a major source of funding. Not all of the technology used by police comes cheap, and the Department of Homeland Security has been a major player in giving cities otherwise expensive gear. Most notably, cell site simulators, known colloquially as stingrays, have become widespread among police departments almost entirely because of federal assistance—primarily through the Department of Homeland Security, though the Federal Bureau of Investigation has been involved as well. Under the Department of Homeland Security Grant Program, the Federal Emergency Management Agency deploys the State Homeland Security Program and Urban Area Security Initiative, budgets with specific availability to urban areas.[7] In 2016, the two programs cost around $982 million.[8] As a result of federal grants, cities that use sophisticated surveillance technology tend to be notable metropolitan areas—as part of the criteria the DHS and FEMA use in awarding grants to urban areas involves an assessment of terror-risk, something that itself is based on metro area population.

Many of the changes that have taken place to allow such extensive surveillance down to the local level can be traced clearly to the September 11th terrorist attacks.[9] American lawmakers and courts have reacted to a perceived threat of terrorism, however real it may in fact be, and have ushered in norms of preventative law enforcement and security-over-privacy on an institutional level.[10] The aforementioned FEMA grants are perfect examples of a focus on preparation and prevention. Though numbers are hard to come by in this area, those that do exist point to a clear rise around the dawn of the 21st century. In 1997, about 20% of police departments used some kind of technological surveillance; by 2007, this number had risen to over 70%.[11]

Finally, it may very well be that the secrecy with which many police departments have been using new surveillance technologies—particularly Stingrays—has been a notable factor in their prevalence. Put simply, there cannot be backlash against a surveillance method if one does not know it is being used. Interactions between police, federal agencies, and the main manufacturer of stingrays (the Harris Corporation) tend to result in pressures on the receiving police by the latter two to keep the capabilities and even the existence of Stingrays secret.[12] For example, when the FBI is involved in the transfer of Stingrays to local PDs, the FBI has the department sign a nondisclosure agreement prohibiting the department from disclosing stingray uses or evidence to judges, or accurately describing the devices when applying for warrants.[13] It is for reasons such as these that we essentially only know about the specific police departments using stingrays because of court orders. Obviously, incidents of community opposition to local police forces’ surveillance have only occurred where the community has been aware of such surveillance.

Reducing this Surveillance

As has been emphasized, the circumstance at hand is new and currently unfolding. Nonetheless, there have been in fact a range of political responses. Political feasibility is key for reforming police surveillance, and these indicate that there are pockets of the country in which this is a politically feasible option.

Conditions of reduction in surveillance

Stability and peace are arguably huge factors involved in gaining political acceptance for transparency or even reduction in surveillance technology/powers. After all, mass-surveillance may have been physically enabled by technology, but it became institutionally ingrained in a post 9/11 world. It may be that time (a good distance from tragedy) can set the stage for a decline in terrorism-focused surveillance. San Francisco’s withdrawal from the FBI’s Joint Terrorism Task Force may be a good example of this, as the move directly counters the dominating rhetoric of being hard on terror, but it is uncertain how well this can be generalized to other cities.

Publicized whistleblowing seems to be the main factor in recent history that can be directly tied to legislative change in surveillance practice. The Edward Snowden revelations of 2013, which were that the National Security Agency had been secretly conducting massive data collections on over a billion people around the world, were shaking to many people. In the wake of the leaks, Congress passed restrictions on metadata collection, and several states passed legislation withdrawing state support for bulk data collection and required warrants for the use of bulk-collecting technology, such as cell-site simulators. In Oakland (see below), the originally planned Domain Awareness Center was significantly scaled back partially because Snowden’s revelations made surveillance a thorny public relations issue. Additionally, it is fairly clear that protests of surveillance only take place when there is knowledge of surveillance in the first place.

As much of the technology comes from grants, cost does not usually seem to be a dominating factor for local police departments in evaluating their surveillance programs.

What Others Have Done

As this policy brief focuses on the national level, understanding plausible options has involved a plethora of cases nationwide. Therefore, I have picked cases that showed notable changes of some kind and that have useful comparative features. Ultimately, the traditional comparative structure of a traditional policy brief cannot hold effectively in this instance, and we must make do with what we have.

