If you liked Survivor, just imagine a new reality-based television show that captures New Yorkers in their most intimate public moments. You'd see politicians' daughters buying drugs in Tompkins Square Park, topless tanning in Central Park, and CEOs stumbling out of midtown bars after having a few too many. Why not? In a culture like ours that thrives on voyeuristic thrills, the show would no doubt be a hit. Viewers in the United Kingdom were treated to a compilation of "juicy bits" from government closed circuit television (CCTV) cameras when Barrie Goulding launched Caught in the Act. The program featured the sexual and other intimate activities of innocent people as well as lawbreakers. Great Britain, the spawning ground of reality television, has 1.5 million television cameras monitoring the public, more than any other nation, but the United States is rapidly catching up.

Thousands of cameras, both publicly and privately owned, dot city streets and parks. Once someone has gone out in public, courts have refused to recognize any expectation of privacy from being watched on camera, either by private (in tort law) or government (under the Fourth Amendment) actors. But people are, by and large, unaware that they are being watched. Nor do they know or control what their images are being used for. The problem is compounded by the potential use of facial recognition technology to compare the faces of people who simply step into public places with the database images of suspects and convicted criminals, an action that in other forms would require probable cause. Law enforcement's track record of disproportionately targeting racial and ethnic minorities, combined with the high rate of false positive matches, makes it likely that innocent people will be arrested without any independent justification.

Electronic surveillance has sent a detectable chill through our public spaces. The first clause of the Fourth Amendment declares, "The right of the people to be secure in their persons, houses, papers and effects, against unreasonable searches and seizures shall not be violated." A faithful reading allows a reasonable expectation of freedom from searching scrutiny of one's person, papers and effects wherever one is, including outside on public streets and in parks. No case has squarely faced this issue, and the current Supreme Court would likely find that electronic monitoring of public areas is not a search subject to Fourth Amendment scrutiny. Chief Justice William Rehnquist has made clear his views that law enforcement interests will generally trump any limited right to privacy while out in public. However, if the Court adopted a more honest test, it would recognize the distinction between what is actually in plain view (and hence reasonable to observe) and what is capable of being seen if technology is trained on the public in ways it doesn't and shouldn't expect.

No unreasonable search occurs when a beat police officer glances around or even stops to survey passersby. But when law enforcement films the public engaging in all the activities of mundane existence, the government crosses the line. Even more invasive is taping video footage, using nightscopes, and using biometric technology to analyze the faces or carriage of presumptively innocent citizens to find matches with criminal suspects. Biometric technology, which has been piloted in Tampa, Florida; in airports in Boston, Providence, San Francisco, and Fresno; and at checkpoints at the U.S.-Mexico border, may appear to be a "perfect search" because it is supposed to single out only known offenders. But in reality, the rate of error vastly exceeds accurate matches.

As part of the 1998 New York Civil Liberties Union Surveillance Camera Project, volunteers walked every block in Manhattan and found 2,397 publicly and privately controlled cameras trained on public spaces. The thickest concentrations are in midtown, the financial district, in lower Manhattan parks, and around housing projects. In one eight-block radius alone, volunteers noted three hundred cameras. But there are likely many thousands more because of the routine security practice of buttressing visible cameras with concealed ones "so everything's covered and it doesn't look like a fortress," a consultant told the Village Voice (Sept. 30-Oct. 6, 1999). Public cameras are typically mounted on traffic and streetlight poles, public buildings, and trees and installed in buses, on subway platforms, and inside subway cars. The full extent of saturation isn't known, because police officials refuse to divulge their locations, claiming that it would undermine law enforcement's effectiveness, and besides, the majority of cameras are nongovernmental, owned by countless private companies.

Bill Brown, founder of the privacy watchdog group The Surveillance Camera Players, gives walking tours of New York City surveillance camera locations. He estimates there are about six thousand cameras in Manhattan. He noted some of the questions, constitutional and commercial, raised by an unmarked camera. "In addition to watching the front door, it's also pointing out into public space and recording information about the passersby: Who are they with? What are they wearing? What are they smoking? Are they drinking Coca-Cola or Pepsi?"

