Thursday, June 13, 2002

     Madam Chairwoman, members of the DC City Council, thank you for the opportunity to testify today at this Joint Oversight Hearing on the proposed video surveillance plans of the Metropolitan Police Department. I appear before you today in both a professional and personal capacity. As Executive Director of the Electronic Privacy Information Center (EPIC), I work on a wide range of privacy issues concerning new technology. I have taught Information Privacy Law at Georgetown University Law Center for more than a decade. I served as Counsel to the Senate Judiciary Committee after law school and focused on new technology issues. I also served on the ABA Criminal Justice Committee that developed guidelines for Technology Assisted Physical Surveillance (TAPS) specifically addressing the question of the use of video surveillance cameras by the police. Finally, I have also lived in Washington for almost twenty years. My wife, a former economist, is now a public school teacher here in the District and our children are in the DC public school system. We have a strong personal commitment to the city, the schools, and our community.

     Let me now thank you and the members of the DC City Council for your efforts to create a legal framework for the use of video surveillance in Washington, DC. No other city has looked so closely at how best to regulate this new technology. I commend you, Councilmember Patterson, for your important work on this issue. It is clear that the deployment of a video surveillance system in Washington is a matter of great public concern.

     First, the use of surveillance cameras in Washington, DC raises far-reaching Constitutional questions that implicate the rights of residents and visitors, and most significantly people who engage in peaceful public protest in our Nation's capital.

     Second, the benefits of video surveillance systems as a means to reduce crime and deter terrorism have been significantly overstated. Studies from London and Sydney make clear that the value of cameras is overstated and that money is better spent on officers than on cameras. Moreover, the particular effort to promote the use of face recognition technology may be one of the biggest corporate boondoggles in recent history, costing taxpayers hundreds of millions of dollars with little benefit in return.

     Third, the proposed guidelines of the Metropolitan Police Department fail to provide adequate protection for residents and visitors in Washington, DC. The rules take too restrictive a view of the expectation of privacy and First Amendment protected activities. The proposed system of public surveillance lacks adequate means of independent oversight. The reporting requirements are vague, the policy on usage and retention is insufficient, the definitions are too narrow and the auditing is too limited.

     If a system of video surveillance is established in Washington, DC, there must be a far more robust legal framework than the one currently before the City Council. The use of electronic surveillance is highly regulated in this country under both federal and state law. The use of video surveillance, which would sweep far more broadly across innocent activity, must be regulated at least as stringently as surveillance of electronic communications. This will require judicial review, public reporting, rules for minimization, and clearer restrictions on the use of any information obtained.

     Fourth, I strongly urge you to consider the creation of an independent commission to evaluate the effectiveness of the plans and to review these proposals on an ongoing basis. Many cities around the world are considering whether to adopt CCTV systems. Washington could help bring the principles of democracy, transparency, and public participation to this significant challenge.


     Some have stated that the legal question concerning video cameras in public spaces is simply whether one has a reasonable expectation of privacy in a public place. I believe that this perspective fundamentally misstates the Constitutional interests at stake in this debate.

     There is no dispute that the police have a critical role in protecting public safety in the Nation's capital. This is particularly true when a large number of people are gathered for political protest. Protesters, as well as residents and tourists, have an interest in ensuring that public assemblies are peaceful and do not endanger persons or property.

     I want to be clear that this concern about surveillance of Constitutionally protected activity is more than theoretical. Shortly after EPIC learned of the plan to install video cameras in public places, we submitted a series of Freedom of Information Act to several agencies in the District. A response that we received from the United States Park Police, attached to this statement, is particularly revealing.

     The documents that we obtained contain individual logs of the aerial surveillance conducted by the DC Municipal Police. I would like to call your attention to the activities of surveillance by the MPD.

Although all of these incidents implicate Constitutional matters, the last example may be the most significant because it makes clear that video surveillance is specifically undertaken of individuals engaged in political protest. This clearly implicates Constitutionally protected freedoms.

     It is also clear that aerial surveillance is increasing. Many of the records we obtained concerned the recent protests in April of this year.

     There will be serious Constitutional questions that arise when images obtained by the police are used in trial against a criminal defendant. Particularly with a technology where it is so easy to manipulate images and so difficult to ensure the integrity of a digital file, very clear rules for retention, chain of custody, and use must be established.


     I'd like to make a few remarks about the value of CCTV to protect public safety. Here again, I would like to provide a bit of perspective on these issues. There is no doubt that technologies and certain common sense practices reduce crime. Locking car doors, for example, is a simple and effective practice that dramatically reduces the risk of car theft. In the realm of criminal investigation, fingerprint identification continues to play a critical role in crime scene investigations.