Note: For these policies, the consequences and results have not yet been understood. Therefore, I have not included statements on their results. However, insofar as privacy is a right, and insofar as shrouding surveillance in secrecy is wrong, it can be said there is at least some inherent benefit to these policies.

Transparency laws

As noted, to fully aim for stripping police of surveillance tools would be highly ambitious and is politically unfeasible for many places. However, transparency laws can enjoy wider support, probably because they do not involve removing something. It is far easier for even a politically moderate person to get behind greater transparency than removing a policeman’s tool.

Santa Clara County

Santa Clara County provides a particularly unique example of local response, in that an approved ordinance works to account for changes in future technology. The ordinance requires that agencies in the county put in place public policies regarding the use of any surveillance technology before it is acquired, let alone used.[14] In addition, it calls for regular reports on how the technology has been used.[15]

California’s statewide transparency bill

In 2015, California signed into law bills that called for greater transparency and regulation of license plate readers and stingrays.[16] Senator Jerry Hill, who originally introduced these pieces of legislation, later tried to introduce another piece of legislation that would have mandated fully public disclosures to be drawn up by cities using stingrays, and if a public plan could not be agreed upon within a certain time frame, the police agencies would be required to stop using the technologies. The bill passed the California senate, but not the Assembly.

Stingray warrant laws

Though few in number, several states have passed laws requiring warrants before stingray use. California was the latest state to do so (in 2015), with the California Electronic Communications Privacy Act, which has hence required police to acquire warrants before using stingrays. At the time, California was the 5th state to pass such a law, after Washington, Virginia, Minnesota, and Utah. Despite the low number of states, the demographics make it clear the issue is largely bipartisan.

Disassociation from federal programs

San Francisco’s withdrawal from the Joint Terrorism Task Force

In this respect, San Francisco has set something of an example: in early 2017, the San Francisco Police Department suspended its participation in the FBI’s Joint Terrorism Task Force (the FBI’s JTTFs are intelligence-driven).[17] It was widely interpreted as a response to Donald Trump’s “Muslim ban,” but it is important to note this was the culmination of a 7 year grassroots effort.[18] Of course, there are many things unique about San Francisco, so it is unclear how well this case can be extrapolated to other cities.

Better regulation for existing tools

Seattle’s facial recognition tools

Seattle is one of the best cities in terms of regulating its facial recognition program. Seattle began using facial-recognition software in 2014, but from its inception the program had strict controls: the police department only uses mug shots based on “reasonable suspicion” the person has committed a crime.[19] The program is audited and does not allow real-time, live camera facial recognition[20]—perhaps the most invasive form of it. However, Seattle seems anomalous in this regard: such strict regulations are unparalleled anywhere in the country.[21] This would appear to be due to the fact that Seattle PD worked with their branch of the ACLU to make sure the technology was being used with a reasonable degree of regulation.

Grassroots opposition

This last category has likely had the largest number of successes. Although all the aforementioned responses have had grassroots origins, the successes seem rather small in total: for example, grassroots opposition exists in every state, but only 5 states have required warrants for the use of Stingrays. However, if one looks at grassroots efforts focused specifically at local surveillance, they will find more success stories. It is for that reason I give attention to one below (with San Francisco being another example, discussed earlier), as it provides useful examples of how local surveillance efforts have been thwarted, minimized, or otherwise regulated by the citizenry.

Oakland’s Domain Awareness Center

In 2013 residents of Oakland began calling attention to the city council’s plan to accept a multi-million dollar federal grant for a massive expansion of surveillance. The Domain Awareness Center would have built upon existing city equipment in the city’s ports and made a city-wide network of license plate readers, gunshot microphones, cameras, and facial-recognition software. Had it not been scaled back by local activism, it is likely it would have been one of the most advanced cases of city-wide surveillance in the hemisphere.