A number of factors make an electronic eye inherently more intrusive than direct observation by a human being. Law professor and legal columnist Sherry Colb notes that you can't stare back to discourage the privacy incursion. Nor can you choose to move away from the area under surveillance if you don't know where it is. People behave differently when they think they are alone. A Tampa police-camera operator Raymond Green told the St. Petersburg Times: "Some things are really funny, like the way people dance when they think no one's looking." On the other hand, even if people are aware of the cameras, this does not alleviate the problems. As Colb signals, awareness that we are being watched by surveillance cameras can put us in a "constant state of apprehension and self-consciousness whenever we are out in public." Sociologist Steven Flusty aptly calls this monitored public space "jittery space." Rather than promoting the ideals of a democratic society, it creates a police state environment that chills the exercise of First Amendment rights and further polarizes race relations.

Police officers have long singled out African-Americans and other people of color for mistreatment. Camera operators may be even more likely to target members of stigmatized social groups for monitoring when the public can't see them. From a control room it's easier to single one person out, and it's less noticeable. Colb identifies targeting as a secondary, but important, privacy harm. When a person is singled out by police with no legitimate basis, he or she is left asking, "What made them think I'm a criminal?"

Criminologists Clive Norris and Gary Armstrong answer that question in their study of CCTV monitoring in Great Britain. In the absence of concrete guidance on whom to monitor, "CCTV operators selectively target those social groups they believe most likely to be deviant." The most over-represented characteristics of those scrutinized were maleness, youth, and blackness. Operators also focused on anyone running or loitering, even though this rarely led to finding criminal conduct (of course, disorderly conduct also attracted the operator's attention); anyone who appeared to be spatially or temporally out of place-the most obvious being drunks, beggars, the homeless and street vendors; people whose clothing style signaled they were trouble-makers; people who challenged the authority of the camera to watch them; and people who seemed to be trying to conceal their identity. Norris and Armstrong found that nearly four out of ten people were monitored for "no obvious reason."

Cameras do not merely serve the purpose of an automated beat cop, contrary to courts' assumptions. They are more like super cops. Video feeds can be linked to a nationwide or international database to scan for matches with criminal suspects (though thus far such efforts have faltered because facial recognition software is so imperfect). Images can be enhanced far beyond the visual capabilities of a human being. The cameras can view faces a hundred yards away, see in the dark with infrared technology, and zoom in to read a letter someone is holding, which police couldn't do without probable cause. Furthermore, they can be placed in a position where a person couldn't be, such as high up on a wall.

The potential for abuse is considerable: the technology could easily be used for sexual voyeurism, racial profiling, or to harass gays, lesbians, and transgendered people. Unlike the eye simply catching a fleeting glance of an embarrassing moment, the camera makes it permanent and possible to disseminate widely. The Detroit Free Press found that Detroit officers used the video database of the Law Enforcement Information Network to help themselves or their friends stalk women, threaten motorists after road rage incidents involving the police as drivers, and intimidate supporters of political rivals.

In weighing the intrusiveness of cameras, courts suggest comparing the privacy incursion against law enforcement needs; we also need further comparison with the proportion of positive outcomes. Of all the people subjected to electronic lineups in stadium scans at the Superbowl, the system succeeded only in flagging some petty criminals, such as ticket scalpers, not the menaces to society, such as terrorists and violent criminals, that officials claimed to be targeting with the searches. In Tampa, an experiment with Face-IT, a face recognition application, never correctly identified a single face in its database of suspects and didn't result in a single arrest, but had many false positives.

Although cameras are billed as necessary to prevent serious crimes, they are used to raise revenue and catch relatively minor offenders. New York City, San Diego, and Washington, D.C., use cameras at bus stops and intersections to ticket illegally parked vehicles and catch speeders and red-light runners. After data were released in San Diego, the court threw out hundreds of traffic tickets. The data showed that accidents at monitored intersections actually increased. The city's vendor company (Lockheed Martin IMS) had shortened the yellow-light time to catpure more offenders. In Great Britain, the Environment Agency installed a tiny camera in a discarded Coke can to catch people in the criminal act of dumping trash, at a cost of 3,500 pounds (or about $5,000).