     But the benefits of CCTV and particularly face recognition technology that may soon follow have been dramatically overstated. A far-reaching study by Dr. Clive Norris and Gary Armstrong of the Centre for Criminology and Criminal Justice at Hull University, "The Unforgiving Eye: CCTV Surveillance in Public Spaces" found that:

The gaze of the cameras do not fall equally on all users of the street but on those who are stereotypical predefined as potentially deviant, or through appearance and demeanor are singled out by operators as unrespectable. In this way youth, particularly those already socially and economically marginal, may be subject to even greater levels of authoritative intervention and official stigmatization, and rather than contributing to social justice through the reduction of victimization, CCTV will merely become a tool of injustice through the amplification of differential and discriminatory policing. (P. 8)

     According to a second study in Sydney, the CCTV system produced one arrest in 160 days. The study concluded, "Before limited resources are spent on surveillance cameras, close attention must be paid to the claimed benefits."

     Even the study conducted by Booz Allen of the National Capital Area raised significant questions about the value of CCTV. A report titled "Counter-Terrorism Plan for National Park Service, National Capital Region", obtained by EPIC under the Freedom of Information Act, stated that:

the most effective countermeasure available to the NPS is the beat officer. No computer or other technological device can replace the human officer whose perceptual system and brain far exceed any other device in coming to a logical analytic and conclusion concerning a potential terrorist situation. (p. vii)

     The report goes on to recommend integrated detection, monitoring and surveillance, including CCTV, which would "allow surveillance of multiple memorials and sites simultaneously," but then goes on to note that "CCTV will not replace the need for an officer to view and assess the situation."

     This report, which is the most comprehensive survey of counter-terrorism measures for the national Capital that has been released to the public to date make clear that, at best, CCTV will be only a partial solution to any steps that are taken to safeguard the National Parks and memorials.

III. Review of the Proposed changes to the DC Code

     I will now go through the specific provisions in the proposed changes in the DC Code and point to the areas where changes should be made.

PURPOSE (Section 2500)

     At the outset, the purpose of the Metropolitan Police Department Use of Closed Circuit Television is not clearly stated. Sect. 2500. Instead we are told that the MPD has deployed a CCTV system "within the Synchronized Operation Command Complex (SOCC) that are highly secured and protected against unauthorized access." This is troubling. What is the purpose of the camera system? It should be stated at the outset. And regarding the MPD's representations that the network is secure, we understand that many of the cameras operate on a wireless feed, which raise obvious security problems.

     The statement that the MPD shall comply with federal and district law is a good starting point but hardly answers the question of what precisely the MPD will do. Sect. 2500.2. Is the MPD aware that the use of audio recording in conjunction with CCTV necessarily triggers the federal wiretap statute? A better formulation of this section would include the specific provisions of law that are currently applicable, as well as acknowledgement of any other provisions that may also apply.

POLICY (Section 2501)

     The Policy statements that follow, such as the "CCTV systems represent a valid use of government's power to protect its citizens and will be utilized to enhance public safety" read more like a press release than a finding of fact or a statement of purposes. Sect. 2501.2. Specifically, I am not aware that a legislative determination has been made that the use of CCTV "represents a valued use of government power" and I do not believe that anyone can assure the city that the system will be used only to enhance public safety. On this particular point, if the lessons of London are clear, it is more accurate to suggest that the use of CCTV will be to extend social and political control over the people who work, live, and visit Washington, DC.

     Sect. 2501.3 suggests that the CCTV proposal may be as much about perception management as it is about actual crime reduction. One would not dispute the aim of preventing crime, but to see that the CCTV systems are also intended to augment the District's efforts to prevent "fear of crime" suggests that plan is very much about public perception. This approach raises a particular problem on the evaluation side: how does one evaluate the success of a CCTV system? Let us imagine that crime has gone up after CCTV is installed but the public, perhaps aided by an effective public relations campaign, believes that the cameras have helped deter crime. How do we evaluate the system under the standard set out by the proposed provision in the DC Code?

     Section 2501.4 should make clear that the system should not be used where it would chill First Amendment freedoms.

     Section 2501.5, which permits the MPD to obtain video feeds from public and private entities, authorizes a general purpose search without judicial review that flatly contradicts the intent of the Fourth Amendment. Should the police be able to obtain a real-time transfer of the banking transactions conducted by a customer at an ATM machine on Connecticut Ave.? Of course not. Then what could be the basis to allow a video feed of the same conduct without judicial approval?

     If the City Council approves this provision, and I urge you not to, then there must be very elaborate systems of public oversight and judicial review. There is also no indication in section 2502 of the limitations on duration. This is clearly necessary.

USAGE (Section 2503)

     The section on Usage, section 2503, is possibly the most critical section of the new chapter. Sect. 2503.1 correctly prevents the use of CCTV to target individuals on a discriminatory fashion. But 2503.2 goes on to say that the CCTV system should be used to observe "locations that are in public view where there is no reasonable expectation of privacy."