The results were as follows: the city council adopted a privacy policy, created a committee to address surveillance reform, and scaled back what was originally going to be a city-wide infrastructure to the city’s port only.[22]

It ought to be noted here that Oakland is unique among cities. Firstly, it is a smaller sized city, with a population around 420,000. Secondly, the history has a long and infamous tradition of police abuse and an equally, if not more famous, tradition of resistance to police—it is the birthplace of the Black Panthers. Grassroots opposition towards Oakland’s DAC could have been uniquely strengthened by local culture and history. Finally, and very importantly, the Snowden leaks broke the news during the fight to scale back the DAC, likely affecting the political favorability of such an advanced surveillance network. It is tempting to compare Oakland to San Francisco, a city that is twice as large but very close and that shares a history: it seems progressive cities and cities with a strong tradition of activism will be the most receptive to grassroots efforts. However, it would be unwise to generalize this to other American cities.

Recommended Options

It is not difficult to find examples of local surveillance being addressed locally (this is not to say the issue is being easily solved, simply that ample cases exist in the world’s third most populated country). Translating the diversity of options into a recommended national one is very difficult, as there is essentially no existing federal legislation regulating the use of invasive surveillance technologies by police departments.

Moreover, it would be useless if not harmful to evaluate national options through comparison to other countries. Most other nations with comparable local surveillance networks are authoritarian regimes, and while Canada and the United Kingdom may be exceptions to this, the amount of confounding variables in even these relatively similar countries makes any meaningful comparative analysis untrustworthy at best.

For this reason, a recommended option provided here uses comparisons between places within the United States.

Federally Mandated Transparency

At the most basic level of feasibility, a transparency law far outstrips stronger regulatory efforts. It is politically feasible (and even easy) to counter efforts at stripping surveillance power from authorities: it seems soft on terror and soft on crime. In contrast, transparency and disclosure are much harder to argue against, mainly because they take nothing away.

Additionally, transparency could unlock the power of grassroots activism at the local level and facilitate local change. It is evident that the most robust instances of reducing local surveillance are primarily bottom-up, grassroots efforts, and the ability to know the extent to which one’s police are using stingrays bolsters the chance locals will take notice and take action.

Transparency measures are the common theme in all the options. Transparency measures do not strongly run up against calls to be harder on terror, and are emerging precedents in the policies of police surveillance. Finally, if California gives any indication, it is that transparency and disclosure can set the ground for more democratic discussions of how to handle the technology.

Appendices

Appendix A: The Cell Site Simulator

Stingray_Harris_handle_side

A cell site simulator, more popularly known as Stingrays, are mass-surveillance devices commonly used by local law enforcement. Stingrays imitate cell phone towers and thereby trick nearby cellphones into connecting to the Stingray. Once this happens, locations and other identifying information is transmitted.[23] However, Stingrays are not precision tools: they may be used in the effort of locating or surveilling a single person, such as a suspect, but in the process they can collect the data of everyone else within a certain radius of the device.

Appendix B: A Note on the Case Studies

A brief reading of the options list would reveal something odd: it is largely California cities, and the exception is simply up North (Seattle). The reason for this is simple: California provides many good traits for the purpose of policy analysis. Firstly, its cities have made the largest strides forward in handling surveillance; secondly, state law has resulted in the disclosure of surveillance technology and subsequently in efforts to reduce this surveillance, so there is a greater body of evidence; third, it is populous and has many metropolitan areas; fourthly, it provides the best chance of comparison in a brief already condemned to struggle with comparisons.

In short, California is a sort of Goldilocks zone for those studying the policies of police surveillance: it has all the right problems, all the right solutions, and documentation of them. It must absolutely be acknowledged that California is a strongly liberal state with liberal cities. However, the repeated successes could be a safe indicator of a still-emerging policy climate.

Appendix C: Georgetown Law’s Center for Privacy and Technology Report

The aforementioned report is a landmark study and has been cited many times in the area of policing and facial recognition. It is one of the few reliable sources on national facial recognition software use and its validity should be interrogated because of its importance.

The study involved over 100 records requests to police departments nationwide, by default making it the most comprehensive survey of its kind. In addition, the authors went through FBI data with the information obtained from the local agencies, to find that 117 million Americans adults have been in a police facial recognition database. As they note, there is practically no larger oversight of this; it is something entirely at the discretion of the local leaders.