In "Privacy and the Reasonable Paranoid," her University of Toronto Law Journal article (Summer 2000), New Zealand scholar Elizabeth Paton-Simpson posits a character, Prudence, who takes extreme precautions to protect her privacy. For example, she keeps her curtains closed day and night; refrains from any public displays of affection, aware that anything she does may be permanently recorded and broadcast to the nation; and never uses public restrooms for fear of being spied on by security or police. I would add that she should wear sunglasses and a wide-brimmed hat at all times when appearing in public because of the risk of false positive matches with terrorist suspects by facial recognition software. Prudence sounds very paranoid, but behaves in accordance with the widespread view of what constitutes a reasonable expectation of privacy in American jurisprudence.

In Katz v. United States, the court ruled that the limits of Fourth Amendment protection "cannot turn upon the presence or absence of a physical intrusion into any given enclosure." Later decisions have made it clear that a legally recognized expectation of privacy hinges on social norms. A person's subjective expectation of privacy is only "legitimate in the sense required by the Fourth Amendment . . . [if] the government's intrusion infringes upon the personal and societal values protected by the Fourth Amendment." Hence, as Sherry Colb notes, the Court has constructed Fourth Amendment doctrine "by reference to the privacy that one might legitimately expect to have from other private actors, independent of any state surveillance."

When people go out in public, especially in a large city such as New York, they expect anonymity. This is perhaps a paradox, but in a crowded street, individuals are unlikely to really be noticed. As privacy activist Bill Brown observes, "These surveillance cameras strike at the single most beautiful quality about New York, which is, quite obviously, the ability to become anonymous." Of course people realize they will be seen, and perhaps recognized by others when they step outside into public view. I do not suggest that police cannot look at what everyone else can. By stepping outside we implicitly consent to be seen and observed by others, but certainly not to be scrutinized, stalked, or recorded. When the government uses cameras to perform these functions of hyper-scrutiny, it is violating social norms and implicating Fourth Amendment protections.

UNFORTUNATELY, the Court has taken a misstep in fleshing out the concept of knowing exposure. In United States v. Knotts, the defendant Darryl Petschen was tracked by police by means of a beeper planted on a container in his car. The Court reasoned that when he traveled over public streets, he voluntarily and knowingly conveyed-to anyone who wanted to watch-where he was going and where he stopped on the way; therefore he had no legitimate expectation of privacy in his movements outside. The Court concluded that warrantless, continuous police surveillance was no intrusion and did not constitute a search. But if the police had continued to track the container as it was moved around inside a private residence, hidden from view, then the monitoring would have become a search. As Colb points out, the Court's analysis is counter-intuitive, because clearly anyone would consider the police tracking your every move once you leave the house more invasive than simply keeping track of where in your garage you park your car. Only the strictest formalist would consider your doorstep the place where privacy begins or ends.

Myriad social rules and practices govern our actions concerning privacy in public. Sociologist Erving Goffman documented some of the rules for territory and personal space in public places and limitations on staring. He noted the example of men using urinals in public restrooms: for a period of time they are in close proximity and must expose themselves. Men take considerable care where they look to avoid violating privacy more than necessary. Similarly, when we meet an acquaintance, there is an obligation to greet the other person. An incidental effect of this rule is protection from a sort of spying observation or eavesdropping. We all learn these rules as part of growing up. Social sanctions such as requests to stop staring, or stop listening, or responding with a counter-stare usually keep violators in line.

Conceptual artist Sophie Calle manipulated generally accepted social practices when she found a man's address book by chance and, before returning it to him, photocopied the contents. She then called all of the people listed, interviewed those who were willing, and published their comments about him in a column in the French newspaper Libération without ever meeting him. The object of this unwanted attention was extremely upset. Though he did leave his address book lying out in the public view, he in no way consented to having his friends, family, and acquaintances contacted and asked about him. These examples illustrate the fallacy of what Colb identifies as the Court's dubious logical move of equating risk with consent, whereby it treats a person who takes a risk of something occurring, even if it is highly unlikely, as having invited its occurrence. As Paton-Simpson notes, "empirically, social factors are relevant in shaping people's expectations of privacy," and the law should realistically take them into account.