     Here a critical point must be made -- the expectation of privacy is dependent on the nature of the observation. A person sitting on a bench during lunchtime reading a newspaper near the reflecting pool on the Mall is certainly aware that they may be viewed by others who pass nearby, but does anyone really believe that this person expects to be observed by a hidden camera 50 yards away, or that as they read the newspaper with no one in sight, that several people could be observing the person from a television monitor in the SOCC? This is the critical flaw in the expectation of privacy analysis set out in the new chapter: the diminished expectation of privacy associated with the presence of others in one's physical vicinity cannot become the standard for hi-powered CCTV system that covertly observes, monitors and records activities for observation by others that cannot be seen and are not known to the subject.

     Such an approach to the expectation of privacy analysis confuses the subjective expectation of the observed with the technological prowess of the observer. It is contrary to the legal analysis and it will set the District on a downward spiral that will transform our wonderful public spaces into broad-based zones of surveillance.

     Section 2503.3 seeks to protect First Amendment rights, but it focuses on the wrong subject. It is not the handbills or fliers that have First Amendment rights, it is the individuals participating in peaceful public protest that have these rights. Section 2503.3 must be altered so as to make clear that CCTV will not focus on the faces of the individuals, that is to say the identity of the individuals, without indication of an actual threat to public safety. Any other standard makes a mockery of this provision. Said differently, if a hundred thousand people participate in a pro-life march how are their First Amendment freedoms protected if the signs they carry are not observed but their faces, and the faces of their friends and children who may join them in protest, are observed and recorded?


     Section 2504 is far too vague. There must be explicit documentation requirements and a clear format for reporting, similar section 2518 of the federal wiretap act. The names of the person conducting the recording should be noted as well as the period of record, the use, and the disposition of any data obtained.

RETENTION (Section 2505)

     The retention provisions in section 2505 lack the fundamental requirement of independent judicial review. In the federal wiretap context, it would not be left to the police to determine whether to extend the period of an authorized wiretap. Such approval would be obtained from the judiciary. Most critically, this is nothing here that limits continual extension of the data retention period by the chief of police. That can not be right.

     There must be a provision that makes clear that in no case may retention be extended beyond 30 days without court approval. Moreover, the indexing provision contemplated in 2505.1 raises privacy concerns if the index is linked to identifiable individuals. In those circumstances, both federal and state law will limit the collection, use, and disclosure of the records maintained.


     I fully support the effort in section 2507 to notify a community about the decision to deploy cameras in residential areas and to provide the community with the opportunity to comment on the plan. But for community input to be meaningful, the notice has to be made widely available, the full capabilities of the system publicized, and final decisions subject to judicial approval. Moreover, a time period must be specified for the camera installation, otherwise we can simply anticipate that over time, more and more camera systems will be deployed across Washington. Specifically, in section 2507.8 the MPD must provide the exact capabilities of the CCTV systems, their precise location, and their use in departmental operations, absent a specific and articulable reason on a case-by-case basis why such information should not provided. How else does a parent know when they kiss their child in the morning at the front door of their home whether their action will be viewed and recorded by a police camera across the street? And if there are going to be video cameras in residential neighborhoods in the District (and that is an issue that residents must carefully consider) the standards for the use of video cameras in residential neighborhoods must clearly be higher than they are in places of public gathering.

     Regarding the signal provision in 2507.9, a mere notice provides no privacy protection. In fact, it will defeat claims of privacy. A far better approach is to provide a means for residents and visitors to contact an independent commission if the believe that the cameras have been misused or are inappropriate. There should be an 800 number to call, unrelated to the MPD, that allows individuals to register complaints about the presence of the cameras and that should be included on any sign regarding video surveillance in the Nation's capital.

     The reporting requirements suggested in 2507.10 and 2507.11 should be far more specific. The annual report should indicate locations of camera, periods of operation, whether recording occurred, the purpose of the recording, the disposition of the tape, the officer in charge, and some evaluation of whether the camera promoted public safety according to scientific procedure and by comparison to other jurisdictions.

AUDITS (Section 2508)

     The outcome of the audits described in 2508.1 should be made easily accessible to the public, perhaps by posting on the Council of the District of Columbia website.

DEFINITIONS (Section 2509)

     The definitions set out in 2599 are far too narrow. For example, in this context, CCTV should mean any CCTV system deployed by the MPD or any other agency of the District of the Columbia or any person acting on behalf of the MPD or any agency of the District of Columbia. The definition for exigent circumstances should also make clear that it exists when it is impracticable to comply with the regulations and should, in any event, be limited in time to 48 hours.


     The DC City Council has begun a very important effort to bring the use of new surveillance technology within the rule of law. This is critically important for the protection of privacy and I want to thank you for your efforts.

     But the proposed revisions will need to go much further to comply with Constitutional requirements, to safeguard the rights of residents, the people who work in Washington, and those who visit our nation's capital.

     The District of Columbia is beginning the process of building a system of hi-tech police surveillance for a city that has long cherished freedom. I urge you to proceed with great caution.


EPIC, Video Surveillance Information Page

The Observing Surveillance Site


Marc Rotenberg, Executive Director
Electronic Privacy Information Center
1718 Connecticut Ave., NW, Suite 200
Washington, DC 20009
1 202 483 1140 (tel)
1 202 483 1248 (fax)