Appendix D: Big Data

“Big data” is a term defined best by Sarah Brayne of University of Texas at Austin. Sarah describes big data as “massive and diverse datasets,” but it is important to note the temporal attachments to this term: it certainly refers to this time period, in which massive amounts of information have been digitized.[24] Big data also has a strong connotation of use: that is to say, big data is always analyzed, manipulated, transferred, sold, etc.

Works Cited (in order of citation)

[1] Clare Garvie, Alvaro Bedoya, and Jonathan Frankle, “The Perpetual Line Up: Unregulated Police Face Recognition in America,” Perpetual Line Up, October 18th, 2016. https://www.perpetuallineup.org/

[2] Adam Bates, “Stingray: A New Frontier in Police Surveillance,” Cato Institute, January 2017. https://object.cato.org/sites/cato.org/files/pubs/pdf/pa-809-revised.pdf

[3] Ibid.

[4] Clare Garvie, Alvaro Bedoya, and Jonathan Frankle, “The Perpetual Line Up: Unregulated Police Face Recognition in America,” Perpetual Line Up, October 18th, 2016. https://www.perpetuallineup.org/

[5] “Stingray Tracking Devices: Who’s Got Them?” American Civil Liberties Union, March 2018. https://www.aclu.org/issues/privacy-technology/surveillance-technologies/stingray-tracking-devices-whos-got-them

[6] Sarah Brayne, “Big Data Surveillance: The Case of Policing,” American Sociological Review 2017, Vol. 82(5) 977-1008. DOI: 10.1177/0003122417725865

[7] FEMA, “Fiscal Year 2016 Homeland Security Grant Program,” FEMA.gov, June 2nd, 2017. https://www.fema.gov/fiscal-year-2016-homeland-security-grant-program

[8] Ibid.

[9] William Bloss, “Escalating U.S. Police Surveillance after 9/11: an Examination of Causes and Effects,” Surveillance and Society Part 1, 4(3): 208-228. https://queens.scholarsportal.info/ojs-archive/index.php/surveillance-and-society/article/viewFile/3448/3411

[10] Ibid.

[11] Phil Ciciora, “Better police surveillance technologies come with a cost, scholar says,” Illinois News Bureau, November 11th, 2013. https://news.illinois.edu/view/6367/204700

[12] Adam Bates, “Stingray: A New Frontier in Police Surveillance,” Cato Institute, January 2017. https://object.cato.org/sites/cato.org/files/pubs/pdf/pa-809-revised.pdf

[13] Ibid.

[14] Eric Kurhi, “Pioneering spy-tech law adopted by Santa Clara County,” The Mercury News, June 2016. https://www.mercurynews.com/2016/06/07/pioneering-spy-tech-law-adopted-by-santa-clara-county/

[15] Ibid.

[16] “Governor Signs License Plate Reader and Cell Phone Intercept Privacy Bills,” Senator Jerry Hill, October 2015. http://sd13.senate.ca.gov/news/2015-10-08-governor-signs-license-plate-reader-and-cell-phone-intercept-privacy-bills

[17] Michael German and Emily Hockett. “San Francisco Sets an Example of How to Resist Surveillance in Trump Era,” Brennan Center for Justice, March 2017. https://www.brennancenter.org/blog/san-francisco-sets-example-how-resist-surveillance-trump-era

[18] Ibid.

[19] Steve Miletech, “Seattle police win praise for safeguards with facial-recognition software,” Seattle Times, October 2016. https://www.seattletimes.com/seattle-news/crime/seattle-police-wins-praise-for-safeguards-with-facial-recognition-software/

[20] Ibid.

[21] Ibid.

[22] Oakland City Council, DAC Privacy Policy. May 2015. https://www.aclunc.org/sites/default/files/20150521-dac_privacy_policy.pdf

[23] “Stringray Tracking Devices,” American Civil Liberties Union, accessed March 2018. https://www.aclu.org/issues/privacy-technology/surveillance-technologies/stingray-tracking-devices

[24] Sarah Brayne, “Big Data Surveillance: The Case of Policing,” American Sociological Review 2017, Vol. 82(5) 977-1008. DOI: 10.1177/0003122417725865

Leave a Reply

Your email address will not be published.