In a case involving a serious crime, United States v. Torres, for which a warrant was obtained, Seventh Circuit judge Richard A. Posner noted "that television surveillance is exceedingly intrusive, especially in combination (as here) with audio surveillance, and inherently indiscriminate, and that it could be grossly abused-to eliminate personal privacy as understood in modern Western nations." While he declined to rule that video surveillance is unconstitutional, Posner did caution that it should not be used as generally as less intrusive techniques that are available. Even if law enforcement obtained a warrant, a search could be unreasonable if it infringed on personal privacy out of proportion to the "likely benefits from obtaining fuller compliance with the law."

Some states, including California, Hawaii, Illinois, Montana, Oregon, and Pennsylvania, have heightened protection of the right to privacy under their state constitutions. But only California and Hawaii's laws specifically limit video surveillance. There is no federal legislation governing video surveillance. Congress has not yet addressed video surveillance, even though it is in many ways more intrusive than wiretapping. For example, a video camera continuously monitors everything that transpires, whereas surveillance by wiretap will only be triggered when someone speaks. As the court observed in Commonwealth v. Kean, a videotape "is likely to contain a richness of detail which could not be successfully communicated by even the most articulate of observers. One does not need a law degree in order to understand that a picture is worth a thousand words."

Eighteenth-century philosopher Jeremy Bentham's panopticon imagined a structure that would facilitate the social control of prisoners. A guard tower would be at the center, with prison cells built in a circle around it so that an unseeable guard at the center could always see the prisoners' silhouettes. The system would be effective whether or not there were always a guard in the tower, because prisoners would be aware that they could be watched at any time. Modern surveillance systems more closely mirror Michel Foucault's vision of social control enforced by myriad public and private institutions, as public and private cameras have created decentralized zones of surveillance. Reporter Mark Boal, writing in the Village Voice (Sept. 30-Oct. 6, 1999), captures the elusiveness of the surveillance problem, and why "we the people" have not rebelled:


Private and police cameras essentially have the same function: to protect private property. And both have a similar effect of making video-monitored space feel controlled, not free and public. But it would feel very different if we knew that Big Brother was watching-if all the cameras were linked and their control centralized in law enforcement command centers. Feelings of paranoia would be justified because private viewers are much less likely to be upset by behavior that police may find egregious. For example, on the highway there is a distinction between driving so badly that people around you will use their cell phones to call the police, and driving well enough to satisfy the police officer cruising along behind you.

Contemporary sociologists argue that surveillance has the potential to control the population by keeping social groups apart. In their separate case studies of Los Angeles, urban sociologists Steven Flusty and Mike Davis discuss how heightened surveillance has undermined the potential for interaction between individuals and different social groups, and essentially eviscerated democratic public space. Public streets, parks, and plazas have traditionally provided an arena to "synthesize new cultures, alternative ways of living and popular forces occasionally strong enough to upset entrenched status quos," notes Flusty. Access to city spaces should be guaranteed by virtue of citizenship, but it is increasingly becoming a privilege conferred by status. Surveillance implies that users are not to be trusted; "flawed consumers and other undesirables" pick up on the message that the space is monitored and controlled by the corporations that foot the bill for the security towers and cameras at Los Angeles malls. In City of Quartz, Mike Davis described how the Los Angeles Police Department "relentlessly" restricts uses of public spaces and the free movement and association of youths by enforcing juvenile curfews, barricading boulevards, and sealing off entire neighborhoods and public housing projects. Available public spaces are increasingly circumscribed as the wealthy elite of Los Angeles construct gated residential communities to exclude the marginalized and ghettoized underclass and use CCTV to police the boundaries and ensure prompt responses to incursions. Flusty underscores the dystopian potential of a fortress city coupled with maximum surveillance:


Police and the military often use surveillance to intimidate protestors. Recent examples include mass demonstrations against the joint World Bank and International Monetary Fund meeting in April 2000, the mainstream Million Family March in October 2000, and protests against George W. Bush's inauguration in January 2001. According to social psychologists, a participant's self-image can be affected by taping, because being a subject of surveillance is unconsciously associated with criminality. The message is clear: "hang with dissenters, and you'll end up in a police video," Boal observes.

Some have argued that taping political protests is actually beneficial to demonstrators, because it protects them from police misconduct. However, the evidence strongly suggests that police develop strategies to avoid becoming the subject of the "disciplinary gaze" themselves. Norris and Armstrong found in their study of CCTV operations in Great Britain that police officers routinely signaled camera operators with a hand gesture to move the cameras away to avoid being caught on camera while standing with other police officers chatting or scuffling with a suspect. Even if police abuses were filmed, the operators certainly didn't forward the tape to the victim; officers were rarely, if ever, disciplined for their actions.

Widespread surveillance of everyday life can have disturbing effects on our psyche. For example, communications professor Carl Botan's 1996 study of workplace surveillance found that employees who knew they were being watched were more distrustful of bosses, suffered from low self-esteem, and became less likely to communicate. The result: "a distressed work force."

On the other hand, some people claim to feel reassured when using an ATM at night if there is a camera monitoring the transaction. A camera won't actually intervene if someone attacks, but perhaps the awareness of being watched deters some would-be robbers. A neighbor's son was recently mugged in the lobby of my building. The landlord responded by mounting cameras, as well as posting signs giving notice that the building is under surveillance. This seemed an appropriate response to supplement the other "security" features of the building: the vestibule and lobby were always clean, well-lit, and visible to the street through plate glass windows. Cameras don't make spaces as safe as these other options, but they may have a place, say, in particularly high-risk locations, such as ATMs, or low-traffic areas, such as lobbies and elevators. There is a balance to be struck between the need for public space where we can be free of prying eyes and preserving personal safety. Where cameras are limited to private indoor or desolate outdoor spaces (such as parking garages), not just set out to monitor public streets or parks, their use is justifiable. Fair warning that one is being taped should accompany video surveillance. This serves two purposes: one is aware of being watched, and so can avoid any embarrassing displays before the camera; second, it may serve a deterrent function.

Spaces that are accessible, not defensive, will be used by more people, thus rendering them safe for all. To which playground would you rather send your child? The one bristling with surveillance cameras or the one that has none? Electronic patrols signal not safety but insecurity. Public space scholar William H. Whyte cites numerous examples to support his assertion that places "designed in distrust get what was anticipated…. The best way to handle the problem of undesirables is to make a place attractive to everyone else."

Video surveillance and facial recognition are easy to sell to a public more paranoid than ever before about terrorists, murderers, rapists, child molesters, cop killers, robbers, and prison escapees in our midst. What could be less objectionable than catching terrorists? Each intrusion is introduced as being a tool against the pariah group or issue: drug testing, once just for high security jobs, is now being implemented against secretaries at your local Board of Education and students who participate in extracurricular activities; EZ Pass was introduced to alleviate traffic congestion, and is now used to track speeding remotely and issue tickets; it could be used in the future to trace people's whereabouts. After the technology has been implemented with one justification, over little public objection, we can be sure that it will be used for a much wider range of everyday activities and against many more people. Sociologist Gary Marx warns, "Once the new surveillance systems become institutionalized and taken for granted in a democratic society," they can be "used against those with the 'wrong' political beliefs; against racial, ethnic, or religious minorities; and against those with lifestyles that offend the majority."

Contrary to the assumptions of privacy jurisprudence, reasonable people do not intend to waive all rights to privacy when they appear in public. Reasonable people trust that their privacy will be more respected than the law recognizes. If we use a contextual approach to determining the contours of Fourth Amendment privacy in public spaces, we see that constant video surveillance is clearly inappropriate and a violation of the government's mandate to leave people free from unreasonable searches and seizures. Fundamental values of autonomy and freedom of association are crucial to the vitality of a democratic society. And public privacy is essential to guarantee equal access to the benefits of privacy. If it is confined to private spaces, it becomes available only to the affluent. Under this regime, of course, homeless people, who have committed no crime, have no defensible right to privacy. There is a whole range of socially productive interactions that can only happen outside-in particular, anything involving people we don't already know, or anything unplanned, unexpected, or kept secret from the people we live with. Our legal regime should protect our trust and the values that underlie it, not undermine it with unchecked spying.


Molly Smithsimon, a New York City attorney at the Community Service Society, directs an initiative for tenants of federally assisted housing.

top  |  print  |  email